Young v. Red Clay Consolidated School District

Annotate this Case
Justia Opinion Summary

Plaintiffs in this case were residents of Red Clay who were unable to access the polls during a special election held by Red Clay Consolidated School District in February 2015. In the election, residents were asked to approve an increase in the school-related property taxes paid by owners of non-exempt real estate located within the district. Red Clay prevailed in the special election, but Plaintiffs claimed electoral misconduct. The Court of Chancery declared that Red Clay violated the Elections Clause of the Delaware Constitution but did not award any greater relief because the violations did not warrant invalidating the special election. The court reached this result through a balancing of factors, including the dysfunction in Delaware’s system for funding public schools, which led to Red Clay facing a significant deficit without a favorable vote.

Download PDF
IN THE COURT OF CHANCERY OF THE STATE OF DELAWARE REBECCA YOUNG, ELIZABETH H. YOUNG and JAMES L. YOUNG, Plaintiffs, v. RED CLAY CONSOLIDATED SCHOOL DISTRICT, Defendant. ) ) ) ) ) ) ) ) ) ) ) C.A. No. 10847-VCL OPINION Date Submitted: February 23, 2017 Date Decided: May 24, 2017 Richard H. Morse, AMERICAN CIVIL LIBERTIES UNION OF DELAWARE, Wilmington, Delaware; John W. Shaw, Karen E. Keller, Jeffrey T. Castellano, David M. Fry, Nathan R. Hoeschen, SHAW KELLER LLP. Counsel for Plaintiffs. Barry M. Willoughby, William W. Bowser, Michael P. Stafford, Margaret M. DiBianca, YOUNG CONAWAY STARGATT & TAYLOR, LLP, Wilmington, Delaware. Counsel for Defendant. LASTER, Vice Chancellor. In February 2015, Red Clay Consolidated School District (“Red Clay”) held a special election in which residents were asked to approve an increase in the school-related property taxes paid by owners of non-exempt real estate located within the district (the “Special Election”). Red Clay prevailed in the Special Election, with 6,395 residents voting in favor and 5,515 against. The plaintiffs are residents of Red Clay who did not vote in the Special Election because they were unable to access the polls. They filed suit, asserting that Red Clay violated the provision of the Delaware Constitution which guarantees that “[a]ll elections shall be free and equal.”1 They also contend that Red Clay’s actions violated the Due Process and Equal Protection Clauses of the Fourteenth Amendment of the United States Constitution.2 This court previously held that the plaintiffs’ theories stated claims on which relief could be granted.3 This decision only addresses their state law claim. It does not reach their federal claims. The plaintiffs proved at trial that to secure a favorable result in the Special Election, Red Clay violated the Elections Clause. Red Clay held seventy-five events on election day, 1 Del. Const. art. I, § 3 (the “Elections Clause”). See U.S. Const. amend. XIV, § 1 (“No State shall make or enforce any law which shall abridge the privileges or immunities of citizens of the United States; nor shall any State deprive any person of life, liberty, or property, without due process of law; nor deny to any person within its jurisdiction the equal protection of the laws.”). 2 3 Young v. Red Clay Consol. Sch. Dist., 122 A.3d 784 (Del. Ch. 2015) (the “Dismissal Ruling”). 1 in the school buildings that served as polling places, that drew families with children to the polls. The purpose and effect of these events was to reward families of Red Clay students for voting. By Red Clay’s own calculation, at least 6,383 people attended these gatherings. Several of the evening events drew hundreds of people. The Elections Clause, a related constitutional provision, and two related statutory provisions evidence an unwavering Delaware public policy against both overt and covert rewards for voting. When a government provides a targeted reward for voting to a group it believes will favor its position, the election is not “free and equal.” The election day events also had the unfortunate consequence of interfering with access to the polls. The many families who attended the events jammed the parking lots at the schools that served as polling places. The evidence at trial showed that at least some elderly and disabled residents did not vote because they could not find accessible parking. Having heard the Red Clay representatives testify, I am convinced that they did not intend to discriminate against elderly and disabled residents. They recognized that the events would generate crowded parking lots, and they took some steps to mitigate this effect, but they failed to anticipate the serious problems that elderly and disabled residents would face. They also did not monitor the parking places designated for voters, as required by Red Clay’s contracts with the Department of Elections, to ensure that they remained available. Delaware case law, a statutory provision governing electioneering, and evidence of custom and practice in Delaware elections demonstrate that for an election to be “free and equal,” voters must be able to access the polls. An election in which the government obstructs the ability of elderly and disabled residents to vote is not “free and equal.” 2 Red Clay also rendered the Special Election unequal by engaging in four months of one-sided get-out-the-vote efforts. Starting in November 2014, Red Clay aggressively targeted the voters it believed would support the tax increase. Red Clay consciously avoided using communication channels that would inform the public as a whole and make the Special Election a debate. In particular, Red Clay used its access to confidential information about Red Clay families to promote voting by parents of Red Clay students. Red Clay’s assigned each school a goal number of “YES” voters, had each school canvass its parents to find those “YES” voters, and used targeted followed-up communications to get these voters to the polls. Red Clay also employed a variety of other tactics to mobilize student families. Although Red Clay directed some communications to the community at large, they were comparably minimal and generally required by law. The Delaware Supreme Court has held that when a school district conducts a referendum, the “expenditure of public funds in support of one side” must remain “within reasonable limits,” and the school district’s speech should not venture “beyond [a] factual presentation” to the point of “overstatement and emotional appeals.”4 In the Dismissal Ruling, I suggested that societal developments since that decision warranted loosening those restrictions. On the facts of this case, however, the extent and intensity of Red Clay’s targeted campaign speech, particularly when considered in conjunction with the election day events, resulted in the Special Election not being “free and equal.” 4 Brennan v. Black, 104 A.2d 777, 790 (Del. 1954). 3 Although this case focuses on Red Clay’s election-related conduct, it stems from dysfunction in Delaware’s system for funding public schools. The proper operation of that system depends on the regime for determining property values for tax purposes. By statute, the assessed value is supposed to reflect a property’s current market value. In practice, assessments in New Castle County remain pegged to values from 1983. This means that Red Clay’s tax base has remained flat for nearly thirty-five years. Red Clay’s operating expenses have not remained flat for nearly thirty-five years. They increase every year, both because of inflation and because society regularly asks the public schools to take on greater burdens. When the value of the tax base is fixed, the only way to raise revenue is to increase the tax rate. By statute, school districts cannot raise the tax rate unilaterally; they must ask district residents to approve the increase. In this case, without a favorable vote, Red Clay faced a looming deficit. Prevailing in the Special Election was therefore crucial, and the Red Clay administrators were under a great deal of pressure to achieve that result. Unfortunately, their understandable desire to obtain adequate funding to fulfill their mission led them to undermine the electoral process. But their actions must be evaluated in light of the difficult situation they faced. The question for decision in this case is not only whether Red Clay violated the Elections Clause, but also whether those violations warrant invalidating the Special Election. Extensive precedent makes clear that proving electoral misconduct does not lead ineluctably to invalidation. In this case, a balancing of multiple factors convinces me that the Special Election should stand. This decision therefore results in a declaration that Red Clay violated the Elections Clause, but it does not award any greater relief. 4 I. FACTUAL BACKGROUND A three-day trial took place from October 31 to November 2, 2016. The parties introduced 327 exhibits. Eighteen fact witnesses and five expert witnesses testified live. The following facts were proven by a preponderance of the evidence. A. Property Taxes And Public Schools Understanding why Red Clay intervened in the Special Election requires some background knowledge about how Delaware funds its public schools. School districts in Delaware receive the majority of their operating revenue from a combination of state and local funds. State funds come from the General Assembly.5 Local funds come from property tax revenue generated within each school district. “Local funds touch every aspect of the school district budget from employee salaries and benefits, to supplies and materials[,] to maintenance and security and transportation.”6 To generate local revenue, owners of non-exempt real property in a school district pay property taxes on the assessed value of their real estate at a rate set by the school board and approved by residents.7 The amount of available local revenue thus depends on two variables: the assessed value of the property and the tax rate per dollar of assessed value. The Delaware Code provides that “[a]ll property subject to assessment shall be assessed at 5 See 14 Del. C. § 1701 et seq. 6 JX 25 at D2548. 7 See 14 Del. C. §§ 1902-03, 1913-18. 5 its true value in money.”8 The Delaware Supreme Court has held that the concept of a property’s “true value in money” is “the same as its fair market value.”9 “Fair market value” is “the price which would be agreed upon by a willing seller and a willing buyer, under ordinary circumstances, neither party being under any compulsion to buy or sell.” 10 The Delaware Code requires the annual preparation of an assessment roll showing the values ascribed to properties for purposes of taxation.11 In New Castle County, where Red Clay is located, the Department of Land Use is obligated to “assess all property subject to taxation by the County and maintain appropriate records.”12 The Department also must “prepare tax rolls, including those required by any . . . school district.”13 To oversee the process of preparing the assessment roll, the Delaware Code establishes a Board of Assessment Review.14 To emphasize the point that its members are 8 9 Del. C. § 8306(a). New Castle Cty. Dep’t of Fin. v. Teachers Ins. & Annuity Ass’n, 669 A.2d 100, 102 (Del. 1995). 9 10 Id. (citation omitted). See 9 Del. C. § 8301 (titled “Annual Assessment of all assessable property and persons”). Each county follows different procedures and is governed by a different chapter of Title 9. Id. Red Clay is located in New Castle County, so this decision focuses on the procedures there. 11 12 9 Del. C. § 1301(15). 13 Id. § 1301(16). 14 9 Del. C. § 1317. 6 supposed to ensure that property is valued at its “true value in money,” the Delaware Code contemplates a per-property fine for departures from that standard.15 The statutory framework calls for the Department of Land Use each year to “prepare and present to the Board of Assessment Review a copy of the assessment roll for [that] year.”16 The Board of Assessment is charged with hearing appeals by individual property owners who have challenged the assessment of the property.17 The Board of Assessment also is charged with “[r]eview[ing] the methods by which the general manager of the Department of Land Use has established the assessments and the results thereof as reflected by the assessment roll.”18 So far, this system makes sense. Property must be assessed at its “true value in money,” which is synonymous with its “fair market value.”19 The Department of Land Use 15 Id. § 8306(b) (“If any board of assessment, or any member thereof, knowingly and willfully vales or assesses any property upon any other standard than its true value in money, each of the members of the board participating therein shall be fined not more than $100.”). 16 9 Del. C. § 1322(a). 17 9 Del. C. §§ 1318(1) & (2). 18 9 Del. C. § 1318(3). 19 Teachers Ins., 669 A.2d at 102. 7 conducts the annual assessment.20 The Board of Assessment oversees the system to keep everything on track.21 Yet in New Castle County, “property assessments are based upon 1983 property values.”22 That fact is sufficiently astounding to merit repeating. Property assessments are not based on fair market value in the year of the assessment, but rather on their value as of June 1, 1983, nearly thirty-five years ago. That date carries talismanic significance because it was when New Castle County’s last general reassessment became effective for tax purposes.23 20 See 9 Del. C. §§ 1322, 8301. 21 See 9 Del. C. § 1318(3). 22 Office of Assessment, Assessment Process & Applicability, http://www.nccde. org/181/Assessment; accord New Castle County v. New Castle County Bd. Of Assessment Review, 970 A.2d 257, 2009 WL 790360, at *3 (Del. 2009) (TABLE); accord JX 25 at D2548 (“Local taxes are collected by New Castle County and are fixed based on 1983 assessed property values.”); Tr. 645 (Floore). 23 A general reassessment involves a systematic effort to determine current valuations for all the taxable property in the county. This concept contrasts with the annual assessment, when the Department of Land Use generally carries forward existing valuations. As part of this process, the Department of Land Use does update some valuations of specific properties. When issuing a building permit, the Department conducts a property-specific reassessment to reflect the value of the improvements, such as a new deck or finished basement. It then adjusts the reassessment to reflect 1983 values. These changes lead to de minimis increases in the tax base. By going without a general reassessment since 1983, New Castle County falls in between its sister counties. Kent County’s last general reassessment was in 1987. Sussex County’s was in 1974. 8 Reflect momentarily on how much has changed since 1983. Back then, the Governor of Delaware was Pete DuPont. The President of the United States was Ronald Reagan. The Soviet Union still existed, and the federal government regarded it as the “evil empire.” A new Star Wars movie was in theaters, but it was Return of the Jedi. MTV still played music videos, and Michael Jackson’s Thriller made its premier. The two teams in Super Bowl XVII were the Washington Redskins and the Miami Dolphins. On a personal note, I started high school. Property values have changed dramatically since the early 1980s. The following data from judicially noticeable sources provides a sense of the magnitude of the change: In 1983, the House Price Index for New Castle County, published by the Federal Housing Finance Agency, was 159.59. By 2016, the House Price Index had climbed to 499.49. The 2016 figure represents an increase of 340% over the 1983 figure and an annualized gain of 9.44%.24 Measured in 2017 dollars, the inflation-adjusted median sales price of a new home in Delaware in January 1980 was $126,455. In January 2000, it was $187,596.25 In January 2010, the median value of all owner-occupied housing in Delaware (not just 24 See Federal Housing Finance Agency, Housing Price Index Datasets: States (Not Seasonally Adjusted), https://www.fhfa.gov/DataTools/Downloads/Pages/House-PriceIndex-Datasets.aspx#qexe. The base year was 1975, meaning that housing values in that year were assigned a value of 100. The index reached 159.59 in 1983, meaning housing values in New Castle County were 60% higher in 1983 than they were in 1975. 25 See United States Census Bureau, Historical Census of Housing Tables: Home Values, available at https://www.census.gov/hhes/www/housing/census/historic/values. html. The Census Bureau presents these values in 2000 dollars. For consistency with present data, this decision recalculates them in 2017 dollars using the Consumer Price Index. See U.S. Dep’t of Labor Bureau of Labor Statistics, CPI Inflation Calculator, https:// www.bls.gov/data/inflation_calculator.htm. 9 new homes) was $273,000.26 The 2010 figure represents an increase of 116% over the 1980 figure and an annualized gain of 3.86%. That was seven years ago. Measured in 2017 dollars, the inflation-adjusted median sales price of a new home in the United States in January 1980 was $196,331. In January 2000, it was $235,214, and in January 2010, it was $244,535. In January 2017, the median sales price of a new home was $308,000.27 The 2017 figure represents an increase of 56.9% over the 1980 figure and an annualized gain of 1.54%. Measured in 2017 dollars, the median value of an owner-occupied housing unit in New Castle County in 2000 was $195,652. In 2015, it was $260,600.28 The 2015 figure represents an increase of 33.2% over the 2000 figure and an annualized gain of 2.2%. Yet despite these significant gains, the assessed value of New Castle County’s underlying tax base remains flat. In Red Clay, the average assessed value of a residential property is stuck at the 1983 level of $80,100.29 26 See United States Census Bureau, Fact Finder: Median Housing Value of OwnerOccupied Housing Units (Dollars), https://factfinder.census.gov/faces/tableservices/jsf/ pages/productview.xhtml?src=bkmk. 27 See United States Census Bureau, Median and Average Sales Prices of New Homes Sold in United States, available at https://www.census.gov/construction/nrs/pdf/ uspricemon.pdf. The Census Bureau’s figures do not account for inflation. This decision recalculates the values in 2017 dollars using the Consumer Price Index. See U.S. DEP’T OF LABOR BUREAU OF LABOR STATISTICS, CPI Inflation Calculator, https://www.bls.gov/ data/inflation_calculator.htm. 28 See United States Census Bureau, American Fact Finder Guided Search: Selected Housing Characteristics, https://factfinder.census.gov/faces/nav/jsf/pages/guided_search. xhtml. 29 JX 25 at D2551. 10 It should be obvious that assessing properties as of 1983 is a far cry from determining their “true value in money” as of the current year.30 The Delaware Supreme Court indicated in 1977 that at some point, assessed values could become so stale as to be statutorily infirm.31 But to date, no one has brought a county-wide challenge.32 30 9 Del. C. § 8306(a). 31 See Bd. of Assessment Review for New Castle Cty. v. Stewart, 378 A.2d 113, 116 (Del. 1977) (observing that no one had argued that “an unreasonable period of time has passed since the last general assessment was made in 1970”). Technically, there is no statutory section in the Delaware Code that requires New Castle County to conduct a new general reassessment. Nevertheless, multiple sections anticipate that general reassessments will take place. Several establish mechanisms for capping at 10% the maximum possible tax increase that could result from a general reassessment. See 14 Del. C. § 1916(b) (providing that if “a subsequent general reassessment of all real estate in the county changes the total assessed valuation of the school district, the local board of education of each such local school district shall calculate a new real estate tax rate which, at its maximum, would realize no more than 10% increase in actual revenue”); 14 Del. C. § 2601(c) (providing similar procedure for adjustment of tax rate based on general reassessment that changes total assessed value of county vocational-technical high school district or county vocational-technical center district); 14 Del. C. § 1707(b)(11) (providing for possibility of general reassessment when defining term “total full valuation” for purposes of Division III equalization funding that state provides to school districts). Another authorizes the New Castle County Council to facilitate challenges by property owners following a general assessment. See 9 Del. C. § 1322(d) (providing that in any fiscal year in which New Castle County proposes to implement a general reassessment, “the County Council may by ordinance establish appropriate and reasonable time periods for the filing of exemption applications; submission, inspection and certification of assessment rolls; notices of assessments; appeals from such assessments; and any other requirements relating to the implementation of the general reassessment”). Of course, an obligation to conduct periodic general assessments might be implicit in a statutory requirement to value each property at its “true value in money.” 9 Del. C. § 8306(a). 32 Individual litigants have objected to the staleness of the 1983 general assessment when challenging the valuations of their own properties. In that context, however, the Delaware courts have refused to intervene because the Delaware Constitution requires uniformity in tax assessments. See Del. Const. art. VIII, § 1 (“All taxes shall be uniform upon the same class of subjects within the territorial limits of the authority levying the 11 Ironically, a well-intentioned state-level scheme for providing additional funds to less-wealthy school districts creates a powerful disincentive for any civic-minded official tax[.]”). To ensure uniformity, New Castle County uses the base year method. See, e.g., Stewart, 378 A.2d at 116. As noted, New Castle County currently uses a base year of 1983. Commerce Assocs. v. New Castle Cty. Office of Assessment, 2016 WL 3457820, at *8 (Del. Super. Apr. 1, 2016), rev’d on other grounds, 2017 WL 1337318 (Del. Apr. 11, 2017). When a property owner asserts that a 1983 valuation is outdated and a property should have a more accurate, current value, the principal of uniformity comes into play and prevents any single property owner from obtaining a valuation different from the base year. Id. Using this reasoning, a series of decisions have rejected challenges to valuations based on 1983 values. See, e.g., Mazen v. City of Dover Bd. of Assessment Appeals, 2016 WL 520996, at *2 (Del. Super. Jan. 22, 2016); RRHC Wilmington, LLC v. New Castle Cty. Office of Finance, 2014 WL 2538886, at *8-9 (Del. Super. May 30, 2014); Bailey v. Bd. of Assessment Review, 2004 WL 1965867, at *5 (Del. Super. Aug. 19, 2004). The uniformity defense might not apply in a case which asserted that New Castle County is currently violating the statutory requirement to value property at its “true value in money” by continuing to use a base year of 1983. Assuming the suit asserts that 1983 valuations are so stale that New Castle County must conduct a general reassessment, then the values going forward would be tied to the general reassessment and made uniform as of a new date. The 1983 base year would no longer be relevant, just as the prior base year of 1974 was no longer relevant after the 1983 general reassessment. Moreover, nothing about the base year method requires that the base year be the same year when the last general reassessment took place. Just as the county currently uses the Consumer Price Index to factor property values back to 1983, it could use the same metric (or a more realestate-specific metric) to bring values forward to a later year or, most appropriately, the current year. See New Castle Cty. v. New Castle Cty. Bd. of Assessment, 970 A.2d 257, 2009 WL 790360, at *1 (Del. 2009) (TABLE) (explaining that the County “used the Consumer Price Index (CPPI) to factor [a 2006-2007-year valuation] back to 1983”). One reason for the lack of a county-wide challenge might be the absence of a motivated plaintiff. A successful challenge that generates a new general assessment almost certainly would cause most assessed property values to increase, subject to the 10% statutory cap. Few people like paying higher taxes. Even fewer are motivated to file suit to fix a dysfunctional system where success means higher taxes for everyone. A plaintiff with more systemic interests, such as restoring the integrity of the school-funding mechanism, might have different motivations. So too might an institutional plaintiff that could represent the interests of minors currently disadvantaged by the system. 12 to take the lead in reassessing property values. Each year, the Department of Education recommends that the General Assembly authorize an aggregate state-wide appropriation based on the number of “units of pupils” in each school district.33 The General Assembly allocates appropriations in three buckets. Division I funds pay for administrators, teachers, and other personnel.34 Division II funds primarily pay for textbooks, furniture, and other classroom equipment, but can be used for any lawful purpose.35 Division III funds are budget equalization funds that are allocated based on a formula designed to provide matching funds to less wealthy districts.36 Because districts that generate less local funding receive more Division III funding, there is a disincentive for any county to take the lead in reassessing property values. Although a general reassessment would yield more local funds, the county’s school districts would receive less state funds. Everyone has a reason to keep the existing values in place. See 14 Del. C. §§ 1702(b), 1703-04. One “unit” is 16.2 students enrolled in kindergarten, first grade, second grade, or third grade, or 20 students enrolled in fourth through twelfth grade. See 14 Del. C. § 1703(d)(2)-(3); Tr. 655 (Floore). 33 34 14 Del. C. § 1702(c). 35 14 Del. C. §§ 1702(d), 1706. 36 See 14 Del. C. § 1707. The formula is complex but essentially pegs state funding to a combination of each school district’s “effort index” and “ability index.” 14 Del. C. § 1707(c). The effort index is the ratio by which the district’s tax burden exceeds the average tax burden across the state. See 14 Del. C. § 1707(b)(3). The ability index is the district’s aggregate property value. As the effort index rises, the State’s contribution rises, but as the ability index rises, the State’s contribution falls. See 14 Del. C. § 1707(c) (“The State share per unit is equal to the authorized amount times the effort index times the quantity of 1 minus .75 times the ability index[.]”). 13 The upshot is that the value of the underlying tax base in each school district remains flat. But the cost of running a school district does not. Inflation reduces the purchasing power of a school district’s budget every year. Even if a school district does not introduce any new initiatives and just maintains the status quo, the absence of regular and systematic reassessments inevitably generates a funding gap. At a macro level, since 1983, the purchasing power of a school district’s tax base in constant dollars has declined by nearly 60%. Put differently, it requires $2.44 in 2017 to buy the same amount of goods and services that $1.00 would buy in 1983.37 This leaves school districts with one lever to pull. The Delaware Code empowers the school board for each district to set the amount of tax per dollar of assessed value that a property owner must pay.38 Using this authority, the school board can increase the tax rate so that the same assessed value generates more revenue. But the school board cannot levy the tax unilaterally. The school board first must “call a special election to be held at the polling place or places designated by the Department of Elections conducting the election.”39 The outcome of the special election determines whether the tax can be levied.40 See U.S. Dep’t of Labor Bureau of Labor Statistics, CPI Inflation Calculator, https://www.bls.gov/data/inflation_calculator.htm. 37 38 See 14 Del. C. § 1902. See Tr. 646-47 (Floore explaining why, under current system, a referendum is “inevitable”). 39 14 Del. C. § 1903. 14 Del. C. § 1911 (“If the majority of the votes cast at the election . . . shall be for additional tax, the tax shall be levied and collected as provided in this chapter.”). Technically there are other, smaller tax categories that the board can raise unilaterally, 40 14 The relentless effect of inflation combined with the increased expectations and demands that our society places on public schools means that school districts in Delaware must regularly seek tax increases from their voters. Generally speaking, in Delaware, a school district needs to prevail in a referendum every three to five years.41 The frequency of tax referendums generates negative reactions. Some residents object as a matter of principle to having their taxes raised. More object if they think their tax dollars are not being used wisely. Delaware’s complex system for funding public schools is not easily understood. The natural reaction of some citizens to regular requests for tax increases is to suspect that school officials are wasting money. A review of decades of referendums reveals that they often fail the first time, then pass when presented a second time after supporters recognize the need to make it a priority to vote.42 without the approval of residents. Tr. 645, 703 (Floore identifying these categories). They are relatively immaterial. To avoid overly complicating an already complicated area, this decision sets them aside. 41 Tr. 660 (Floore testifying that contemplating a referendum every four years is “pretty standard for every school district” in Delaware). Appendix A compiles a list of school referendums in Delaware, since 1980, drawn from publicly available articles in The News Journal. It includes both operating expense referendums and capital expense referendums, which must be held separately. See 14 Del. C. § 2122. It also includes referendums seeking approval to transfer tax revenue from one account to another. See id. The appendix, which is likely under-inclusive, indicates that on average, Delaware school districts have obtained approval for 7.3 tax increases since 1980. That amounts to one tax increase every five years. Counting only operating referendums, Delaware school districts has obtained approval for an average of 4.4 tax increases since 1980. Red Clay is slightly above average on both counts, having obtained approval for eight tax increases, including six for operating expenses, during that period. 42 See Appendix A. Of the operating referendums for which vote figures are publicly available, sixty-six were approved and fifty-one defeated. When a referendum failed, the 15 The Delaware public schools, including Red Clay, recognize the problems created by this system. When responding to questions from legislators about Red Clay’s conduct during the Special Election, Superintendent Mervin Daugherty started by calling out the underlying problem with Delaware’s mechanism for funding public schools: It is important to note that the election process and requirement for a Referendum is one that is set by the legislature, not school districts. The public school districts believe the established system is an ineffective way to fund education, especially in light of recent years in which the State has cut school district frequently prevailed in a second referendum held shortly thereafter. The second referendum nearly always generated higher turnout. For example: In Brandywine School District, which borders Red Clay, voters defeated an operating referendum in March 1987 by a vote of 6,841 to 6,313. In October 1987, the voters approved it by a vote of 11,938 to 6,589. In 1993, the voters defeated an operating referendum by a vote of 9,827 to 8,866. In 1994, they approved it by a vote of 14,579 to 10,669. In April 2007, they defeated an operating referendum by a vote of 4,800 to 4,322. In June 2007, they approved it by a vote of 7,584 to 6,305. In June 2007, they defeated an operating referendum by a vote of 3,892 to 3,729. In March 2016, they approved it by a vote of 9,500 to 5,780. In Colonial School District, which also borders Red Clay, voters defeated an operating referendum in 1992 by a vote of 3,900 to 2,988, then approved it in 1993 by a vote of 7,082 to 5,228. In February 2013, they defeated an operating referendum by a vote of 2,484 to 925, then approved it in June 2013 by a vote of 3,005 to 2,938. In Christina School District, which also borders Red Clay, voters twice defeated an operating referendum in 2015, first by a vote of 6,076 to 2,119, then by a vote of 5,968 to 5,074. In 2016, voters approved the referendum by a vote of 6,770 to 6,625. In Red Clay itself, voters defeated an operating referendum in 2007 by a vote of 6,220 to 4,822, then approved it in 2008 by a vote of 8,550 to 7,414. See Appendix A. It makes intuitive sense to me that busy parents who support their children’s schools might nevertheless fail to vote in a referendum, expecting it to pass, then make an extra effort to vote after seeing the initial referendum fail. I confess to having been guilty of this on at least one occasion. 16 funding for public schools. We recognize, however, that we all must work within the confines of the existing law.43 I would go one step further. The referendum process results from a combination of a legislative framework and a property assessment system that currently does not function as the statutes contemplate. In my view, Delaware’s statutory framework is not supposed to force school districts into a vicious cycle of regular referendums. If New Castle County conducted periodic general reassessments—which the Delaware Code appears to contemplate and which seems to have been the practice until 1983—then the underlying tax base would rise as property values increased. The same tax rate would generate more money for the school district, and the district would not have to seek a tax increase as frequently. Or if New Castle County simply used the current year as the base year and brought values forward to the current year rather than back to 1983—whether using the Consumer Price Index or some other measure—then at least the values would increase by inflation year-over-year, and school districts would not have to call referendums just to keep up. Or the General Assembly could solve the problem with legislation that would create a more serviceable framework.44 But without action by the political branches, school districts must resort to the only tool they have: the referendum process. 43 JX 176 at D14675. 44 The General Assembly could devise a wide variety of solutions. Like the United States Census, a statute might require a general assessment in every tenth year. Cf. U.S. Const. art. 1, § 2. Or to give flexibility to the counties, a statute might contemplate a general assessment not less frequently than every tenth year. Nor is there any particular magic to a 17 B. Red Clay Is Forced To Seek A Tax Increase. In summer 2014, Red Clay’s senior administrators concluded that the district needed more local tax revenue to cover its operating expenses. They did not make this decision lightly. They decided to “go to referendum” (the colloquial phrase) only because there was no alternative. ten-year period. Before the 1983 assessment, New Castle County conducted a general assessment that became effective in 1974, nine years earlier. See McGinnes v. Dep’t of Fin., 359 A.2d 166, 167 (Del. 1976). Before that, it appears that New Castle County conducted a general assessment that became effective in 1970, just four years earlier. See Stewart, 378 A.2d at 117. A statute governing mobile homes mandates periodic reassessment every five years. See 9 Del. C. § 8351. Mandatory periodic reassessments also would help mitigate the incentive to game the Division III funding formula. Other structures could be more responsive to market values. For example, a statute could require a new general assessment if average sale price deviated by a particular amount from assessed values. Cf. 14 Del. C. § 1707(b)(11) (using ratio of assessed value to sales price to adjust from assessed value to “total full valuation”). Or a statute could use the most recent arms’ length sale price as the assessed value, then equalize property values for constitutional purposes by applying an adjustment factor to bring the sale price current to the existing calendar year. One also could envision more significant changes to the overarching system for funding public schools. For example, Delaware’s vocational schools are authorized to raise tax rates without a referendum, subject to maximum caps imposed by the legislature. See 14 Del. C. §§ 2601-02. Obviously these are not the only possibilities. Minds more sophisticated in legislative and property tax matters than mine can doubtless come up with superior solutions. The point is that many alternatives are available. The options are not limited to either (i) conducting general reassessments annually or on some other short-term basis, which would be both cost prohibitive and wasteful, or (ii) maintaining the current status quo of doing nothing since 1983. 18 Red Clay had held its last operating referendum in 2008, when voters approved a twenty-five cent increase phased in over three years.45 Because of New Castle County’s approach to property tax assessments, Red Clay’s local operating revenue had not increased materially since 2010. Over those four years, inflation deprived Red Clay’s local operating revenue of 5% of its purchasing power.46 Meanwhile, Red Clay’s student population grew by 10%.47 Measured by units of pupils, Red Clay’s enrollment grew by 17%.48 Red Clay received approximately 40% of its operating budget from local funds and 60% from the state.49 The state funds had not made up the growing local shortfall. By 2014, Red Clay received less discretionary state funding than it had in 2008.50 To operate under these conditions, Red Clay cut costs. By 2014, however, Red Clay projected that without additional revenue, the district would end 2016 with a deficit. By 2018, the cumulative deficit would reach $24.7 million.51 To balance its budget without new revenue, Red Clay would need to cut approximately $9 million in recurring expenses, 45 JX 25 at D2548. Red Clay held a separate capital referendum in 2012 to fund construction of a new school and building improvements. Id. at 2552; Tr. 644 (Floore). See U.S. Dep’t of Labor Bureau of Labor Statistics, CPI Inflation Calculator, https://www.bls.gov/data/inflation_calculator.htm. 46 47 JX 25 at D2548; Tr. 646, 655-56 (Floore). 48 JX 258 at D2169. 49 Tr. 645 (Floore). 50 JX 27 at D2602; Tr. 651-52 (Floore). 51 JX 258 at D2176; Tr. 661-63 (Floore). 19 representing 15% of the operating budget.52 This would require eliminating at least forty teachers. It also would mean fewer school resource officers, reading specialists, and other staff positions. The cuts would affect virtually every aspect of programming, including after-school sports, technology programs, and arts offerings.53 After evaluating their options, Red Clay’s senior administrators concluded that the district had “stretched as long as we can.”54 They recommended that the Red Clay Board of Education (the “School Board”) call a special election. In October 2014, the School Board scheduled the Special Election to take place on February 24, 2015. The purpose of the Special Election was to obtain approval from district residents to raise the tax rate on non-exempt real property by a total of thirty-five cents per $100 of assessed value. The proposal initially called for the rate to rise by $0.25 in 2016, $0.05 in 2017, and $0.05 in 2018. It was later modified to $0.20 in 2016, $0.10 in 2017, and $0.05 in 2018.55 At the time, the owner of an average taxable parcel in Red Clay paid $1,419 per year in school-related property taxes. After the three-year phase-in, the owner of an average parcel would pay approximately $280 more per year, or roughly $23 more per month. The proposal thus contemplated an approximately 20% increase in the school-related taxes paid 52 JX 25 at D2554. 53 JX 27 at D2607; Tr. 665 (Floore). 54 JX 25 at D2548; see also JX 27 at D2602. 55 JX 324 at D18441-42; PTO ¶ 8; Tr. 323 (Johnson); Tr. 668-69 (Floore). 20 by the average property owner. The first-year increase would generate an incremental $10.5 million for Red Clay. The second-year increase would generate another $5.2 million, and the third-year increase would add another $2.6 million. When fully implemented, the tax increase would yield $18 million annually in operating funds.56 Once a referendum is called, the Department of Elections designates buildings to be used as polling places.57 When selecting polling places, the Department of Elections gives “prime consideration” to the suitability, convenience, and accessibility of the locations for voters.58 The statute identifies public schools as recommended polling places. 59 The Department of Elections designated twenty-five polling places for the Special Election.60 Twenty-three were Red Clay public schools.61 56 See JX 283; Tr. 681 (Floore). 57 14 Del. C. §§ 1072(b), 1903. 58 14 Del. C. § 1072(b). 59 Id. 60 Tr. 408 (Lippincott). 61 See PTO ¶ 26. Appendix B contains a table listing the schools designated as polling places and the abbreviated names that this decision uses to refer to the schools. 21 C. The Campaign Team In Delaware, school tax referendums frequently fail.62 In 2007, a Red Clay referendum failed. The district made $8 million in cuts and took out a loan to cover expenses. The district held its last successful referendum the following year.63 Red Clay’s financial situation made the Special Election a high stakes matter.64 Red Clay’s administrators thought that obtaining a successful vote was an “uphill campaign by its very nature.”65 Not surprisingly, given the stakes, Red Clay planned and carried out a campaign to secure passage of the tax increase. Superintendent Daugherty was the final decision-maker, 62 See Appendix A. 63 Tr. 643-44 (Floore). Tr. 184-86 (Nash telling reporter that Red Clay “had a lot at stake” in the referendum); JX 81 at D11596 (teacher noting, “We NEED this referendum to pass. PERIOD”); id. at D11597 (principal noting, “This referendum is very important to ALL of us in Red Clay”); JX 89 (principal stating, “The success of this referendum is VITAL”); JX 96 at D4599 (principal stating the referendum “is critical”); JX 119 (Superintendent Daugherty noting, “There is a lot riding on this one vote”); JX 176 at D14676 (Superintendent Daugherty describing the referendum as concerning “critical funding for our school district”); JX 233 (principal’s message to parents stating that with a failed referendum, “[c]uts would be drastic and across the board”); JX 242 (Superintendent Daugherty describing the referendum as “so important”); JX 300 (president of School Board describing referendum as “critical”). 64 JX 176 at D14677; accord Tr. 721 (Floore: “[A]sking people to increase their taxes is an uphill battle. By its very nature, anybody who asks to do that is at an extreme disadvantage.”); see also JX 119 (“[R]aising taxes is about the hardest thing to ask any community.”); JX 242 (“[I]t is not easy to ask our community to raise taxes.”); JX 259 (“A referendum is not an easy task . . . .”); JX 310 at D8612 (“Passing a referendum isn’t easy . . . .”); JX 50 (“We anticipate a close race and your support of Red Clay schools could make all the difference.”). 65 22 but he delegated primary day-to-day responsibility to three senior administrators: Ted Ammann, the Assistant Superintendent for District Operations; Patti Nash, the Public Information Officer, and Jill Floore, the Chief Financial Officer.66 Each administrator’s role in the campaign matched their job with the district. Ammann took charge of operations. He ensured that the school principals supported the campaign, instructed them to develop and submit school-specific referendum plans, reviewed and signed off on their plans, and oversaw their implementation. He mobilized Red Clay personnel to volunteer for various referendum-related projects, including making get-out-the-vote calls, stuffing report card envelopes with referendum materials, and putting out pro-referendum signs.67 He caused Red Clay personnel to generate the student call lists that the callers used.68 He also dealt with a variety of logistical matters.69 He was See, e.g., Tr. 179 (Nash testifying she was “at the center of the referendum organizing at Red Clay” and “[o]ne of its leaders”); JX 90 (Daugherty delegating planning of event on day of Special Election to Ammann); JX 133 (Daugherty delegating distribution of pro-referendum video clips to Nash); JX 169 (Daugherty complementing senior administrators for their work on Special Election); Daugherty Dep. 17-20, 23-24, 31 (explaining that he has “never been the point person” for a referendum; identifying Ammann and Nash as the lead personnel for special elections). 66 See, e.g., JX 72 (arranging “clerical help” for stuffing report cards); JX 145 (organizing placement of pro-referendum signs). 67 See JX 84 at D19668 (“All building principals received an updated call list from Ted [Ammann] this week.”); JX 112 (Ammann directing preparation of call lists and describing them as “critical”). 68 69 See, e.g., JX 52 (ensuring that pro-referendum signs were ordered); JX 62 (arranging transportation for Red Clay personnel to volunteer on election day); JX 130 (ordering pro-referendum stickers). 23 in charge of election day arrangements and interactions with the Department of Elections, including arrangements for parking.70 Nash took charge of communications. She developed and ran a telephone campaign to identify likely “YES” voters among Red Clay student households, then follow up with those likely “YES” voters to get them to the polls.71 She organized and ran a social media campaign.72 She spearheaded the creation of campaign materials, including proreferendum video clips.73 She also helped organize a pro-referendum rally on the night before the Special Election.74 Floore provided expertise regarding Red Clay’s finances. Internally, she served as a third key decision maker. Externally, she explained the district’s finances and consequent need for the Special Election to numerous groups and the media.75 The Red Clay administrators did not openly lead the campaign. They believed there were limitations on what district employees could do, such as not directly asking residents 70 See, e.g., JX 88. 71 Tr. 176, 189 (Nash). 72 Id. at 176. 73 For video clips, see JX 39; JX 40; JX 57. For other materials, see JX 35 (providing scripts for telephone campaign); JX 86 (distributing referendum fact sheet to be placed in report cards); JX 97 (distributing draft Steering Committee email). 74 See JX 59; JX 60; JX 72. 75 Tr. 673-75 (Floore describing her role); see JX 267 (listing meetings at which Special Election was discussed); see, e.g., JX 83; JX 154. 24 to vote “YES.”76 They also believed that by establishing a broader organization that included Red Clay parents, they could expand the campaign’s reach and generate greater support.77 The administrators therefore formed a Referendum Steering Committee.78 To chair the committee, they recruited two community members: Yvonne Johnson, whose 76 See, e.g., Tr. 189 (Nash testifying that district employees could not ask residents to vote “YES” while on district time); Tr. 191 (Nash testifying she was “told that District communications, our newsletters, our website, could not specify how to vote”); JX 65 (principal noting that parents who are members of the parent-teacher association can “say to vote yes”); JX 78 (Nash stressing that a teacher could not say “Vote Yes” or post information promoting a “YES” vote on a school page or during the school day, but could do so on a personal page, on their own time); JX 79 (Superintendent Daugherty correcting assistant principal who planned to use faculty meeting time for teachers “to make parent phone calls in regards to the referendum”); JX 81 (teacher noting that district could not require staff to make pro-referendum calls or engage in pro-referendum activities); JX 84 (principal informing parent, “I have some ideas as to ways you can help – since we (as employees) can’t tell people to vote yes, but you as a parent can certainly do so. : )”); JX 108 (parent stating she was “under the impression that [parents] were making calls as well [as teachers] because we could actually ask [voters] to vote yes”); JX 140 (principal telling staff, “While we can’t tell people to vote yes, we can say how passing the referendum will help children”); JX 150 at D11328 (parent stating, “I was under the impression we were making calls as well because we could actually ask them to vote yes”). Contrary to the premise that staff members could not ask parents to vote yes, Red Clay communications frequently made this ask. See, e.g., JX 144 (“Your students will benefit from the following with a Vote ‘For’ . . .”); JX 247 (“Vote ‘yes’ to keep Skyline a school of excellence!”); Tr. 630 (Ammann agreeing that the Skyline principal asked families to “Vote ‘yes’”); see also JX 233 (“WE NEED YOUR VOTE!”); JX 259 (letter telling recent Red Clay graduates about the “opportunity to support . . . your . . . school district by voting YES”); id. (“[C]ome to the polls to vote Yes on February 24 . . . .”). 77 Tr. 193-94 (Nash explaining that having communications come from parents would make them seem “more real” and would “resonate more” with voters). 78 See JX 16 (Floore emailing potential committee members); Tr. 177 (Nash agreeing that “part of [her] job as public information officer for the District was to sit on the steering committee”). 25 children had attended Red Clay schools, and Nate Schwartz, whose children were still attending Red Clay schools.79 They also formed a political action committee called “Friends of Red Clay Referendum,” chartered for the purpose of “rais[ing] funds to support successful referenda in the Red Clay school district.”80 During and after the campaign, Red Clay representatives tried to depict these organizations as grassroots, parent-led efforts rather than something Red Clay engineered.81 In fact, the entire campaign was a Red Clay operation. Of the nineteen members of the Steering Committee, fourteen were current Red Clay employees: five were Red Clay administrators, three were principals, five were teachers, and one was a current member of the School Board. Of the five who were not current Red Clay employees, one was a retired teacher and former president of the School Board.82 The treasurer of the political action committee and the signatory on all of its filings was a Red Clay employee,83 79 See JX 193; Tr. 322 (Johnson). 80 PTO ¶ 31; see JX 41; JX 68. Tr. 324 (Johnson: “So the steering committee was a grassroots effort to educate the community [about] why we needed to go to referendum and to get the support to get the referendum passed.”); JX 50 (Red Clay communication to residents stating, “This is a community led campaign . . .”); JX 176 at D14675 (Superintendent Daugherty telling legislators, in response to inquiries about election misconduct, that campaign was “led by parents”); see also JX 154 (News Journal article reporting that “a steering committee chaired by parents” was “working to ‘get the word out’”); JX 172 (parent claiming in response to legislative inquiry that Red Clay effort was a “grassroots campaign”). 81 82 See JX 25 at D2546; Tr. 177-78 (Nash). 83 PTO ¶ 32. 26 and Nash drafted the invitation for its fundraiser.84 Internally, Red Clay personnel did not distinguish between the district administration and the Steering Committee. The Baltz principal testified that “the steering committee and the district kind of meshed together.”85 The core day-to-day leadership team for the campaign was Ammann, Nash, Floore, Johnson, and Schwartz.86 When push came to shove, Ammann, Nash, and Floore were the day-to-day decision-makers, with Superintendent Daugherty having final authority.87 D. The Overall Campaign Strategy To obtain a favorable vote, the Red Clay team developed a campaign strategy that involved identifying groups that were likely to favor the tax increase, then engaging in 84 See JX 30. 85 Tr. 114 (Penoyer). See Tr. 348-49 (Johnson testifying that “the core group . . . was [herself], [her] co-chair, and . . . Pati Nash, Jill Floore, and Ted Ammann,” and that the “five . . . coordinated closely on every step of the referendum process”); see also, e.g., JX 59 (reflecting decision by core group after communications among group); accord JX 67; JX 90; JX 123. The team also benefitted from the input of Richard Przywara, a professional campaign consultant, who provided advice as a volunteer. Tr. 202-03 (Nash). 86 87 By making this finding, this decision does not mean to diminish the importance of the parents and community members who volunteered. The record demonstrates that many parents and community members worked extremely hard and contributed meaningfully to achieving the outcome that Red Clay sought. In particular, Johnson and Schwartz devoted many hours to the referendum. Johnson also participated actively in the litigation and testified at trial. In my view, this type of community involvement and support for the public schools is highly commendable. In defending its election-related conduct, however, Red Clay sought to use the Steering Committee and the involvement of parents and community members to diminish its role as a state actor. The record demonstrates that in substance, Red Clay ran the campaign. 27 efforts to convert those likely supporters into actual votes.88 In crafting and carrying out the campaign, the Red Clay team viewed individuals with ties to the Red Clay schools as more likely to support the tax increase. In particular, they believed that parents of current Red Clay students would be supportive.89 Their reasoning was straight-forward: Parents likely want a good education for their children, so they should be willing to support a tax increase to achieve that goal.90 The Red Clay team appears to have adopted this view as a matter of common sense, based on their intuitions about human nature and their experience from past referendums in which mobilizing the parent vote provided the key to success.91 Data that the Red Clay team gathered during the campaign confirmed their assessment.92 Tr. 181 (Nash testifying that Red Clay prioritized “the people you think are the most likely to be positive voters”); id. (Nash testifying that in its telephone campaign, Red Clay “was targeting or prioritizing positive voters”). 88 See Tr. 180 (Nash testifying that “there is a belief that parents are more likely to support the referendum”); Tr. 351 (Johnson testifying to her understanding “that very few parents would oppose the tax increase”); id. (Johnson confirming that she “wanted parents in the schools on the day of the referendum because [she] wanted them to vote”); Tr. 49798 (principal of Richardson Park stating, “I believe that if [parents] came out, they would vote yes”); Tr. 638 (Ammann testifying that “[t]he referendum steering committee had talked about the importance of contacting parents, and that would be what they would do”). 89 90 See Tr. 351-52 (Johnson explaining that parents should want to support their children’s schools). See JX 242 (“The 2008 referendum passed for one reason—because parents and schools worked together. The same holds true today.”); JX 169 (Superintendent Daugherty observing after Special Election that “[s]chool leaders and staff worked extremely hard to gain parent support”); JX 193 (Johnson observing before Special Election that “we will not pass this without support from our charter and choice parents”); see also JX 97 (“In the past, referendums have failed because of apathy.”). 91 92 See JX 303 (results from electronic survey of McKean parents showing support); JX 87 (Skyline principal noting that parent calls yielded sixty-six expected “YES” votes 28 At the same time, the Red Clay team viewed elderly and retired residents as less likely to support the tax increase.93 Their reasoning was again straight-forward: Many elderly and retired residents live on fixed incomes, so they cannot easily accommodate additional expenses. The Red Clay team appears again to have adopted this view as a matter of common sense, based on their intuitions about human nature and their experience from other referendums in which elderly and retired residents vocally opposed tax increases. 94 and five expected “NO” votes”); JX 99 (Heritage secretary informing Johnson that parent calls yielded 309 “YES” votes); JX 148 (email to dean of Cab Calloway reporting that parent calls yielded nine “definite yes votes” and “zero no votes”); JX 313 at D20468 (Johnson reporting to team ten days before Special Election that “[p]reliminary numbers we are getting from schools look very positive, we are trending in the right direction and getting very few opposition votes from our parents”). See, e.g., Tr. 217 (Nash agreeing that she believed that seniors generally “would be more likely to vote against a referendum”); Tr. 675 (Floore explaining that at the initial Steering Committee meeting, one of the speakers gave an example of a senior citizen on a fixed income as a likely “NO” voter with whom there would be “no point” in engaging); JX 310 at D8610 (Johnson stating her belief that the campaign would “not . . . reach out to [retired] folks as they could bring out the no vote”). 93 See Tr. 179 (Nash testifying that the “conventional wisdom” is that “seniors vote against referendums”); id. at 206 (same); see also JX 104 (Nash asking for input on letter that “kind of mention[s] the senior vote, without saying seniors”); JX 124 (delaying submission of supportive letters to the editor because of concerns that “the letters might ‘wake up’ the ‘no’ voters”); JX 134 (debating whether “the ‘pro’ letters [to the editor] would wake up even more seniors??”); JX 309 at D8598 (Johnson expressing concern about speaking with “retired folks when I thought we were not going to reach out to these folks as they could bring out the no vote”). 94 The Red Clay team’s views found support in the reaction to a proposed reduction in a tax credit that Delaware provides to seniors. Recognizing that many senior citizens live on a fixed income and have a limited appetite for paying school-related taxes, the State of Delaware has a program that gives seniors a credit on their state income taxes for amounts paid for school property taxes. Under the 2014 status quo, Delaware provided a credit for 50% of the school-related taxes paid by a property owner 65 years of age or older, up to a 29 Importantly, the Red Clay team did not adopt a simplistic view in which they expected every parent to vote “YES” and every elderly resident to “NO.” They recognized that many factors affect how an individual votes. A parent might oppose the tax increase because of a negative experience with a public school, personal financial limitations, or a strong ideological opposition to tax increases. Or an elderly voter might support the tax increase because of a positive experience with a public school, ample income, or a civicminded belief that funding public education benefits the community. 95 But the Red Clay team understood that there are tendencies in the electorate, and they wanted to mobilize the groups that were most likely, on average, to support the tax increase. At bottom, the Red Clay team believed that by increasing the number of parents who voted, they increased their chances of success. They likewise believed that by decreasing—or at least not increasing—the number of seniors who voted, they increased their chances of success.96 cap of $500. In January 2015, then-Governor Jack Markell proposed a budget for the 201516 fiscal year that would have reduced the income tax credit, either by lowering the percentage to 25% or by reducing the cap to $250. See Jon Offredo & Jonathan Starkey, Senior Tax Breaks Hit in Budget Proposal, The News Journal, Jan. 30, 2015, at A1, A14. Supporters of public schools worried that this proposal would make senior citizens less likely to vote for a tax increase. See, e.g., JX 105 (inquiring about effect of reduction in tax credit); JX 123 at D1799, D1800 (email from senior citizen stating that he intended to vote “NO” in part due to the proposed reduction in tax credit; Floore forwarding email and commenting that “we knew we wouldn’t be able to change some people’s minds”). 95 See Tr. 675-76 (Floore describing variation in electorate). 96 The social science literature indicates that school districts employ similar strategies widely and effectively. See, e.g., Stephanie Dunne et al., Endogenizing the Median Voter: Public Choice Goes to School, 93 Public Choice 99 (1997) [hereinafter Endogenizing the Median Voter]. As discussed in the section on remedies, the parties did not address any of the extensive social science literature on pertinent issues, nor did they 30 To this end, Red Clay’s strategy focused primarily on mobilizing parents of existing Red Clay students and getting them to the polls. But the Red Clay team also identified and targeted other groups that were likely to be supportive, such as parents with children under five years of age, graduates of the Red Clay schools, and current students who were old enough to vote. Conversely, Red Clay focused on not doing things that might mobilize the opposition, particularly seniors. The Red Clay team documented their overall strategy in a presentation to the Steering Committee during that group’s first meeting, held on November 6, 2014. In advance of the meeting, Superintendent Daugherty made the official nature of the meeting clear when some principals asked whether they needed to attend. He responded, “Everyone should attend – They must realize the importance of this Referendum.”97 Kelly Penoyer, the Baltz principal, understood that the Steering Committee meeting was a mandatory event for all Red Clay principals.98 Eric Mathis, the Richardson Park principal, only attended the meeting because it was mandatory.99 As described in the presentation to the Steering Committee, the Red Clay campaign strategy had four principal planks: introduce the type of statistical analysis that the literature conducts and which would have been highly informative. See Part III.B.4, infra. Consequently, while noting the existence of this scholarship, this decision does not take it into account. 97 JX 28. 98 Tr. 114 (Penoyer). 99 Tr. 489 (Mathis). 31 • Work from past experience • Work on the Yes not the No • Inform and engage all parents • Vote Goals100 “Work from past experience” meant that in prior referendums, success resulted from mobilizing parents and other supportive groups.101 “Work on the Yes not the No” meant that converting members of supportive groups into actual votes was more important than trying to convince negative voters to change their minds.102 “Inform and engage all parents” recognized that parents of Red Clay students were the group most likely to support the tax increase. “Vote Goals” referred to specific numbers of likely “YES” voters that the Red 100 JX 25 at D2555. Cf. JX 97 (“In the past, referendums have failed because of apathy.”); JX 242 (“The 2008 referendum passed for one reason—because parents and schools worked together. The same holds true today.”); JX 169 (Superintendent Daugherty observing after Special Election that “[s]chool leaders and staff worked extremely hard to gain parent support”); JX 193 (Johnson observing before Special Election that “we will not pass this without support from our charter and choice parents”). 101 See Tr. 675 (Floore: “[I]f somebody has made up their mind, you’re not going to change their mind. That’s what I understood that to be. Work on the people [to whom] you can make your case.”); id. (explaining that at the initial Steering Committee meeting, one of the speakers gave an example of a senior citizen on a fixed income as a likely “NO” voter with whom there would be “no point” in engaging); JX 53 (Johnson stating, on behalf of the Steering Committee, “We appreciate your support in getting the vote out. We believe overall that folks support the referendum however, it is getting them out to the polls to vote that we must focus on”); JX 61 at D5290 (Brandying Springs PTO minutes stating, “Ms. Yvonne Johnson . . . shared the importance of parents getting out to vote . . . . Dr. Newton shared information about the importance of college students being eligible to request absentee ballot[s]”). 102 32 Clay team established for each school to turn out based on the school’s student population and voting patterns from past referendums.103 Implementing this strategy required a school-by-school effort. The assembled principals were told that they each “needed to have an event” on the day of the Special Election.104 They also were told that each school would have a parent leader whose responsibilities included: • Texting, Social Media and Letters to the Editor • Coordinating call sessions o Get out the vote information calls o Reminders 2 days before • Yard Signs • Working with Principals for Referendum Day Events105 The school principals were in charge of implementing the campaign plan at each school. Ammann required that each principal prepare a School Referendum Plan on a form he created. The form required that each principal provide (i) a communication plan, (ii) a list of activities leading up to the Special Election, and (iii) “school activities planned for the 103 See JX 27 at D2608 (listing voting totals from 2008 operating referendum and 2012 capital referendum, and setting target vote totals for Special Election); JX 61 at D5290 (“Each school has an identified target number of parental votes.”). 104 Tr. 114 (Penoyer testifying that her understanding from the meeting was that “every school needed to have an event on the day of the referendum”). 105 JX 25 at D2556. 33 day of the 2015 Referendum.”106 Every school principal submitted a plan, and every school held at least one and typically multiple “Referendum Day Events.”107 E. Specific Campaign Tactics Red Clay mounted a vigorous election campaign that involved a series of tactics designed to generate a favorable vote. These included (i) hosting seventy-five “Referendum Day Events” at the schools designated as polling places to bring families of Red Clay students to the polls, (ii) calling the households of Red Clay students to identify likely “YES” voters, then following up with the likely “YES” voters to get them to the polls, and (iii) sending targeted communications to Red Clay households and other groups that Red Clay believed would be likely to support the tax increase. At the same time, Red Clay avoided communications with groups that Red Clay believed would oppose the tax increase, particularly seniors, to avoid “waking up” the opposition vote.108 1. The Family-Focused Events Red Clay’s signature tactic was to hold events on the day of the Special Election, in the schools designated as polling places.109 The events were designed to appeal to families of Red Clay students, thereby providing the parents with an inducement to come to the 106 See JX 307 (compilation of all plans). 107 Tr. 625, 628 (Ammann). See JX 124 (reflecting concerns about “‘wak[ing] up’ the ‘no’ voters”); JX 134 (discussing how to avoid “wak[ing] up even more seniors”). 108 109 See JX 301 (table compiling all events). 34 polls.110 Because of their purpose and effect, the Dismissal Ruling called them the “FamilyFocused Events.” Ammann oversaw this aspect of the campaign. He required that each school principal submit a School Referendum Plan on a form he created, and the plan had to identify “school activities planned for the day of the 2015 Referendum.” 111 Ammann instructed the principals that “activities should be scheduled between the hours of 10:00 am and 7:00 pm.”112 Ammann provided examples such as “homeroom breakfast, lunch with your child, [and] after school activities.”113 Ammann reviewed and approved the plans, sometimes asked for changes, then sent them to Johnson for her review.114 Each of the twenty-three schools designated as polling places held Family-Focused Events on the day of the Special Election.115 The one school that was not designated as a See Tr. 351 (Johnson confirming that she “wanted parents in the schools on the day of the referendum because [she] wanted them to vote” and that she “didn’t want any parents to come to an event and then leave without voting”); Daugherty Dep. at 203 (agreeing that the Family-Focused Events were “get out the vote events”). 110 111 See JX 307 (stating timeframe for scheduled activities on plan forms). 112 Id. 113 Id. 114 Tr. 595-97, 621-28 (Ammann describing review process). Ammann tried at trial to suggest that he was simply an administrative conduit and was not substantively approving the plans. That testimony was contrary to the weight of the evidence. 115 PTO ¶ 27; see JX 301 (table compiling all events); Tr. 625 (Ammann testifying that on day of Special Election, all schools that were designated as polling places held events). 35 polling place did not hold any family-oriented events on the day of the Special Election.116 The principal of that school wrote in his School Referendum Plan, “Since we are not a polling site, I will support [Stanton] for any activities.”117 In total, the Red Clay schools held seventy-five Family-Focused Events on the day of the Special Election.118 During the school day, many schools scheduled luncheons for parents.119 In the evening, many schools scheduled fun activities for children. Baltz hosted a “Pajama Jammie Jam” dance party with pizza and a raffle.120 Linden Hill hosted “Blizzard Blues Beach Bingo.”121 Marbrook hosted a “Winter Carnival.”122 Shortlidge hosted “Family Line Dancing.”123 Skyline hosted a family pizza dinner and a staff versus 116 PTO ¶ 28; see Tr. 624-26 (Ammann testifying that the principal of the one school that was not designated as a polling place reallocated resources to support events being held at two of the schools that were polling places). 117 JX 307 at D3716. 118 See JX 301 (spreadsheet identifying events, target audience, likely attendance, and other details). See, e.g., JX 248 (“Honor Roll Luncheon” at Stanton); JX 250 (“Rockin’ The Referendum Day” at Warner with “Science Show” luncheon); JX 301 (“Taco Tuesday” lunch with parents and “Sixth Grade Luncheon,” among other events, at Brandywine Springs; luncheons with parents at Forest Oak; “Lunch with Your Child” at Richey; “Bring a Special Person to Lunch” at Shortlidge; “Literacy Coffeehouses” at Baltz; “Author’s Teas” at Heritage). 119 120 See JX 301; Tr. 120-25 (Penoyer describing Pajama Jammie Jam). 121 See JX 234; JX 301. 122 See JX 235; JX 301. 123 JX 301. 36 students basketball game.124 A.I. duPont and Mote held “Family Fun Nights.”125 Highlands held a “Family Fitness Night.”126 Free food was provided for attendees at twenty-four of the events.127 At two other sites, the parent-teacher organizations solicited pizza orders from parents in advance.128 Needing to pick up a pizza that you had pre-ordered gave parents good reason to attend those schools’ Family-Focused Events. The principals and staff understood that the purpose of the Family-Focused Events was to get likely “YES” voters to the polls. The principal of A.I. duPont wrote in his plan, “We will find opportunities during the studnet [sic] day to draw families in.”129 A teacher at Heritage explained to a parent that the “Author’s Teas” were “one of the activities to help get parents in to school so they will vote in the referendum.”130 124 Id. 125 Id. 126 Id. 127 Id. 128 See Tr. 618-19 (Ammann explaining that at Linden Hill and Marbrook, parents pre-ordered pizza in advance); JX 234 (pizza order form for Linden Hill); JX 235 (pizza order form for Marbrook). 129 JX 56. 130 JX 102 at D5960; see JX 301 (listing six “Author’s Teas” during the day at Heritage). 37 The principal at Richardson Park was particularly candid in his communications. He planned a luncheon event at which parents would get a free meal.131 He instructed a volunteer to remind parents “to vote first then eat.”132 For the evening, he planned a bingo night with over $1000 in prizes, two movies, and “1 free uniform pass for each voting adult.”133 He testified that because Richardson Park was a uniform school, the “no-uniform pass” was about the most valuable reward he could offer his students.134 For the evening event, he instructed a volunteer to organize a team to “catch” parents attending the “Bingo/Movie night” and “make sure they vote prior to ‘having fun.’”135 The principal believed that these events would “make it a very incentivized night for parents and kids to come out.”136 He also believed that “if parents come out most will vote yes (I hope).”137 131 See generally JX 149; JX 150. JX 149 at D11324; see also JX 150 at D11327 (“I’m thinking having 2 or 3 parents during the day . . . [and] about twice that during the evening to both help steer people to the auditorium lobby to vote and encourage them to vote yes . . . .”). 132 133 JX 108; see also JX 66 (noting that Richardson Park was hosting Bingo and movies and distributing no-uniform passes); JX 80 (soliciting donations for use as prizes). Tr. 492-94; id. at 494-95 (Q: “And if a parent didn’t vote, they didn’t get a pass. Correct?” A: “Technically.”). 134 135 JX 149 at D11324. 136 JX 66. 137 JX 108. 38 Johnson perceived the direct connection between the Family-Focused Events and getting parents out to vote. On February 14, 2015, ten days before the election, she sent out the following email to the Red Clay team, including the Red Clay principals: School Events!!! All your school events for the day of the referendum should be planned and invitations/announcements to your entire school community should have been sent home via email, phone, or paper. Principals, when your families arrive I know you will be reminding your families to go into that poll and vote! Very important, please be sure to get these folks to the polls before the event begin[s]. The polls could be closed when the event is over!138 After the election, Nash told a reporter from NBC10 News that the Family-Focused Events were “get-out-the-vote events.”139 Superintendent Daugherty agreed that the FamilyFocused Events were “get-out-the-vote events.”140 At trial, some Red Clay witnesses attempted to suggest that Red Clay held the Family-Focused Events for the general community to showcase the Red Clay schools.141 That testimony was contrary to the weight of the evidence. The Family-Focused Events were not advertised to the general community, only to the families of the schools hosting the events.142 For some events, the schools only permitted family members to attend 138 JX 313 at D20467. 139 JX 185. 140 Daugherty Dep. 203; accord JX 176 at D14676 (Daugherty referring to FamilyFocused Events as “[g]et out the vote activities”). 141 Tr. 182-84, 188 (Nash); Tr. 346 (Johnson); Tr. 621 (Ammann). 142 Tr. 183-84 (Nash). 39 because of security concerns.143 During her post-Special Election, pre-litigation interview, Nash noted that the events technically were open to voters without children, but she qualified her statement by observing, “I don’t know that they [i.e., voters without children] would . . . have [a] desire to come to Family Bingo Night.”144 She described the events more accurately as “parent events that [were] open to every parent.”145 For purposes of this litigation, Red Clay listed the seventy-five Family-Focused Events held on the day of the Special Election. Only three identified the “community” or the “public” as part of their target audience: a high school drama production, a musical showcase, and an arts concert. For every other event, the target audience was students, parents, and families.146 2. Identifying “YES” Voters Among Red Clay Families A second Red Clay campaign tactic was to identify likely “YES” voters in Red Clay student households, then follow up with targeted get-out-the-vote communications designed to get them to the polls. Red Clay did not make similar efforts to communicate with other voters. 143 Tr. 123, 145-46 (Penoyer). JX 185. Nash confirmed at trial that she “just didn’t know whether or not someone who has no connection to the school and no connection to those children would want to participate with them in an event.” Tr. 187. 144 145 JX 185. 146 JX 301. 40 Beginning in December 2014, every Red Clay school made an initial round of scripted calls to Red Clay student households to determine how the adults were likely to vote.147 Each school was assigned a specific goal for the number of “YES” voters it needed to identify; the calls were used “to find your goal ‘yes’ votes.”148 During an early debate over making direct telephone calls to households, Nash explained the purpose of the calls: The phone calls aren’t meant as a way to keep in touch through the campaign, that can occur with social media, etc. What we have done in the past was use the calls to find our goal yes votes and remind them to vote. It allows us to track where we are, and whether we are close to our goal. I think the phone campaign – even in our era of social media – remains really important.149 Throughout the campaign, internal communications stressed using the calls to identify the goal number of “YES” voters so they could be targeted later with reminders to vote.150 Daugherty Dep. 149 (“They all made parent calls . . . . [E]very school said they completed it.”). 147 148 Tr. 208 (Nash); accord Tr. 636 (Ammann). During the November 2014 parent leader meeting, which every principal was required to attend, Red Clay gave the principals target vote counts specific to each school. JX 27 at D2608. 149 JX 24; accord Tr. 208 (Nash); see JX 27 at D2608. See Tr. 189, 208 (Nash testifying about plan to identify “YES” voters and then follow up with reminders directed at those voters); id. at 209 (Nash discussing need for caller “to keep track of the number of parents that told them they would be voting yes”); JX 58 (informing team at one school that “We need to get a tally of YES votes” and describing plan for contact, identification, and follow up); JX 81 at D11597 (principal explaining that the purpose of the calls is “to inform parents of the upcoming referendum and to get an estimate of how many people we can count on to support the referendum”); id. (“In order to pass the referendum, we need to reach out to our parents to inform them of the referendum. It is also important that we tally the number of votes for the referendum.”); id. at D11598 (“Our school goal is to get 375 yes votes for the referendum! Parent responses will help us estimate where we stand.”); id. (explaining that reminder calls would be placed “to parents who respond with yes”); JX 87 (principal of Skyline reporting on results of teacher calls to 185 families that resulted in “66 yes votes (35%)[,] 150 41 The Red Clay team asked teachers to make the initial voter identification calls.151 The team adopted this strategy because “most parents will listen to what teachers have to say.”152 Consistent with the purpose of identifying a goal number of “YES” voters, teachers could send emails instead of making personal calls only if the teachers asked for and 5 no votes and 114 no contacts”); JX 96 (principal explaining that results of survey would be shared “with the Referendum committee”); JX 99 (representative from Heritage reporting on vote totals); JX 113 at D14079 (principal of Conrad explaining that “[m]any of your [sic] are making phone calls to collect as many ‘YES’ voters as we can”). See JX 66 (discussing plan to “have teachers call first and then have parents follow up as needed”); JX 79 (discussing plan for teachers “to make parent phone calls in regards to the referendum”); JX 81 at D11598 (“Homeroom teachers are responsible for calling your class list.”); JX 87 (“Yesterday Skyline teachers began making calls . . . .”); JX 91 (principal at Baltz informing staff about making calls, stating, “We are asking that each homeroom teacher be responsible for making all initial phone calls to their homeroom”); JX 96 (principal at Brandywine Springs instructing teachers to send email survey, stating, “It is time sensitive and we have to begin to determine how many families we have to make personal contact with given our survey results”); JX 108 at D11406 (principal of Richardson Park writing, “[T]he expectation we were given [by Red Clay] was that teachers would be calling”); Tr. 117-18 (Penoyer testifying that she instructed homeroom teachers to be responsible for making initial phone calls to parents using a script provided by the district and that she asked every school staff member to join a call-a-thon); Daugherty Dep. 142-43 (expressing his view that individual teachers should call their parents and could better connect with the voters). 151 152 JX 66; accord Tr. 498 (Mathis); see Daugherty Dep. at 143 (agreeing that the district felt that “individual teachers would be better able to connect with voters”). 42 received responses saying how the adults in the home planned to vote.153 Parent volunteers supplemented the teachers’ efforts using the same script.154 To facilitate the calls, Ammann had Red Clay personnel generate call lists from classroom rosters.155 The lists identified the student, the adults in the family, and their contact emails and phone numbers.156 Red Clay organized call center nights at its schools, See JX 81 at D11598 (“You can send an email . . . but you MUST ask for the parents to respond back in an email about whether we can count on them to support the referendum.”); JX 96 (distributing email script with survey for homeroom teachers to send to families); JX 98 (email distributing survey to Brandywine Springs parents “IN AN ATTEMPT TO COLLECT DATA ON OUR SCHOOL’S SUPPORT” for referendum); JX 102 (teacher from Heritage emailing parents in his class and stating, “I am contacting you via this email message rather than calling your home or cell phone but I do need a response. In other words, I will have to call if I don’t hear from you – sorry!”; stating, “Please let us know we can count on your support FOR the Referendum by hitting ‘Reply’ and writing ‘Yes’ and the number of voters we can count on from your home”); JX 218 (McKean electronic poll for Special Election asking parents for identifying information and whether parent supports the referendum); JX 233 (Heritage survey stating, “Please let us know we can count on your support FOR the Referendum by hitting ‘Reply’ and writing ‘Yes’ and the number of voters we can count on from your home”); JX 276 (email distributing survey to McKean parents “IN AN ATTEMPT TO COLLECT DATA ON OUR SCHOOL’S SUPPORT” for the referendum); compare JX 74 (email to Conrad parents about Special Election; asking recipients to take survey), with JX 75 (Ammann criticizing survey for failing to “actually ask whether or not a parent plans to support”). For an example of survey results, see JX 303 (summary of data from electronic survey distributed by McKean). 153 154 See JX 311 (Johnson distributing telephone campaign script to parent volunteers); accord Tr. 208-09 (Nash). 155 JX 112 (Ammann directing creation of lists, stating that “these calls are critical”). 156 See, e.g., JX 33 (Linden Hill); JX 34 (Skyline); JX 63 (Baltz); JX 111 (Brandywine Springs); see also JX 79 (referring to distribution of lists with “parent contacts organized by classroom”); JX 81 at D11598 (“District will provide call lists, scripts, and FAQs to assist with telephone campaign.”); JX 84 at D19668 (“All building principals received an updated call list from Ted [Ammann] this week.”); JX 87 (reporting 43 many of which provided dinner for the callers.157 Although teacher participation was nominally voluntary, Ammann charged the principals with getting the calls done. They in turn put pressure on their teachers, some of whom felt “bullied or guilted” into working on the Special Election.158 Nash drafted the script that the Red Clay callers used.159 At the outset, the caller identified herself as a parent, teacher, or other capacity “AND volunteer.” 160 The caller then asked the person reached how they intended to vote. If the person expressed support, the caller thanked them, told them their vote was important, and gave them information about when and where to vote. If the person did not seem supportive, the caller would ask that Skyline teachers were making calls “using the phone lists provided”) JX 91 (telling teachers that “[c]lass lists with phone numbers and speaking points will be available”); JX 108 (principal of Richardson Park explaining that teachers “were given lists of the students” to call); JX 112 (Ammann managing distribution of call lists and describing the calls as “critical”). 157 See, e.g., JX 81 at D11597 (Forest Oak); JX 87 (Skyline); JX 91 (Baltz); JX 107 (Linden Hill); JX 113 (Conrad). 158 JX 81 at D11596; see JX 170 at D11484 (teacher informing union negotiator that “we at AI were ‘bullied’ into making phone calls to support the referendum”); see also JX 81 at D11598 (principal asking staff to report back on how they would accomplish parent calls); JX 91 (principals asking staff to let them know if a staff member would not attend a nominally voluntary call session); JX 96 (principal asking staff to make calls and send emails, stating, “If you are uncomfortable doing this, please notify an administrator asap”); cf. JX 79 (principal informing staff that faculty meeting time would be allotted to making parent phone calls for Special Election; Superintendent Daugherty overruling after teachers’ union representative objected). 159 Tr. 208-09 (Nash); Tr. 633-35 (Ammann). 160 JX 311 (emphasis in original). 44 about their concerns and try to address one or two points. If the answer was “no” or they were still undecided, the caller would “thank them for their time and hang up” without giving information about when and where to vote.161 After each call, the caller recorded how the person planned to vote so that Red Clay could send follow-up communications to “YES” voters.162 As planned, Red Clay used the voter identification data it obtained for subsequent get-out-the-vote efforts. On the Sunday and Monday before the Special Election, Red Clay made follow-up reminder calls and sent reminder emails to the households that expressed 161 JX 91 at D15185-86; accord JX 222; JX 228; JX 274; see Tr. 209-11 (Nash describing script). See Tr. 189, 208-11 (Nash testifying about identifying “YES” voters through initial calls, followed by get-out-the-vote reminders to “YES” voters); JX 24 (explaining purpose of calls in terms of getting out the vote); JX 58 (describing plan for contact, identification, and follow up); JX 81 at D11597 (explaining that the purpose of the calls is “to inform parents of the upcoming referendum and to get an estimate of how many people we can count on to support the referendum”); JX 87 (principal of Skyline asking where to turn in call sheets); JX 113 at D14078 (principal at Conrad informing callers that “[c]all sheets are due Thursday, February 12th. Currently we have 364 confirmed YES voters with 68% of sheets collected”); id. at D14079 (principal of Conrad noting that get-out-the-vote plan contemplated “Reminder Phone Calls to ‘YES’ voters”). 162 At trial, Red Clay witnesses asserted that the purpose of the calls was merely to educate parents and provide them with factual information about the Special Election. Tr. 337-38 (Johnson); Tr. 717-18 (Floore). That testimony was contrary to the weight of the evidence. 45 support.163 On February 14, 2015, ten days before the Special Election, Johnson gave the Red Clay team the following instructions: Callers!! Everyone should have made their first round of calls. You should all be thinking about the reminder calls. The reminder calls should be done Sunday (2/22) or Monday (2/23) before the referendum. Calls [are] to remind the “yes” voters to vote. As a reminder we do not call back folks that said they are voting “no.”164 In addition to personal calls, Red Clay used the School Messenger system to send automated follow-up calls to “YES” voters reminding them to vote.165 Red Clay also sent text messages on the day of the Special Election to “YES” voters reminding them to vote.166 Only the “YES” voters received the follow-up calls and reminders.167 3. Other Targeted Communications Directed To Red Clay Parents In addition to the voter identification campaign, Red Clay used other communication channels to target the parents of Red Clay students.168 The effort began in 163 See Tr. 189 (Nash); JX 113; JX 138; JX 141; JX 147; JX 148; JX 223; JX 313; see also JX 150 at D11328 (discussing whether parents or teachers would make the followup calls). 164 JX 313 at D20467. 165 Tr. 212-14 (Nash); see, e.g., JX 223 (script of call from Dickinson School Messenger system for “[l]aunch on 2/23/2015”). See Tr. 189 (Nash); JX 113 at D14078 (“Team will text YES voters the morning of the referendum . . . .”). 166 167 Tr. 211 (Nash). 168 See generally JX 307 (compilation of School Referendum Plans including each school’s communication plan). 46 November 2014 and continued steady through January 2015, then ramped up dramatically during February 2015. The effort included the following activities: Red Clay representatives spoke about the Special Election at PTA or PTO meetings, school-organized information sessions, and school activities, such as concerts and athletic events.169 Principals sent flyers and newsletters home with students.170 Principals sent reminder emails to parents using their school-email distribution systems.171 Principals sent telephone messages to parents using their School Messenger autodialer systems.172 Red Clay staff members stuffed copies of a “Referendum Fact Sheet” in students’ report card envelopes because “report cards are one piece of mail that parents look at right away.”173 Superintendent Daugherty sent a letter in January 2015 to all Red Clay families asking them to support the tax increase.174 169 Tr. 326-27 (Johnson); see, e.g., JX 67 (Schwartz speaking at H.B. duPont Choral Concert); JX 237 (Johnson speaking at PTA meeting); JX 251 (Warner hosting “Tea and Talks” to discuss Special Election); JX 307 (passim); JX 311 at D10352 (Johnson noting, “We have been attending many of your PTA/PTO/Boosters meeting[s]”). 170 Tr. 333 (Johnson); see, e.g., JX 245; JX 246; JX 247; JX 255; JX 256; JX 307 (passim). 171 JX 141 (principal of Richardson Park instructing staff to send emails and explaining that he wanted to “flood the parents with reminders”); accord Tr. 490 (Mathis); see, e.g., JX 89; JX 100; JX 144; JX 146; JX 233; JX 307 at D3704. 172 See, e.g., JX 223; JX 238; JX 307 (passim). 173 JX 86; see Tr. 116-17 (Penoyer testifying about instructing teachers at Baltz to stuff flyers in report card envelopes). 174 JX 242. 47 Red Clay held a pep rally on the night before the Special Election.175 On the day of the Special Election, Superintendent Daugherty made an automated call to all parents using the School Messenger system.176 On the day of the Special Election, Red Clay teachers placed stickers on their elementary school students as a reminder for parents to vote.177 On the day of the Special Election, Red Clay organized teams to distribute push cards at morning drop-off lines and afternoon pick-up lines. Volunteers and district employees walked car-to-car to encourage parents to go inside and vote.178 Each of these communications focused on parents of current Red Clay students. None of them were directed to the electorate as a whole. 4. Targeted Communications Directed At Other Likely “YES” Voters Another Red Clay campaign strategy was to identify other groups that were likely to favor the tax increase. Red Clay sent targeted communications to those groups in an effort to generate favorable votes. 175 See JX 72; JX 84 at D19688; JX 139 at D20526; JX 249; JX 253. Nash believed that holding a pep rally would increase positive voter turnout. Tr. 206-07 (Nash). Approximately 4,500 people attended. Tr. 339 (Johnson). 176 JX 158; JX 159. 177 Tr. 490 (Mathis); see JX 130; JX 139 at D20526. 178 Tr. 213-15 (Nash); Tr. 340-41 (Johnson); JX 123; JX 137 at D13901; JX 139 at D20526; JX 149; JX 308. 48 The Red Clay team identified recent Red Clay graduates as one group of voters who would be likely to support the referendum.179 In December 2014, Red Clay had each high school principal send a letter to Red Clay graduates asking them “to support your former classmates and school district by voting YES to fund a much needed operating referendum.”180 The letter asked graduates living in Red Clay to come to the polls and provided information about where and when to vote; the letter asked graduates who were “away at college or serving in the military” to vote by absentee ballot.181 The letter stressed the need to “[m]ake your voice count” because “[w]e anticipate a close race and your support of Red Clay schools could make all the difference.”182 The Red Clay team also identified parents of pre-school-age children as likely to support the tax increase. On February 18, 2015, Red Clay sent a letter from Superintendent Daugherty to parents living in the district with children under the age of five.183 The letter described the Special Election as a “critical upcoming vote” and argued that “[a]s a future Red Clay parent, you have perhaps the largest stake in this referendum and ensuring our See JX 120 at D79055 (“[P]art of [the team’s] strategy was to reach out to alumni students from Red Clay . . . .”); see also JX 61 at D5290 (“Dr. Newton shared information about the importance of college students being eligible to request absentee ballot[s].”) 179 180 JX 50. 181 Id. 182 Id. 183 PTO ¶ 30; JX 254. 49 schools offer the highest quality education to our children.”184 The letter described various programs and said they were “the very things we are in danger of losing.” 185 The letter asked the parents “to come to the polls on Feb. 24 from 10 a.m. to 8 p.m. at any Red Clay school and cast a vote for your child’s future.”186 In the final days before the vote, the president of the School Board suggested “reaching out to the local colleges with [education] programs to gain support for the referendum” on the theory that “[i]f it doesn’t pass, we will not only have no job openings, but we will have experienced teachers laid off and first [in] line for the next openings.” 187 He specifically recommended this as a possible strategy “to get votes in the last days.” 188 Floore quickly worked up a “blurb” and asked the Red Clay staff who worked with Red Clay’s student teachers to email it out to their colleges and universities.189 The email included a link so that Red Clay residents could easily request absentee ballots. 190 184 JX 254. 185 Id. 186 Id. 187 JX 120 at D7906. 188 Id. at D7905. 189 Id.; Tr. 707-10 (Floore agreeing with email). 190 JX 120 at D7905. 50 The Red Clay team also recognized that current Red Clay students who were old enough to vote were likely “YES” voters.191 Ammann organized this part of the effort. In January 2015, he provided each principal of a Red Clay high school with a list of students who would be eighteen on the date of the Special Election. Ammann further facilitated student voting by telling the principals that “[i]f they [the students] [d]o not have a driver’s license or other id, your counselor can print or sign something from eschool showing address and age.”192 5. The Ghostwritten Media Campaign Yet another Red Clay campaign tactic involved ghostwriting content for parents. Nash organized a social media campaign in which she generated much of the content, but made it appear as if it came from parents. To implement the campaign, Nash created a Facebook account called “Red Clay Parents for Students” and a Twitter handle called “RedClayParents.”193 She organized parents to submit posts, designated several people at each school to share and like the posts, and established a “Response Team” of five parents to counter any negative posts.194 She 191 See, e.g., JX 56 (principal of A.I. duPont stressing in his School Referendum Plan that “we will make sure our seniors (18yrs+) understand they have the right to vote”); JX 143 (teacher at McKean explaining procedures for having 18-year-old students vote). 192 JX 64. 193 JX 25 at D2558; accord Tr. 195 (Nash testifying that a Facebook page called “Parents of Red Clay” was “created specifically for the . . . referendum”). Tr. 197-99 (Nash); JX 76 (Nash distributing “Referendum Social Media Plan”); see also JX 106; JX 272. 194 51 mapped out a schedule that contemplated having posts submitted every other day from February 2 until February 15, every day from February 15 through February 24, and every half-hour on February 24.195 Nash drafted posts for parents to submit as their own.196 She similarly prepared ghostwritten letters to the editor—the old-school version of social media—for parents to submit.197 The lack of attribution was intentional: Nash believed “it was more real and would resonate more with voters if it came from parents.”198 6. Relations With Likely “NO” Voters In contrast to Red Clay’s efforts to mobilize likely “YES” voters, the Red Clay team took steps to avoid bringing out likely “NO” voters.199 Red Clay recognized that there were identifiable groups of opposition voters and, although the Red Clay team rarely came out and said it, they associated the opposition primarily with elderly and retired residents. For example, in February 2015, Nash prepared a draft of the letter that eventually went out to parents of pre-school-aged children. The draft stated, “As a future Red Clay parent, you JX 272 (Referendum Social Media Plan detailing “who will post,” “what will be posted,” “when will it be posted,” and “where will it be posted”). 195 196 Tr. 197-98 (Nash); see, e.g., JX 32, JX 57; JX 103; JX 106; see also JX 95. See generally JX 76; JX 78; JX 123. 197 Tr. 199, 200-01 (Nash); see, e.g., JX 57; JX 103; JX 106; cf. JX 124 (discussing timing of letters to the editor). 198 Tr. 194 (Nash). 199 Tr. 352 (Johnson). 52 have a large stake in this referendum, larger perhaps than the many residents who typically come out to vote against any school funding increase.”200 Nash sent the draft to Floore and Ammann, noting that it “kind of mention[s] the senior vote, without saying seniors,” and asking if it went “too far.”201 The text did not appear in the final version of the letter.202 To avoid mobilizing the “NO” vote, the Red Clay team minimized and, when possible, delayed district-wide communications: The team delayed having parents send letters to the editor in support of the referendum because “the letters might ‘wake up’ the ‘no’ voters.”203 Even after the News Journal ran a story about the Special Election, the team considered waiting until anti-referendum letters to the editor appeared.204 Once they did, the team debated whether “the ‘pro’ letters would wake up even more seniors??”205 The team limited the number of posts about the Special Election on the district’s social media sites as part of “a conscious decision to keep it from becoming a debate.”206 For the same reason, despite making four pro-referendum videos, the Red Clay team decided not to distribute them on the district’s social media platforms.207 200 JX 273. 201 JX 104. JX 119; see JX 309 (Johnson expressing concern about a planned “bus trip to retired folks” and Floore reassuring her that the trip would target “retired red clay educators who support us” and not “random retirees”); Tr. 352-54 (Johnson discussing same). 202 203 JX 124; accord Tr. 201-02 (Nash). 204 See Tr. 201-02 (Nash); JX 103; JX 124. 205 JX 134. 206 JX 133; Tr. 215-16 (Nash). 207 JX 133; Tr. 221 (Nash). 53 When filming a pro-referendum video with Floore and Johnson, Nash told them to focus on the “big picture” rather than on the details of the campaign because publicizing the details “MAY TRIGGER AN ORGANIZED OPPOSITION . . . .”208 The team sought to avoid discussing the Special Election on a particular radio talk show because the host’s audience historically had been “the anti-vote.”209 The team delayed putting up any yard signs until the Sunday evening before the vote, because widespread signage “could bring out the no voters.”210 When the team did put out signs, they only placed them on school property near pickup and drop-off locations so they could serve “for our voters as a reminder which is mostly families.”211 The desire to avoid stirring up opposition extended to distributions of the Red Clay Record, the district’s newsletter. The Red Clay team initially planned to send the February 2015 edition only to the student households in the district.212 Later, the Steering Committee decided to send the February 2015 edition to all Red Clay residents. 213 That decision was driven by (i) an inquiry from the Department of Elections about whether Red Clay was sending a mailing to the entire district, and (ii) a desire to pre-empt the charge that Red Clay was running “a secret campaign.”214 208 JX 73. 209 JX 135; see Tr. 219-20 (Nash). 210 JX 137 at D13900; see also JX 114 (discussing timing of setting up signs); JX 139 at D20526 (same); JX 145 (same). 211 JX 137 at D13901; see also Nash Dep. 154-55. 212 JX 100. 213 See JX 101; Nash Dep. 30-31. 214 See JX 109; Nash Dep. 31, 37-39. 54 The Red Clay team’s concern about senior opposition proved prescient. After Red Clay mailed the February 2015 edition of the Red Clay Record, and after the News Journal began running stories on the Special Election, seniors began objecting.215 One described the referendum as “yet another attack on the Senior Citizens of the district, state, and the nation.”216 Another said he would vote against the referendum because “Our Governor wants to cut the school tax for senior citizens.”217 A News Journal article quoted an opponent who was a retiree. She complained that “[t]here are so many costs that people want to place on us seniors, and we just can’t afford it.”218 A News Journal article about a contemporaneous referendum in the Christina School District quoted a community leader who similarly objected that “[w]e have a lot of seniors saying ‘we can’t afford this increase right now.’”219 Johnson told the other members of the Red Clay team that objections from seniors had been “coming in since the red clay record landed.”220 215 See, e.g., JX 132 (Johnson noting arrival of opposition); JX 134 (parent informing Nash that “there were 2 [letters to the editor] in today about seniors and school tax”); JX 153 (post on Facebook page for opposition group stating, “Residents of Red Clay also stand to get punched in the wallet”); JX 163 (post on Facebook page for opposition group stating, “Get out there and VOTE NO!”). 216 JX 132 at D13973. 217 JX 123 at D1800. 218 JX 154. 219 JX 155. JX 132; see JX 134 (Nash responding to question about whether “‘pro’ letters to the editor would wake up even more seniors” by stating, “I think they [the seniors] are awake”). There is some ambiguity as to whether Red Clay sent the December 2014 edition of the Red Clay Record to all district residents or only to student households. Lippincott’s 220 55 F. Referendum Day: Red Clay’s Perspective On February 24, 2015, Red Clay held the Special Election. From Red Clay’s perspective, the day went largely as planned. The seventy-five Family-Focused Events, the targeted communications, and the get-out-the-vote efforts drew large numbers of parents to the schools that served as polling places. According to Red Clay, at least 6,383 people attended the Family-Focused Events, a figure that uses the low end of Red Clay’s estimates and does not include attendees at twenty-two events where the attendance was listed as “Unknown.”221 Many of the evening events drew hundreds of people, including approximately 300 people at Baltz, 300 people at Linden Hill, and 700 people at Brandywine Springs.222 Testimony at trial focused on the Family-Focused Events at Baltz and Richardson Park. Baltz held three Family-Focused Events: two “Literacy Coffeehouses” around lunch time and a “Pajama Jammie Jam” in the evening.223 The coffeehouses involved families later inquiry and the lack of any meaningful response from “NO” voters in December suggests that the December edition only went to families. Ammann testified that the December edition went to all Red Clay residents. Tr. 580-81 (Ammann). Accepting that it did, the December edition did not go out of its way to focus a reader’s attention on the Special Election. The front page contained a small, approximately one-inch square graphic stating “VOTE! Operating Tax Referendum. February 24, 2015. Polls Open 10 a.m. to 8 p.m. Every Red Clay School (except Central).” JX 281 at D14388. Inside, on page two, the newsletter had a half-page article on the Special Election. Id. at D14389. 221 JX 301. 222 Id. 223 JX 301. 56 visiting the school so that students could present what they were learning. They were solely for the parents, not for the public. Red Clay estimated that approximately fifty people attended.224 Food was provided free of charge.225 The “Pajama Jammie Jam” was held from 6:00 to 8:00 p.m. It was a fun event so that students could wear their pajamas and dance.226 Food again was provided free of charge.227 For security reasons, the Pajama Jammie Jam was only open to Baltz students and their families; it was not open to the community.228 Baltz sent home approximately three flyers to its families to advertise the event. Baltz also used the Alert-Now system to send a recorded telephone message to its families to remind them about the event.229 Red Clay estimated that approximately 300 people attended the Pajama Jammie Jam.230 Baltz stationed staff members at a table at the main entrance to the school to greet families as they entered.231 A sign behind the table stated, “If you care for the Baltz Bear 224 Tr. 118-19 (Penoyer); JX 301. 225 JX 301. 226 Tr. 120 (Penoyer). 227 JX 301. 228 Tr. 123, 145-46 (Penoyer). 229 Id. at 121-22. 230 Id. at 122; JX 301. 231 Tr. 132 (Penoyer). 57 Vote Yes!”232 Although there was a separate entrance for voting, voters often used the main entrance.233 Upon entering the building, each family received a check-off card.234 The card had three boxes labeled, respectively, “I ate,” “I voted,” and “I danced.” Below the boxes were lines for the “Student Name,” the “Parent Name,” and a phone number.235 One of the purposes of the card was to give each family a checklist of the things they were able to do.236 Teachers and staff had stamps to check off items on the card. They asked to look at the cards and encouraged families to check off activities that they had not yet completed, such as voting.237 At the end of the night, the checklists served as raffle tickets, and the person whose card was selected won a prize.238 Richardson Park held two Family-Focused Events: a luncheon for parents during the day and a bingo and movie night in the evening.239 Approximately seventy-five parents 232 JX 162; see Tr. 136-37 (Penoyer). 233 See Tr. 132-35 (Penoyer); Tr. 300 (McHugh). 234 Tr. 124 (Penoyer). 235 JX 275. 236 Tr. 124-26 (Penoyer). 237 Id. at 128, 130; see Tr. 302-03 (McHugh). 238 Tr. 131-32 (Penoyer). 239 JX 301; Tr. 489 (Mathis). 58 attended the luncheon, and approximately 300 people attended the bingo and movie night.240 The bingo games offered prizes valued at over $1,000. More importantly, each child who attended the event received a “no-uniform pass” for “each adult who voted.”241 If a parent did not vote, then the child was not supposed to get a no-uniform pass, but the school did not monitor that closely. The passes served their purpose by incentivizing people to get to the polling place. Because the passes could be used on picture day, they functioned as a reward for the whole family.242 During both events, Richardson Park had parents circulating to “tell people to vote yes.”243 Richardson Park used parents for this purpose because the principal did not believe that he or his staff could ask people to vote “yes.”244 Teachers also circulated during the events. They told parents how important the vote was and “how little time it will take to make a difference for students.”245 240 JX 301; Tr. 499 (Mathis). 241 Tr. 491 (Mathis); see JX 141. 242 Tr. 495, 499 (Mathis). 243 Tr. 500 (Mathis). 244 Id. at 503-04. 245 Id. at 507. 59 G. Referendum Day: The Community Witnesses’ Perspective For a series of witnesses from the community, the Special Election did not go smoothly. Plaintiff Rebecca Young described her experience attempting to vote at North Star. She brought her parents with her so that they could vote as well. At the time of the Special Election, Rebecca Young was sixty-seven years old, her father James Young was ninety years old, and her mother Elizabeth Young was eighty-eight years old. Both of Rebecca’s parents have mobility issues. They initially tried to vote in the morning at approximately 10:00 a.m. As Rebecca explained at trial, “The parking lot was completely packed, jammed, congested. There was no place to park where I could reasonably expect my parents to walk into the -- the polls.”246 They tried again around 3:00 p.m. and found the same situation.247 Because of the lack of parking, Rebecca and her parents were unable to vote. They had never encountered that problem before.248 Mary O’Neill, a non-party witness, testified about her experience attempting to vote at Marbrook. She attempted to vote twice, first at approximately 10:30 or 11:00 a.m., and 246 Tr. 8 (Young). 247 Id. at 8, 10. 248 Id. at 11. Red Clay called a community witness to rebut the testimony about parking problems at North Star. Tr. 230-34 (Landseder). The documentary record reflected that the witness had not voted at North Star on the day of the Special Election. JX 161. It seems likely that the witness confused the Special Election with another occasion on which he voted at North Star. His testimony also concerned the parking situation around 1:30 p.m., which was not when the Youngs tried to vote. 60 a second time around 3:30 or 3:45 p.m.249 Each time, O’Neill found a packed parking lot and no spaces.250 O’Neill suffers from fibromyalgia and lower back pain, so when she could not find a convenient space, she “gave up” and “[w]ent home.”251 She called the Department of Elections the next day to complain, and she also contacted her state representative. She felt she had “every right to vote” but was prevented because Red Clay had “‘events’ going on at the school.”252 State Representative Deborah Hudson testified as a fact witness in addition to serving as one of the plaintiffs’ experts. She also testified about voting at Marbrook. She arrived at Marbrook between 5:30 and 6:30 p.m. and found that the parking lot was full. She circled two or three times, then waited for someone to leave. She also saw a bus that was parked in the circle in front of the school.253 When she entered the school, she saw “[b]ig crowds” and “a lot of people” that included “[m]ore adults than [she was] used to seeing” for an election.254 But despite the crowds, there was not a long line to vote.255 Someone asked her if she was there to vote or to attend the Winter Carnival, and she 249 Tr. 89-90 (O’Neill). 250 Id. 251 Id. at 88, 90. 252 Id. at 90-91. 253 Tr. 27-28, 34-35, 53-54 (Hudson). 254 Id. at 28, 34. 255 Id. at 28. 61 decided to check out the carnival. In the “cafetorium,” she saw a well-run event with games for children and people in the center of the room eating. Her immediate reaction was that the event was a planned activity that Red Clay arranged in an attempt to get “persuadable” voters to vote in favor of the tax increase.256 David Pickering, a non-party witness, testified about his efforts to vote. He left work around 5:00 p.m. and was planning to pick up his parents and take them to vote at H.B. duPont. When he drove by the school on his way to get them, he saw that the parking lot was full of cars. He told his parents that “the polls were very busy,” and they decided at that point not to try to vote.257 After that, Pickering drove to Skyline. When he got there, he saw another full parking lot. He called a friend who was voting inside at the time. His friend told him that “the voting line wasn’t bad,” but that “[t]here w[ere] a lot of people in there for some kind of event.”258 Pickering then drove to Marbrook where the “[p]arking lot was packed.”259 He also saw buses in the circle driveway. He eventually parked in the fire lane. When he got inside, there was virtually nobody voting; everyone was there for an event. After voting, Pickering called the Department of Elections to complain and emailed 256 Id. at 29-30, 32, 38-39, 41. 257 Tr. 73 (Pickering). 258 Id. at 74. 259 Id. 62 State Senator Karen Peterson. He “felt it wasn’t right to have the parking lot so full of cars when [he] was trying to get in there to vote.”260 Mary Ellen Fitzpatrick, a non-party witness, testified about her experience voting at Linden Hill. She drove to the school around 1:30 p.m. She found a “packed” parking lot, cars parked illegally in the circle in front of the school, and cars lining both sides of the street leading to the school. On a second trip through the lot, she decided to park at the end of a line of spaces in the circle, even though it was not a legal space.261 No other spaces were available. When she got inside, the multipurpose room was full of people, and she initially thought it would take a long time to vote. Then she saw a separate sign saying “vote here,” where there were approximately four people in line. She voted and left. She had never encountered similar problems with parking when trying to vote.262 Russell Schnell, a non-party witness, testified about his experience voting at McKean. He arrived at approximately 5:30 p.m. He discovered that the parking lot was “completely full” and that one of the sidewalks was under construction. After spending approximately five minutes circling the lot several times, he parked beside a construction barricade in an area that was not a parking spot.263 He expected there would be a long line to vote, but there were only three or four people voting. After he got home, he sent a note 260 Id. at 78. 261 See Tr. 96-97 (Fitzpatrick); see also JX 284 at D425. 262 Tr. at 100-01 (Fitzpatrick). 263 Tr. 148-49, 156 (Schnell). 63 to the American Civil Liberties Union because he was concerned about how the election was being conducted.264 Sean Boyle, a non-party witness, testified about his experience voting at Brandywine Springs. He arrived around 7:00 p.m. The parking lot was full, and he drove around for five to ten minutes before finding a spot. Other cars were also circling the lot. When Boyle entered the school, he saw lines of people waiting to enter a choral production. To the right, he saw the voting area, where there were perhaps three people waiting to vote. A few days later, he saw an article in the News Journal about problems with voting, so he reached out to Senator Peterson and described his experience.265 Several Red Clay witnesses who were involved with the Special Election sought to rebut the testimony from the community witnesses. Johnson testified that she visited a series of schools on the day of the Special Election and did not see any parking problems, including at Cab Calloway around noon, Marbrook around 1:30 p.m., H.B. duPont around 2:00 p.m., Brandywine Springs from 3:00 to 4:30 p.m., Linden Hill from 5:30 to 6:30 p.m., and North Star around 6:30 p.m.266 Only two schools held Family-Focused Events during those times, and those events only drew twenty-three and twenty-four attendees.267 264 Id. at 154-55. 265 Tr. 162-65, 167-69 (Boyle). 266 Tr. 340-45 (Johnson). See JX 301 (twenty-three attendees at Marbrook’s Fifth Grade Author’s Day; twenty-four attendees at North Star’s Read-In Night). 267 64 Johnson’s testimony did not rebut the community witnesses’ accounts. Floore testified that she voted at North Star around 10:00 a.m. and did not see any parking problems. 268 The plaintiffs testified that the parking lot was full at the time. The accounts can be reconciled if the witnesses were slightly off about the time. The Youngs may have been at North Star somewhat later in the morning. North Star held six Family-Focused Events during the day, but Red Clay could not say when any of them took place.269 If some took place after Floore left when the Youngs tried to vote, then both witness accounts can be credited. Ammann was in charge of ensuring that there was adequate parking at the polling places, and he understood that Red Clay needed to ensure access for voters. 270 He recognized that with the large turnout Red Clay hoped to generate, there could be parking problems at some locations. To avoid congestion at Forest Oak and Linden Hill, he had teachers park in nearby commercial lots.271 For similar reasons, at Richardson Park, teachers and staff were not permitted to park in the front and side lots.272 Ammann also coordinated with the Department of Elections to meet their parking requirements. Their regulations contemplated using five “Voter Only” parking signs to 268 Tr. 677 (Floore). 269 See JX 301. 270 Tr. 631-32 (Ammann). See JX 88; JX 116; Tr. 616 (Ammann testifying that Red Clay “transferr[ed] some teachers from the remote parking to Linden Hill”). 271 272 JX 140; JX 142; JX 149; JX 151. 65 secure ten to twenty spaces.273 The Department asked Ammann to have the spaces monitored so that they remained available for voters. Ammann prepared a parking plan for each school, and the Department of Elections approved his plans.274 The process fell short in the implementation. Ammann admitted that he did not instruct the principals to monitor the parking situation, which led to the principal at Marbrook not taking any action to address the buses that were parked in her school’s circle driveway for a period of time on election day.275 Ammann testified that he instructed the custodial staff at each school to monitor the parking lots to ensure that the spots designated for voters remained available, but there is no contemporaneous evidence that he gave that instruction, even though Ammann used email as a primary means of communication.276 Ammann did not follow up on this instruction during the Special Election, and no one explained how custodians would have distinguished a voter’s car from an event attendee’s, 273 JX 121. 274 See, e.g., JX 266 (plan for North Star); JX 182 (plan approval). 275 Tr. 632 (Ammann testifying that he did not speak with the principals about monitoring parking access). The buses were from Charter School of Wilmington and parked at Marbrook for part of the afternoon as part of an activity. Id. at 631. Having heard the Red Clay witnesses testify, I find that there was no intent to use buses to interfere with the ability of elderly or disabled residents to access the polls. 276 Id. at 632-33 (Ammann asserting that he told the custodians to monitor the parking). 66 particularly when those categories overlapped. Superintendent Daugherty testified that Red Clay did not monitor who was using the parking spaces.277 H. Complaints About The Election Very few issues were reported to the Department of Elections on the day of the Special Election itself.278 On the day after the Special Election, the Department of Elections was “slammed” with complaints.279 Voters also contacted their state legislators, including Senator Peterson and Representative Hudson.280 On March 3, 2015, the Board of Elections for New Castle County met to consider whether to certify the results of the Special Election.281 The Department of Elections reported on the complaints it had received. The Board of Elections determined that it did not have authority to investigate the complaints or take them into account when certifying the election.282 It referred isolated issues to the Attorney General for investigation. The 277 Daugherty Dep. 205. 278 JX 177 at D303. 279 JX 166 (“We are really getting slammed about events yesterday . . . .”). 280 Tr. 48 (Hudson); Peterson Dep. at 11-16. 281 PTO ¶ 7. 282 This court agreed. A companion decision to the Dismissal Ruling dismissed the plaintiffs’ claims against the Board of Elections, holding that “under the statutory scheme that existed at the time of the Special Election, the Board of Elections had no authority to consider legal violations when certifying the results beyond fraud that was apparent on the face of a ballot or certificate.” Young v. Red Clay Consol. Sch. Dist., 2015 WL 5853762, at *4 (Del. Ch. Oct. 2, 2015). 67 Attorney General only reviewed the events for criminal violations and determined not to pursue any criminal charges. I. The Certified Results On March 10, 2015, the Board of Elections certified the results of the Special Election. There were 6,395 votes in favor of the tax increase and 5,515 votes against.283 The winning margin of 880 votes amounted to 7% of those who voted. The 11,910 votes cast represented about 7% of the approximately 165,000 residents in the district and about 12% of its 93,905 registered voters.284 The winning margin thus amounted to less than 0.5% of the district’s residents and less than 1% of its registered voters. As previously noted, Red Clay estimated that the Family-Focused Events drew at least 6,383 people to the polls. That figure uses the low end of Red Clay’s estimates for three events and does not include any attendees for twenty-two events where Red Clay listed the attendance as “Unknown.”285 Using the high end of Red Clay’s estimates, the attendance figure rises to 6,593 people. The average attendance figure for the fifty-three events for which Red Clay provided estimates was 120 people. Using that number for the twenty-two events where Red Clay listed the attendance as “Unknown” adds another 2,640 attendees. A more realistic estimate for the Family-Focused Events is that they brought 283 PTO ¶ 6; JX 178; JX 181. 284 Dkt. 144; see Tr. 252-53 (Ratledge). 285 JX 301. 68 between 9,023 and 9,233 people to the polling places. The mid-point of that range is 9,128 attendees. Red Clay’s low-end attendance figure of 6,383 means that the number of people who attended the Family-Focused Events was approximately equal to the 6,395 votes that Red Clay received in favor of the tax increase. Using the low-end attendance figure, the winning margin of 880 votes represented approximately 14% of the people who attended the Family-Focused Events. The more realistic attendance figure of 9,128 means that the number of people who attended the Family-Focused Events exceeded the 6,395 votes that Red Clay received in favor of the tax increase. Using the high-end attendance figure, the winning margin of 880 votes represented approximately 10% of the people who attended the Family-Focused Events. J. Procedural History On March 27, 2015, the plaintiffs filed their initial complaint and sought an expedited hearing on an application for a preliminary injunction that would block the tax increase from taking effect. I denied the motion to expedite, holding that even if the tax increase went into effect, a post-trial remedy could be crafted if the plaintiffs prevailed. On April 13, 2015, the plaintiffs filed an amended complaint. Red Clay moved to dismiss the amended complaint for failing to state a claim on which relief could be granted. After briefing and oral argument, the matter was submitted for decision on July 10. 69 On October, 7, 2015, I issued the Dismissal Ruling, which held that the plaintiffs had stated a claim on which relief could be granted. The matter proceeded to trial. After post-trial briefing, post-trial argument was held on February 23, 2017. K. The Present Status Of The Tax Increase On July 1, 2015, Red Clay began receiving incremental revenue from the tax increase. During the 2015-16 tax year, Red Clay received approximately $10.5 million in additional tax revenue, reflecting the initial $0.20 increase. During the 2015-16 tax year, Red Clay received approximately $15.8 million in additional tax revenue, reflecting both the initial $0.20 increase and the next $0.10 increase. The full $0.35 tax increase will go into effect on July 1, 2017.286 II. LEGAL ANALYSIS The Dismissal Ruling evaluated the plaintiffs’ complaint. In doing so, it set out a legal framework for analyzing the legal issues presented by the case. As often happens, the evidence presented at trial differed in some respects from the less-informed allegations of the complaint. At the pleading stage, the plaintiffs primarily challenged the FamilyFocused Events as a means of enhancing turnout by parents of Red Clay students while suppressing turnout by elderly and disabled residents. They alleged that Red Clay engaged in excessive advocacy and targeted campaign speech, thereby violating the principles 286 PTO ¶¶ 13-14. 70 established in Brennan v. Black.287 They were not aware of the full scope of Red Clay’s targeted campaign speech. In light of the Dismissal Ruling and the evidence presented at trial, the plaintiffs no longer meaningfully challenge Red Clay’s broadly directed campaign speech. They challenge the Family-Focused Events and Red Clay’s targeted campaign speech. A. The Primacy Of The Plaintiffs’ State Law Claim The plaintiffs assert that Red Clay’s electoral interventions violated both federal and state law. As their federal theory, they contend that Red Clay violated the Due Process and Equal Protection Clauses of the Fourteenth Amendment of the United States Constitution.288 As their state law theory, the plaintiffs contend that Red Clay violated the Elections Clause. Red Clay has asked the court to consider the state constitutional claim first and only reach the federal constitutional claims if necessary. This approach comports with Delaware Supreme Court jurisprudence, which generally gives primacy to specific guarantees in the Delaware Constitution.289 It also makes logical sense. The Dismissal Ruling concluded that 287 104 A.2d 777 (Del. 1954). 288 In the Dismissal Ruling, this court held that the Equal Protection Clause provided a more logical framework for analyzing the plaintiffs’ claims. See 122 A.3d at 830-31. The plaintiffs nevertheless continued to advance a claim under the Due Process Clause, which they supported in their pre- and post-trial briefing with more thorough legal analysis than at the pleadings stage. 289 See id. at 811-13 (discussing Delaware Supreme Court precedent that gives primacy to provisions of the Delaware Constitution and declines to interpret them in lockstep with federal analogs). See generally Randy J. Holland, The Delaware State 71 “[t]he Elections Clause has independent content that is more protective of electoral rights than the federal regime.”290 It is therefore possible that an electoral intervention could pass muster under the federal constitution and yet violate the more specific guarantee provided by the Delaware Constitution.291 Constitution 32-34 (2011); Randy J. Holland, State Constitutions: Purpose and Function [hereinafter Purpose & Function], in The Delaware Constitution of 1897: The First One Hundred Years 3, 17 (Harvey Bernard Rubenstein et al. eds., 1997) [hereinafter First 100 Years]. 290 122 A.3d at 813. 291 A different outcome seems unlikely in this case. Under United States Supreme Court precedent, a selective incentive for voting is subject to the same constitutional analysis a selective burden. Hopper v. Bernalillo Cty. Assessor, 472 U.S. 612, 619 n.8 (1985); see, e.g., Harlan v. Scholz, 210 F. Supp. 3d 972, 977-79 (N.D. Ill. 2016) (enjoining law that would require high-population urban counties to provide same-day voter registration while not requiring low-population rural counties to do the same); Garza v. Smith, 320 F. Supp. 131, 137 (W.D. Tex. 1970) (invalidating statute that allowed blind persons to have an assistant in the voting booth but did not extend the same advantage to illiterate voters), vacated on other grounds, 401 U.S. 1006 (1971). Imposing a burden on an identifiable group of voters violates the Equal Protection Clause unless the government can justify the selective burden with a sufficiently compelling state interest. See Anderson v. Celebrezze, 460 U.S. 780, 782 (1983) (invalidating Ohio’s filing deadline for third party candidates, finding that it “place[d] a particular burden on an identifiable segment of Ohio’s independent-minded voters”); Bullock v. Carter, 405 U.S. 134, 144 (1972) (holding that high filing fees for candidates imposed by Texas law were unconstitutional because of the “obvious likelihood that this limitation would fall more heavily on the less affluent segment of the community, whose favorites may be unable to pay the large costs required by the Texas system”); Obama for Am. v. Husted, 697 F.3d 423, 436-37 (6th Cir. 2012) (affirming injunction blocking statute that established shorter early voting period for nonmilitary voters). “[I]t is especially difficult for the State to justify a restriction that limits political participation by an identifiable political group whose members share a particular viewpoint, associational preference, or economic status.” Anderson, 460 U.S. at 793. “‘Fencing out’ from the franchise a sector of the population because of the way they may vote is constitutionally impermissible.” Carrington v. Rash, 380 U.S. 89, 94 (1965); see id. (invalidating Texas law that excluded members of the Armed Forces who moved their residency to Texas from voting in Texas elections as long as they remained in the Armed 72 This decision consequently starts with the Elections Clause. Because the plaintiffs have proven that Red Clay violated the Elections Clause, this decision does not reach the federal claims. B. Distinguishing Among The Types Of Electoral Interventions The plaintiffs challenge Red Clay’s electoral interventions as a whole, contending that they violated the Elections Clause when viewed in the aggregate. For analytical clarity, however, it is helpful to consider the Family-Focused Events separately from Red Clay’s other interventions, all of which involved varying degrees of government campaign speech. Government involvement in elections creates difficult line-drawing problems. A government that intervenes in elections to protect electors and prevent private parties from using force, coercion, bribery, or intimidation does not undermine the legitimacy of the election. That type of government activity enhances the electoral process by making it possible for more electors to participate freely and express their views. Different and more difficult issues arise if a government uses its powers to encourage or facilitate voting by electors who might be thought to favor the government’s positions, or to discourage or interfere with voting by electors who might be thought to oppose the government’s views. Along similar lines, a government that Forces); Patriot Party of Allegheny Cty. v. Allegheny Cty. Dep’t of Elections, 95 F.3d 253, 262, 264, 269 (3d Cir. 1996) (invalidating Pennsylvania law that prevented crossnomination and fusion only for minor political parties because it severely burdened two identifiable groups—the minority party and those who wished to vote for its candidates— and “constitute[d] the type of ‘invidious discrimination’ prohibited by the Fourteenth Amendment”); Stoddard v. Quinn, 593 F. Supp. 300, 305, 309 (D. Me. 1984) (invalidating statutory filing deadline because it placed a burden on an “identifiable segment of Maine voters—those who may be dissatisfied with the major parties’ choices”). The Dismissal Ruling held at the pleading stage that Red Clay had not justified its electoral interventions with sufficiently compelling state interests. See 122 A.3d at 833-37. Since then, Red Clay has not advanced any additional interests, and the factual record at trial showed that Red Clay’s interventions were more extensive that what the plaintiffs had been able to allege in their complaint. 73 provides factual information to voters acts in an election-enhancing capacity. The potential for concern grows as the tenor, volume, and extent of the government’s advocacy increases. Whether listeners object may well depend on whether they agree with the substance of the government’s views. What one voter might call helpful information, another might label propaganda.292 Recent allegations that a foreign state influenced a national election by altering the information available to voters has brought home the extent to which subtle interventions can affect an electoral outcome. This decision distinguishes targeted incentives and disincentives for voting from government campaign speech. The latter term refers to “speech to the public (rather than to other government entities) that expresses the official view of a governmental branch or body, such as speech issued collectively in the form of a resolution or proclamation, or speech by an official empowered to speak for that governmental entity.”293 Government campaign speech can take a variety of forms. The least problematic type of government campaign speech is directed broadly to the electorate as a whole and openly identified as coming from a government source. Red Clay engaged in this type of campaign speech. Examples included: The two mailings of the Red Clay Record. 292 Dismissal Ruling, 122 A.3d at 799. 293 Helen Norton, Campaign Speech Law with a Twist: When the Government is the Speaker, Not the Regulator, 61 Emory L.J. 209, 213-14 (2011). Norton distinguished government campaign speech during referendums from government campaign speech in favor of particular candidates and excluded the latter from her definition. Id. at 214. In the Dismissal Ruling, I observed that the term logically extends to both contexts, but I agreed that the two scenarios present different policy issues such that government speech during an election involving candidates should receive greater scrutiny. See 122 A.3d at 800 n.7. 74 Postings on the district’s official website and official social media channels. Videos distributed through the district-sponsored cable television channel. Presentations by self-identified Red Clay representatives at public meetings and workshops that were advertised and open to the public. Signs in Red Clay schools or in the community asking for support. Statements by Red Clay representatives to the press. Red Clay also engaged in broadly directed government campaign speech that did not openly identify its source. Instead, Red Clay took steps to mask this speech to look like it originated from parents. Examples included (i) the social media campaign that Nash ran using the Facebook account called “Red Clay Parents for Students” and the Twitter handle called “RedClayParents” and (ii) ghostwritten letters to the editor. These forms of campaign speech became problematic because Red Clay masked their source. A more serious type of electoral intervention involves “targeted government campaign speech, which is directed to identifiable groups within the electorate.” 294 By communicating more vigorously with identifiable subsets of the electorate, the government can encourage voting by groups that are likely to favor its position, thereby increasing turnout among those groups.295 This type of intervention involves the government making 294 Dismissal Ruling, 122 A.3d at 800. 295 Id. at 802. 75 distinctions among groups and attempting to affect the outcome of the election by shaping the demographic characteristics of those who vote.296 Red Clay engaged in extensive campaign speech that was targeted at identifiable groups. Most of it targeted parents of Red Clay students, whom Red Clay believed would support the tax increase. These efforts included: Superintendent Daugherty’s letter to Red Clay student households in January 2015. Presentations by Red Clay representatives at events that parents primarily attended, such as school concerts and athletic events. The flyers and newsletters that Red Clay personnel sent home with students. The “Referendum Fact Sheet” that staff members stuffed into student report cards. The reminder emails that principals sent to parents using their school-email distribution systems. The reminder messages that principals delivered to parents using the School Messenger auto-dialer system. Superintendent Daugherty’s referendum-day reminder message delivered to parents using the School Messenger auto-dialer system. The “VOTE” stickers that Red Clay school teachers put on elementary students. The teams that Red Clay organized to distribute election push cards and encourage parents to vote during morning drop-off and afternoon pick-up. Red Clay also targeted other groups that Red Clay believed would support the tax increase. These efforts included: The December 2014 letter to Red Clay graduates. 296 Id. at 801. 76 The February 2015 letter to parents of children under five years of age. The February 2015 email to local colleges with teacher education programs. The facilitation of voting by high school students who were old enough. An even more serious type of targeted government campaign speech moves from identifiable groups to individual voters. This type of speech involves the government seeking to identify how particular individuals will vote and then taking steps to convert supportive individuals into actual votes. It therefore involves the government discriminating among individuals and attempting to affect the outcome by influencing their behavior. The Dismissal Ruling did not discuss individualized government campaign speech because the plaintiffs did not know it had occurred. During discovery, the plaintiffs learned that Red Clay gave each school a specific number of “YES” voters to identify among its parents. When Red Clay representatives reached “YES” voters, they provided information about how to vote and added the voters to a list for follow-up. When Red Clay representatives reached “NO” voters, they did not provide information about how to vote. Just before the election, Red Clay contacted the “YES” voters to get them out to vote. So far, each level of government intervention has only involved campaign speech. The focus of the speech has narrowed, starting with the electorate as a whole, then moving to identifiable groups, and then reaching individual voters, but the exclusive medium has been speech. A qualitatively different intervention targets an identified group with a concrete incentive or disincentive to vote. Like targeted government campaign speech, this type of intervention involves the government discriminating among identifiable groups by 77 choosing to favor some and disfavor others. It also involves the government trying to influence the outcome of the election by shaping the demographic characteristics of those who vote. But the intervention goes beyond speech by putting a thumb on the scales of the decision to vote. Much of the legal scholarship and case law has involved disincentives for voting. Historical examples involved poll taxes, literacy tests, and other hurdles that were superficially neutral but operated as impediments to identifiable groups.297 More recent efforts involve voter identification requirements, which appear neutral but may operate in practice as a disincentive for identifiable groups.298 297 See generally Christopher S. Elmendorf, Structuring Judicial Review of Electoral Mechanics: Explanations and Opportunities, 156 U. Pa. L. Rev. 313, 365-76 (2007). The Twenty-Fourth Amendment to the United States Constitution abolished the poll tax in federal elections. U.S. Const. amend. XXIV, § 1. The United States Supreme Court invalidated the poll tax in state elections because it discriminated among voters on the basis of wealth. Harper v. Va. State Bd. of Elections, 383 U.S. 663, 666-67 (1966). The Voting Rights Act Amendments of 1970 largely prohibited the use of literacy tests, and the United States Supreme Court held that the ban fell within Congress’ enforcement powers under the Fourteenth and Fifteenth Amendments. Oregon v. Mitchell, 400 U.S. 112, 118 (1970). 298 See, e.g., Crawford v. Marion Cty. Election Bd., 553 U.S. 181 (2008); N.C. State Conference of NAACP v. McCrory, 831 F.3d 204 (4th Cir. 2016); Veasy v. Abbott, 796 F.3d 487 (5th Cir. 2015) (rehearing en banc pending); Frank v. Walker, 768 F.3d 744 (7th Cir. 2014); Common Cause/Ga. v. Billups, 554 F.3d 1340 (11th Cir. 2009); Michael D. Gilbert, The Problem of Voter Fraud, 115 Colum. L. Rev. 739, 743-50 (2015); Charles Stewart III, Voter ID: Who Has Them? Who Shows Them?, 66 Okla. L. Rev. 21, 23-30 (2013); Spencer Overton, Voter Identification, 105 Mich. L. Rev. 631, 633 (2007). Empirical data on the effect of voter identification laws on turnout is mixed. See Gilbert, supra, at 749-50 (compiling literature); Robert S. Erikson & Lorraine C. Minnite, Modeling Problems in the Voter Identification—Voter Turnout Debate, 8 Election L. J. 85, 98 (2009) 78 But a government also can provide an incentive to an identifiable group. This type of intervention transfers something of value to a subset of the electorate in return for the act of voting. Conceptually, it resembles a bribe. As the Dismissal Ruling recognized, some have argued that to increase turnout, governments should be able to provide a reward for voting to all voters.299 Several states allow payments for voting, so long as the payment is not designed to induce a voter to favor a particular candidate or result.300 No one appears to favor a government having the ability to provide rewards to a particular subset of the electorate that the government believes will favor its preferred outcome. (describing then-existing science regarding vote suppression as “incomplete and inconclusive”). See Natalie J. Lockwood, International Vote Buying, 54 Harv. Int’l L.J. 97, 116 (2013) (citing Simeon Nichter, Vote Buying or Turnout Buying: Machine Politics and the Secret Ballot, 102 Am. Pol. Sci. Rev. 19 (2008)) (arguing that vote buying may be effective in encouraging voter turnout); Anthony B. Sanders, In Defense of Vote Buying: How “Nader Traders” Can Defeat Rent Seeking, 26 Hamline J. Pub. L. & Pol’y 43 (2004) (arguing that vote trading, as opposed to vote buying, could increase voter turnout without increasing rent seeking or interest group capture); Richard L. Hasen, Vote Buying, 88 Cal. L. Rev. 1323, 1326, 1332-34 (2000) [hereinafter Vote Buying] (arguing that paying rewards to boost turnout would be beneficial when evaluated using norms of equality and efficiency but could pose concerns for community self-governance). 299 300 See Cal. Elec. Code § 18522 (prohibiting any person from providing consideration to a voter to induce or reward the voter to vote for or against a particular candidate, or to refrain from voting generally); Dansereau v. Ulmer, 903 P.2d 555, 560 (Alaska 1995) (“Although [Alaska law] prohibits a person from paying another person to vote for a particular candidate, proposition, or question, no Alaska Statute prohibits a person from compensating another person for voting per se.”). See generally Vote Buying at 1355. 79 As described in the Statement of Facts, the seventy-five Family-Focused Events were intended to and operated as an inducement for the families of Red Clay students to vote. Red Clay held Family-Focused Events at every school that was designated as a polling location. Their purpose was to draw parents of Red Clay students to the polls. The incentive was a tangible reward in the form of child-focused activities. Although the events were nominally open to everyone, Red Clay recognized that adults without children would not have a desire to attend. Red Clay’s decision to convey the rewards in the form of events did not change their status as a reward for voting. Consider the free food that was made available at twenty-four of the events, or the carnivals and family fun nights that various schools held. Instead of hosting the dinner and carnivals themselves, the schools could have given families coupons to a local restaurant, arcade, or family fun park. Or in lieu of providing coupons, the schools could have handed out money and encouraged families to spend it at a local restaurant, arcade, or family fun park. If the schools had handed out coupons or money, then the exchange of something of value for the act of voting would have been readily apparent. By hosting the dinner and carnivals themselves, the schools simplified the exchange by providing the benefits directly.301 I recognize that some may not immediately perceive the Family-Focused Events as a reward for voting. Or some may feel that because public schools need money, and because public schools are indisputably a public good, the end justifies the means. A thought experiment involving a different issue may offer perspective: 301 Dismissal Ruling, 122 A.3d at 803. 80 Imagine a state-wide referendum on an issue relating to gun rights, such as a proposal to allow universal concealed carry. Posit that the state Department of Elections designates gun ranges as polling places, which is permitted by law. Envision that the separate state agency that focuses on gun regulation engages in a state-wide, pro-referendum campaign that culminates in each gun range offering every adult of voting age free range time and a box of free ammunition on the day of the referendum. As with the Family-Focused Events, the gun range events are facially neutral and open to everyone. Nevertheless, I would expect many would regard this type of intervention as a means of increasing turnout by voters who would favor the referendum. The same is true for the Family-Focused Events. What matters for legal analysis is not whose political ox is gored, but the effect on the electoral process. In practice, the Family-Focused Events did more than just operate as an incentive to vote for parents of Red Clay students. They also functioned as a disincentive for voting by the elderly and disabled. The Family-Focused Events caused the school parking lots to be packed with cars. The full lots made it difficult for elderly and disabled residents to find accessible parking spots. Although Red Clay did not purposefully create this obstacle, the Family-Focused Events had this effect. Because the Family-Focused Events provided an incentive for voting by parents of Red Clay students while creating a disincentive for voting by the elderly and disabled, this decision analyzes them separately from Red Clay’s government campaign speech. In ultimately deciding whether Red Clay’s conduct affected the election, this decision considers Red Clay’s interventions as a whole. 81 C. The Family-Focused Events The Elections Clause provides that “[a]ll elections shall be free and equal.” 302 The plaintiffs proved that the Family-Focused Events violated the Elections Clause. By holding seventy-five events on the day of the Special Election, in the schools that served as polling places, Red Clay provided a targeted incentive for parents of Red Clay students to vote. At the same time, Red Clay created a disincentive for elderly and disabled residents to vote. The Family-Focused Events rendered the Special Election unfair and unequal. 1. The Family-Focused Events As A Targeted Incentive For Voting The Family-Focused Events operated as targeted incentives for voting by parents of Red Clay students. By providing a reward for voting to the group most likely to support the referendum, Red Clay violated the Elections Clause. a. The Relevant Content Of The Elections Clause An election in which the government provides a particular group with targeted incentives to vote is neither fair nor equal. This conclusion flows from multiple sources. i. The Text Of The Elections Clause And Pertinent Judicial Interpretations The Dismissal Ruling analyzed the text of the Elections Clause and Delaware decisions that have discussed voting rights.303 The one Delaware decision to address the Elections Clause states that its purpose is “to ensure that the right of citizens to vote in an 302 Del. Const. art. I, § 3. 303 See 122 A.3d at 837-38, 855. 82 election is unfettered.”304 Other Delaware decisions on voting rights call for interpreting election-related laws liberally to protect voters.305 These decisions suggest that the Elections Clause should be interpreted broadly, but they do not provide more concrete guidance. The Dismissal Ruling also examined decisions from other jurisdictions that have interpreted their comparable constitutional provisions.306 These decisions observe that the operative question under the Elections Clause is whether the outcome represented “a full, fair, and free expression of the popular will upon the matter, whatever it may be.”307 They also teach that that it does not matter “whether the cause that prevented persons legally entitled to vote from exercising the right of suffrage was due to some imperfection or insufficiency in the statute regulating the conduct of elections or to fraud, intimidation, 304 Abbott v. Gordon, 2008 WL 821522, at *19 (Del. Super. Mar. 27, 2008). 305 Republican Party of Del. v. Dep’t of Elections, 792 A.2d 224, 226 (Del. Super. 2001) (“Election laws are to be construed liberally because of their importance to the public’s right to vote.”); Bartley v. Davis, 1986 WL 8810, at *9 (Del. Ch. Aug. 14, 1986) (Allen, C.) (“Election laws are not merely technical creatures creating or regulating private rights. They are of transcending public importance, touching upon—indeed giving vitality to—the most fundamental of our rights.”), aff’d, 519 A.2d 662 (Del. 1986). 306 See 122 A.3d at 838-41. 307 Wallbrecht v. Ingram, 175 S.W. 1022, 1026 (Ky. 1915). The State of Kentucky has a particularly well-developed elections-clause jurisprudence. Dismissal Ruling, 122 A.3d at 838; see Gunaji v. Macias, 31 P.3d 1008, 1016 (N.M. 2001) (citing Kentucky as having “the most developed jurisprudence of any state on what [the free and equal elections] clause means in relation to ballot problems”). 83 violence, bribery, or other wrongdoing.”308 One precedent states that “[e]lections are free when the voters are subjected to no intimidation or improper influence, and when every voter is allowed to cast his ballot as his own judgment and conscience dictate.”309 These decisions indicate that the Elections Clause protects against improper, external influences on voting, but they offer little more. Having reviewed these authorities, the Dismissal Ruling sought more specific sources of authority: When imbuing terms like [“free” and “equal”] with content, I believe a judge should strive to be guided by more than the judge’s own subjective views about what they should mean. The judge instead should look to embodiments of those concepts that have been deeply and widely endorsed, such as indications from other provisions of the Delaware Constitution (including its overall structure), state statutes, and longstanding doctrines of common law.310 308 Wallbrecht, 175 S.W. at 1026. 309 People v. Hoffman, 5 N.E. 596, 599-600 (Ill. 1886). The State of Illinois is another jurisdiction with a well-developed elections-clause jurisprudence. Dismissal Ruling, 122 A.3d at 839. Dismissal Ruling, 122 A.3d at 841. The Dismissal Ruling also noted that “several of the older cases interpreting the Elections Clauses of other states cite Thomas M. Cooley, A Treatise on the Constitutional Limitations which Rest Upon the Legislative Power of the States of the American Union (8th ed. 1927). It seems likely that this treatise and others would shed additional light on the meaning of the Elections Clause.” Dismissal Ruling, 122 A.3d at 842 n.65. The parties did not take the hint, but in the interest of completeness, I secured a copy of Cooley’s treatise. The work states that “[t]o keep every election free of all the influences and surroundings which might bear improperly upon it, or might impel the electors to cast their suffrages otherwise than as their judgments would dictate, has always been a prominent object in American legislation.” Cooley, supra, at 1389 (footnote omitted). It cites the state free-and-equal elections clauses and related statutes as “looking to the accomplishment of [this] general purpose.” Id. at 1390. The treatise mentions, as one of the specific restrictions adopted in support of this end, that “the treating of an elector, 310 84 For purposes of analyzing incentives to vote, the most pertinent sources are the history of the Elections Clause, a related provision in the Delaware Constitution that criminalizes bribery in elections, and Delaware statutes addressing similar electoral misconduct. These authorities point towards a longstanding Delaware public policy against voters receiving rewards for voting in elections. ii. The Legislative History Of The Elections Clause If the meaning of a constitutional provision is unclear, a court may consider its legislative history.311 The legislative history of the Elections Clause “is largely coterminous with the development of the Delaware Constitution.”312 Delaware has had four constitutions, adopted respectively in 1776, 1792, 1831, and 1897. They are not separate and independent, but rather linked.313 The Constitution of 1897 continues in force today. “The antecedents of the Elections Clause can be seen in Delaware’s first with a view to influence his vote, is in some States made an indictable offense.” Id. After summing up various examples, the treatise observes that “generally all such precautions as the people in framing their organic law, or the legislature afterwards, have thought might be made available for the purpose, have been provided with a view to secure the most completely free and unbiased expression of opinion that shall be possible.” Id. at 1391. The treatise thus supports examining critically and approaching with skepticism referendumday events at polling locations that are designed to influence a subset of the electorate. 311 In re Request of Governor for Advisory Opinion, 950 A.2d 651, 653 (Del. 2008); Opinion of the Justices, 290 A.2d 645, 647 (Del. 1972). 312 Dismissal Ruling, 122 A.3d at 819. 313 Maurice A. Hartnett, III, Delaware’s Charters and Prior Constitutions [hereinafter Delaware’s Charters], in First 100 Years 21, 23. 85 constitution.”314 “In May of 1776 the Continental Congress passed a resolution that advised the colonies to form new governments.”315 “A widespread concern before the Declaration of Independence had been a desire for popular control over the process of governing.”316 The Constitutional Convention that convened in August 1776 began by drafting a Declaration of Rights and Fundamental Rules of the Delaware State (the “Declaration of Rights”), which the delegates adopted on September 11.317 The Declaration of Rights emphasized the role of the people and the importance of elections. Section 1 stated that “all government of right originates from the people, is founded in compact only, and instituted solely for the good of the whole.”318 Section 6 contained a predecessor to the Elections Clause, which stated that “the right in the people to participate in the Legislature, is the foundation of liberty and of all free government, and for this end all elections ought to be free and frequent.”319 When the delegates adopted Delaware’s first constitution on September 20, 1776, they incorporated the Declaration of Rights.320 Article 30 stated: “No article of the declaration of rights and fundamental rules of this state, agreed to by this 314 Dismissal Ruling, 122 A.3d at 820. 315 Purpose & Function at 5. 316 Id. at 6. 317 Holland, Delaware State Constitution at 6. 318 Declaration of Rights & Fundamental Rules of the Delaware State § 1 (1776). 319 Id. § 6. 320 Holland, Delaware State Constitution at 7; Delaware’s Charters at 28. 86 convention, . . . ought ever to be violated on any pretense whatever.”321 “After the ratification of the United States Constitution, states began the process of re-writing their own constitutions.”322 Delaware’s Constitutional Convention took place in two sessions, one in 1791 and a second in 1792.323 Most significantly for present purposes, the drafters of the Constitution of 1792 updated the Declaration of Rights.324 The Elections Clause remained part of the new declaration.325 The next significant development for the Elections Clause was the Constitution of 1897.326 “Concern about the legitimacy of state and local elections played a significant role in the prompting Delaware to convene a constitutional convention.327 “[E]lections and 321 Del. Const. of 1776, art. 30. 322 Dismissal Ruling, 122 A.3d at 820. 323 Holland, Delaware State Constitution at 10-11. 324 Dismissal Ruling, 122 A.3d at 821. 325 Id. at 820. 326 The intervening Constitution of 1831 made relatively minor changes to the Constitution of 1792, which principally involved reorganizing the judiciary. Holland, Delaware State Constitution at 12-13. The Constitution of 1831 is thus “better seen as a modification of the 1792 Constitution.” Id. at 15. See, e.g., id. at 24 (“[R]eforming the political system had helped spur calls for the convention. Vote buying and election fraud were considered rampant.”); William H. Williams, Delaware in the 1890s, in First 100 Years 45, 53 (describing the decade of the 1890s as “probably the most politically corrupt in the state’s history”); Henry R. Horsey et al., The Delaware Constitutional Convention of 1897, at 58 [hereinafter Delaware Constitutional Convention of 1897], in First 100 Years 55, (noting that in the years leading up to the Constitutional Convention of 1896-97, “abuse of the poll tax and rampant votebuying threatened to undermine the foundations of representative government”); Rodman Ward Jr. & Paul J. Lockwood, Bill of Rights: Article I [hereinafter Bill of Rights], in First 327 87 election politics, which had been a powerful impetus for calling the convention, were never far from [the delegates’] minds.”328 One of the convention’s standing committees was the Committee on Securing the Purity of the Ballot.329 Despite their focus on electoral issues, the delegates did not make changes to the Elections Clause. The report from the committee charged with reviewing the Declaration of Rights stated: This [Declaration of Rights] is regarded, astonishingly and with great unanimity, by the Members of the Convention, as almost the same document. Gentlemen of the Convention are so earnest and anxious that they may transmit this valuable relic of the former centuries to their children and grandchildren, and they might point to themselves with pride, that they have left it simply intact, scarcely a dot from the i or a cross from the t being omitted.330 Rather than changing the Declaration of Rights, they addressed elections elsewhere.331 One key provision became Section 7 of Article V of the Constitution of 1897 (the “Anti-Bribery Clause”), which defined substantive election offenses and fixed the penalties 100 Years 73, 80 (“The delegates at the 1897 Convention met in the shadow of a very recent history of election bribery in the state.”); Joseph T Walsh & Thomas J. Fitzpatrick, Jr., Judiciary: Article IV, in First 100 Years 121, 123 (“The period immediately preceding the convention was noted by some historians as a time of political scandal, particularly in the election process.”). Delaware Constitutional Convention of 1897, at 61; see id. at 62 (“Another of the convention’s primary concerns was to rescue the state’s election process from the rampant fraud which pervaded elections in the Reconstruction period.”). 328 329 Id. at 61. 330 4 Debates and Proceedings of the Constitutional Convention of the State of Delaware 2386 (1958). 331 Delaware Constitutional Convention of 1897, at 64; Bill of Rights at 80-81. 88 for violations. It stated: Every person who either in or out of the State shall receive or accept, or offer to receive or accept, or shall pay, transfer or deliver, or offer or promise to pay, transfer or deliver, or shall contribute, or offer or promise to contribute, to another to be paid or used, any money or other valuable thing as a compensation, inducement or reward for the giving or withholding, or in any manner influencing the giving or withholding, a vote at any general, special, or municipal election in this State, or at any primary election, convention or meeting held for the purpose of nominating any candidate or candidates to be voted for at such general, special or municipal election . . . shall be deemed guilty of a misdemeanor, and shall be fined not less than one hundred dollars nor more than five thousand dollars, or shall be imprisoned for a term not less than one month nor more than three years . . . ; and shall further for a term of ten years next following said person’s sentence, be incapable of voting at any such general, special, municipal or primary election or convention or meeting . . . .332 Under this clause, it became a constitutional offense to provide money or any other valuable thing as a means of influencing in any manner the giving or withholding of votes. The breadth of the Anti-Bribery Clause reflected the “practices of early nineteenthcentury politicians who roused their followers to partisan enthusiasm with plentiful and free liquid requirements.”333 Since the voter had to be a taxpayer, the candidate who had money was tempted to help a poor but faithful supporter by paying his tax bill. He might go a bit further, by giving him a drink of liquor and perhaps two dollars. Sometimes a hat, a coat, or a pair of boots would do as a gift, or a dress for a man’s wife.334 332 Del. Const. art. V, § 7 (amended 1999). 333 John A. Munroe, History of Delaware 173 (5th ed. 2010). 334 Id. 89 “Another path for corruption was the Voters’ Assistant Law of 1891, which was intended to provide a legitimate service for illiterate or otherwise handicapped voters by authorizing each political party to provide a voters’ assistant in a polling place who could help a voter who asked in marking his ballot.”335 The voter who wanted to sell his vote would request assistance in marking his ballot. When he had properly marked and deposited his ballot, he would be given some token as a sort of receipt for his vote, along with instructions as to where to find the party treasurer, who would exchange an agreed sum for the token. In one case that became notorious because it was described in a court of law, the payoff token was a salted chestnut, handed at the polls to a voter who would then exchange it for his payoff—usually five or ten dollars—at a nearby livery stable.336 To address these practices, the Anti-Bribery Clause made it a constitutional offence to use “any money or other valuable thing” to “influenc[e]” voting or as a “compensation, inducement or reward” for voting.337 A violation of the Anti-Bribery Clause does not require a “corrupt motive.”338 335 Dismissal Ruling, 122 A.3d at 823. 336 Munroe, History of Delaware at 173; see Delaware Constitutional Convention of 1897, at 63 (explaining that “a candidate or party could pay the tax for a voter who would not otherwise qualify (or could simply provide the voter with a falsified tax receipt)” and that “it was common practice to provide citizens with a tax receipt, together with a premarked ballot and perhaps a drink or other small gift once the duty was performed”). 337 Del. Const. art. V, § 7. 338 State v. Collins, 42 A. 619, 622 (Del. Ct. Gen. Sess. 1898); see Holland, Delaware State Constitution at 203. The Anti-Bribery Clause applies to “any general, special, or municipal election.” Del. Const. art. V, § 7. The Delaware Code describes a school referendum on taxes as a “special election.” See 14 Del. C. § 1903 (“Before any school board levies a tax . . . it shall . . . call a special election to be held at the polling place or places designated by the Department of Elections conducting the election.”). By its plain language, the Anti-Bribery Clause would seem to apply to a school referendum. But as 90 When the delegates adopted the Anti-Bribery Clause, statutes proscribing bribery in elections already existed in Delaware. “Despite the existence of these statutes, the delegates to the Constitutional Convention of 1896-97 regarded the integrity of elections as a serious problem.”339 So they made it a constitutional criminal office to provide “any money or other valuable thing” as a “compensation, inducement or reward” for voting.340 Viewed as a whole, the evolution of the Delaware Constitution from 1776 until 1897 evidences consistent concern for the integrity of the electoral process, particularly in response to widespread nineteenth-century practices in which voters received items of value in return for votes. The legislative history indicates that that an election in which certain voters receive an inducement for voting is not “free and equal.” noted in the Dismissal Ruling, the Delaware Attorney General opined in 2006 that the AntiBribery Clause did not apply to a school referendum. See Del. Op. Att’y Gen. 06-IB04, 2006 WL 1242015 (Mar. 23, 2006). The opinion relied on the text of the Anti-Bribery Clause, without discussing the authority that refers to a school tax referendum as a “special election.” In the same opinion, the Attorney General suggested that the plaintiffs might have a cause of action under Section 5162 of Title 15. That section also applies in “any general, special or municipal election,” 15 Del. C. § 5162, creating tension with the opinion’s prior analysis of the Anti-Bribery Clause. 339 Dismissal Ruling, 122 A.3d at 825; see id. at 824-25 (discussing statutes). 340 The delegates removed the right of jury trial for the constitutionalized criminal offense. “Although it seems to be little more than a historical anecdote today, [this] may have been the most vigorously debated section of the 1897 Constitution.” Bill of Rights at 80; see id. at 81-82. The delegates resisted limiting the fundamental right to trial by jury, but they ultimately regarded the step as necessary “because juries were unwilling to convict in political cases.” Id. at 80 (citing 1 Debates and Proceedings of the Constitutional Convention of the State of Delaware 503 (1958)). 91 iii. Related Statutory Frameworks The Delaware statutory frameworks that govern elections also provide guidance for applying the Elections Clause. Civil and criminal statutes that regulate elections reflect public policy determinations about what free and equal elections should look like. Criminal statutes in particular provide evidence of the floor for permissible electoral conduct.341 In other areas of the law, courts have used criminal statutes to inform how the civil law should apply342 and to flesh out constitutional parameters.343 Delaware has several statutory frameworks for challenging particular types of elections.344 For most elected offices, the statutory procedure permits the losing candidate to challenge the election “[w]hen the person whose right is contested has given to any elector . . . any bribe or reward or shall have offered any bribe or reward for the purpose of 341 To reiterate a point made in the Dismissal Ruling, this decision is not seeking to enforce the criminal laws as such. This is not a criminal case, the task of enforcing the criminal statutes lies with the Attorney General, and his office enjoys “broad discretion as to whom to prosecute.” Albury v. State, 551 A.2d 53, 61 (Del. 1988). This court neither has jurisdiction over criminal proceedings, nor the equitable authority to involve itself in criminal proceedings. See Econ. Cleaners v. Green, 184 A. 225, 226 (Del. Ch. 1936) (Wolcott, C.) (“[A]s a general rule courts of equity have no jurisdiction to interfere by injunction with the enforcement of the criminal laws of the State by its duly constituted officers.”). 342 See Dismissal Ruling, 122 A.3d at 843 (discussing role of criminal statutes in establishing the standard for negligence). See id. (discussing judicial consideration of criminal statutes to determine scope of the right to privacy). 343 344 See 15 Del. C. § 5901 (challenge to election as member of General Assembly); 15 Del. C. § 5921 (challenge to election of persons chosen as electors of the President or Vice President of the United States); 15 Del. C. § 5941 (challenge to other offices). 92 procuring his or her election.”345 This provision is part of Title 15, the purpose of which is “to assure the people’s right to free and equal elections, as guaranteed by our state Constitution.”346 Widespread conduct that violates the provisions of Title 15 is therefore inconsistent with the “right to free and equal elections.” It follows that an election in which one side provides voters with widespread rewards for voting violates the Elections Clause. Title 14 of the Delaware Code proscribes certain conduct in school elections. It includes the following prohibition: No person who receives or accepts or offers to receive or accept, or pays, transfers or delivers, or offers or promises to pay, transfer or deliver, or contributes or offers or promises to contribute to another to be paid or used, any money or other valuable thing as a compensation, inducement or reward for giving or withholding or in any manner influencing the giving or withholding a vote at any public school election, shall vote at such election . . . .347 Based on this provision, if a participant in a school election has engaged in a widespread practice of providing “any money or other valuable thing” as an inducement for voting, then the election has not been “free and equal” under the Elections Clause. As with the evolution of the Delaware Constitution, the statutory schemes governing elections indicate that the Elections Clause prohibits inducements to vote. An election in 345 15 Del. C. § 5941. 346 15 Del. C. § 101A. 347 14 Del. C. § 1079(a). The statute permits a resident who has been challenged under the statute to vote if the person executes an oath or affirmation attesting to compliance with the statute. 14 Del. C. § 1079(a)-(b). 93 which certain voters receive money or other valuable things as an inducement for voting is not “free and equal.” b. Application To The Trial Record The Family-Focused Events provided parents of Red Clay students with an inducement to go to the polls and vote in the Special Election. The inducements did not appeal to other parts of the electorate. As such, the Family-Focused Events functioned as targeted rewards for voting and violated the Elections Clause. From the outset, the Family-Focused Events were a central part of Red Clay’s campaign strategy. Superintendent Daugherty required that all Red Clay principals attend the campaign kick-off meeting on November 6, 2014,348 when the principals were told that they “needed to have an event” on the day of the Special Election.349 They also were told that each school would have a parent leader whose responsibilities included “[w]orking with Principals for Referendum Day Events.”350 Ammann required that each principal submit a School Referendum Plan that included “school activities planned for the day of 348 JX 28 (Superintendent Daugherty telling all principals to attend); Tr. 114 (Penoyer testifying that she understood the meeting was mandatory); Tr. 489 (Mathis testifying that he only attended because it was mandatory). 349 Tr. 114 (Penoyer testifying that her understanding from the meeting was that “every school needed to have an event on the day of the referendum”). 350 JX 25 at D2556. 94 the 2015 Referendum.”351 Every school principal submitted a plan.352 Ammann reviewed the plans, approved them or asked for changes, then forwarded them to Johnson.353 Every school that was a polling place held at least one—and typically two or more— Family-Focused Events.354 The one school that was not designated as a polling place did not hold any family-oriented events on the day of the referendum.355 The principal of that school explained in his School Referendum Plan, “Since we are not a polling site, I will support [Stanton] for any activities.”356 In total, the Red Clay schools held seventy-five Family-Focused Events on the day of the referendum.357 These events were intended to bring parents into the schools to vote.358 They were not designed to appeal to the community in general. Baltz held a 351 See JX 307 (compilation of all plans). 352 Tr. 625, 628 (Ammann). 353 Tr. 595-98 (Ammann). 354 JX 301; Tr. 625 (Ammann). 355 PTO ¶ 28; see Tr. 624-26 (Ammann testifying that the principal of the one school that was not designated as a polling place reallocated resources to supporting two of the schools that were polling places in hosting their events). 356 JX 307 at D3716. 357 See JX 301(spreadsheet identifying events, target audience, likely attendance, and other details). See Daugherty Dep. 203 (confirming that the Family-Focused Events were “get out the vote events”); JX 176 at D14676 (Superintendent Daugherty referring to FamilyFocused Events as “[g]et out the vote activities”); JX 56 (principal of A.I. duPont writing in his School Referendum Plan, “We will find opportunities during the studnet [sic] day to draw families in”); JX 102 (teacher at Heritage explaining to parent that the “Author’s 358 95 “Pajama Jammie Jam”—a pajama dance party with pizza and raffles.359 Other evening events similarly targeted students and their families, such as “Bedtime Stories,” a “Winter Blues Beach Bingo,” a “Winter Carnival,” and “Family Fun Night.”360 Some Family-Focused Events were nominally open to the public, but Red Clay did not expect them to attract anyone other than students and their families.361 Of the seventyfive Family-Focused Events, only three identified the “community” or the “public” as part of their event’s target audience.362 Some events, like the Pajama Jammie Jam, were closed to the public due to security concerns.363 Although all of the Family-Focused Events functioned as rewards for parents to visit the polls, some schools offered more tangible rewards. One example was the “no-uniform” passes at Richardson Park. Each student received one no-uniform pass for each “voting Teas” were “one of the activities to help get parents in to school so they will vote in the referendum”); JX 313 at D20467 (Johnson stressing that during the Family-Focused Events, principals should be “reminding your families to go into that poll and vote!”); Tr. 351 (Johnson confirming that she “didn’t want any parents to come to an event and then leave without voting”). 359 Tr. 120, 125 (Penoyer). 360 See JX 301; JX 234; JX 235. JX 185 (Nash acknowledging that “I don’t know [that people without children in the district] would, you know, have a desire to come to Family Bingo Night”); Tr. 186-87 (same). 361 362 JX 301. 363 Tr. 123, 145-46 (Penoyer). 96 adult.”364 There was no limit on how many passes a child could accumulate, so if a child brought four voting adults, the child received four passes.365 The principal explained that because Richardson Park was a uniform school, the “no-uniform pass” was perhaps the most valuable reward he could provide for his students.366 He believed the passes also served as a reward for the students’ families, because one of the no-uniform days was “Picture Day,” which provided “a chance for families to get their child’s picture out of uniform!”367 Richardson Park also offered prizes in its bingo games valued at over $1,000. The principal believed that the passes, prizes, and activities would “make it a very incentivized night for parents and kids to come out.”368 Baltz took a similar approach. When a family entered the school, they received a checklist with boxes for “I Ate, I Voted, I Danced.” Below the boxes were lines for the “Student Name,” the “Parent Name,” and a phone number.369 One of the purposes of the card was to give each family a list of things to do.370 Teachers and staff had stamps to check 364 Tr. 491, 494 (Mathis); see JX 66; JX 80; JX 108; see also Tr. 494-95 (Mathis) (Q: “And if a parent didn’t vote, they didn’t get a pass. Correct?” A: “Technically.”). Tr. 494-95 (Mathis); see JX 108 (“[I]t will be 1 free uniform pass for each voting adult . . . 4 adults – 4 passes . . . .”). 365 366 See Tr. 492-94 (Mathis testifying about importance of pass). 367 JX 108; Tr. 495 (Mathis). 368 JX 66. 369 JX 275. 370 Tr. 125, 126-27 (Penoyer). 97 off items on the card. At the end of the night, the completed checklists doubled as raffle tickets.371 Although the Baltz principal denied that families had to check off all three boxes to enter the raffle, a community witness testified that adults near the voting area were checking off the boxes, and other adults were asking the children whether their parents voted before they entered the line for free pizza.372 The trial record established that the Family-Focused Events operated as a targeted reward that induced parents of Red Clay students to vote. As such, they violated the Elections Clause. 2. The Family-Focused Events As Impediments To Voting The Family-Focused Events also operated as impediments to voting by the elderly and disabled. Through this mechanism, they violated the Elections Clause. a. The Relevant Content Of The Elections Clause The Elections Clause prohibits widespread impediments to voting. An election in which a government inhibits voting by particular groups is neither fair nor equal. That is particularly true when the group is part of a protected class, like the elderly and disabled.373 371 JX 275; Tr. 124-25 (Penoyer). Compare, e.g., Tr. 127 (Penoyer) (“[It] didn’t have to be all checked. It doesn’t matter. It’s elementary school. . . . [T]hey can check it, or they don’t have to check it.”), with Tr. 302-03 (McHugh) (“[T]here was a woman standing there talking as people came out with their children to make sure—I don’t know the exact words, but something to the effect of, you know, ‘Did your parents vote?’ You know, ‘Here’s your ticket and’—to go in and get pizza.”). 372 373 See 6 Del. C. § 4504(a). 98 The conclusion that the Elections Clause prohibits widespread impediments to voting flows from multiple sources. i. Traditional Legal Authorities Both case law and analogous statutory frameworks indicate that a government violates the Elections Clause when it inhibits voting by particular groups. The first source is the Abbott decision, which held that the purpose of the Elections Clause is “to ensure that the right of citizens to vote in an election is unfettered.”374 The court dismissed the claim in that case because the candidate had not alleged that voters’ “access to the polls was disturbed”375 The Abbott decision indicates that sufficiently serious and widespread impediments to accessing the polls could result in an election that was not “free and equal.” Another source of authority is Section 1087 of Title 14, which prohibits “[e]lectioneering as described in § 4942 of Title 15 . . . in any public school election.”376 Section 4942(a) of Title 15 states: No election officer, challenger or any other person within the polling place or within 50 feet of the entrance to the building in which the voting room is located shall electioneer during the conduct of the election. No political headquarters or gathering shall be permitted within that building during the conduct of the election.377 374 Abbott, 2008 WL 821522, at *19 (citing State ex rel. James v. Battersby, 56 A.2d 527, 532 (Del. Super. 1947)). 375 Id. at *20. 376 14 Del. C. § 1087. 377 15 Del. C. § 4942(a) (the “Anti-Electioneering Statute”). 99 Under the statute, electioneering includes but is not limited to political discussion of issues . . . or partisan topics, the wearing of any button, banner or other object referring to issues . . . or partisan topics, the display, distribution or other handling of literature or any writing or drawing referring to issues . . . or partisan topics, the deliberate projection of sound referring to issues . . . or partisan topics from loudspeakers or otherwise into the polling place or the area within 50 feet of the entrance to the building in which the voting room is located.378 Based on this provision, when widespread electioneering interferes with voters’ ability to access the polls, the election has not been “free and equal” for purposes of the Elections Clause.379 ii. Evidence of Custom and Practice An equally important source of authority regarding impediments to voting is Delaware custom and practice.380 Barbara Lippincott of the Department of Elections 378 15 Del. C. § 4942. 379 Dismissal Ruling, 122 A.3d at 845. See Agostini v. Colonial Tr. Co., 44 A.2d 21, 22 n.1 (Del. Ch. 1945) (“[T]he existence of a usage or custom may call for, and influence, a construction of the statute.”); see also In Re Adoption of Keith M.W., 79 P.3d 623, 637 (Alaska 2003) (Matthews, J. concurring) (“The meaning attached by people affected by an act may have an important bearing on how it is construed.” (citation and quotation marks omitted)); Bd. of Exam’rs in Optometry v. Spitz, 479 A.2d 363, 370-71 (Md. 1984) (“In this case, the Court is faced with an ambiguous statute that has been given a long-standing interpretation by those affected by it and arguably by the public at large . . . . Such long-standing practice should not be disturbed in the absence of an unambiguous statute.”); People v. Nguyen, 161 Cal. App.3d 687, 692 (Cal. Ct. App. 1984) (“Requisite standards of certainty can often be fleshed out from otherwise vague statutory language by reference to . . . long established or commonly accepted usage . . . .” (internal alterations and quotations omitted)); Hennessey v. Pers. Fin. Corp. of New York, 26 N.Y.S.2d 1012 (N.Y. Sup. Ct. 1941) (“A practical construction given a statute by the public generally, as indicated by a uniform course of conduct over a considerable period of time, and acquiesced in and approved by a public official charged with the duty of enforcing the act, is entitled to great weight in . . . interpretation.”); 2A 380 100 testified at trial. She has overseen thirty-five school referendums since 1999, including the Special Election.381 To govern the use of Red Clay schools as polling places, the Department of Elections and Red Clay entered into a “standard polling place contract” for each of the schools where voting took place.382 Lippincott explained that under the contracts, schools are permitted to have events during the referendum, and the Department of Elections understands that “generally events will be held on the day of [a] referendum.” 383 This makes sense, because a school referendum does not take place on a weekend or holiday. School is in session, and school activities continue. Each polling place contract provides, however, that any activities conducted in the building designated as a polling place must not “interfere with the voting process.”384 Each contract likewise provides that the activities cannot “hinder[] access to the voting area and/or building.”385 Lippincott confirmed that a school cannot “hold an activity that Norman J Singer & Shambie Singer, Sutherland Statutes and Statutory Construction § 49.5 (7th ed. 2008) (“[C]ourts look to ‘established usage’ and the meaning attached by people affected by an act, and the public at large, for persuasive interpretative guidance.” (footnotes omitted)). 381 Tr. 379, 390 (Lippincott). 382 Id. at 392; see JX 282 (compiling contracts). 383 Tr. 396 (Lippincott). 384 JX 282. 385 Id. 101 interferes with the voting process.”386 She similarly confirmed that the “activities within the building also cannot hinder access to the voting area and/or the building.”387 During the lead-up to the Special Election, Lippincott warned Ammann against conflicts between activities at the schools and the voting process.388 The standard polling place agreements and the Department of Elections’ custom and practice make clear that the limitation on obstructing voter access applies to any entrance to the building where the voting room is located and requires a fifty-foot electioneeringfree zone around the voting room, including inside the building where the voting room is located. Each of the polling place agreements designated the school building as the polling place for purposes of the Special Election.389 Lippincott agreed that for the Special 386 Tr. 398; accord id. at 399. 387 Id. 388 JX 36; see Tr. 400-01 (Lippincott). 389 Each contract began by listing the school and its street address, then included the following text: “We hereby agree to permit the above named building to be used as a polling place for the RED CLAY CONSOLIDATED SCHOOL DISTRICT REFERNDUM.” JX 282. Other Delaware statutes similarly indicate that when an election statute refers to a “polling place,” it means the building as a whole where voting will take place. See 14 Del. C. § 1072(b) (requiring Department of Elections to “designate and procure the buildings within the district . . . which shall be used as polling places for any public school election” and stating that “[w]henever possible, such polling place shall be located in public buildings, which shall include suitable schools”); id. at § 1073 (making election inspector “responsible for the setup, operation and closing of the polling place”); 14 Del. C. § 1082 (describing procedures for voters “at the polling place”); 14 Del. C. § 1088 (identifying the persons who shall be “admitted within the voting room”). 102 Election, each school building was the operative polling place.390 Each of the polling place agreements also specified that electioneering cannot take place “within 50 feet of any entrance to the building in which the voting area is located.”391 This restriction applied by its terms to any entrance to the building, not just those specifically designated for use by voters. Lippincott confirmed that the ban on electioneering applied not only to the specific entrances designated for voting, but to “any entrance used by a voter.”392 Each of the polling place agreements further specified that any activities held in the buildings could not involve “[g]atherings advocating or opposing any proposal or candidate on the ballot.”393 Lippincott agreed that the ban on electioneering is “not just limited to the area outside the building.”394 Lippincott further agreed that because the tax increase was a proposal on the ballot for the Special Election, an event advocating in favor of the proposal should not have been held in a school that was designated as a polling place.395 Lippincott also explained that the need to ensure access to the polls extends to parking. The Department of Elections expects that the spaces designated for voters will be reserved for voters and not used by individuals attending other activities at the polling 390 Tr. 398-99. 391 JX 282. 392 Tr. 388. 393 JX 282. 394 Tr. 407. 395 Tr. 406-07. 103 place.396 When communicating with Ammann about the spots, Lippincott instructed Ammann to monitor the spaces “so that event attendees will not occupy those spaces.” 397 Lippincott’s testimony and the content of the polling place agreements indicate that an election is not “free and equal” when a government holds gatherings inside the buildings designated as polling places, on the day of the election, that promote the government’s position on an issue being considered by the voters. Gatherings of this type interfere with the ability of electors to vote by turning the polling places into partisan outposts. The interference becomes more pronounced when residents are subjected to electioneering activities from within the polling places. The same sources indicate that an election is not “free and equal” when widespread activities inside the polling places have the practical effect of impeding physical access to the polls. b. The Family-Focused Events Interfered With Voter Access. The plaintiffs proved at trial that the Family-Focused Events interfered with voter access to a degree that violated the Elections Clause. The Family-Focused Events were partisan gatherings held in polling places on the day of the election, they resulted in electioneering activity inside the polling places, and they produced parking problems that hampered the ability of the elderly and disabled to access the polls. Tr. 402-03 (Lippincott). “[T]he voter spots are only for people while they are actively voting. If they are going to be going to an event, they should park in a voter spot, go in and vote, and then they come out and move their car . . . .” Id. 396 397 JX 121. 104 In substance, the Family-Focused Events were partisan gatherings. Red Clay held the seventy-five Family-Focused Events on the day of the Special Election in the school buildings designated as polling places. The events were designed to support the tax increase that was on the ballot, contrary to the terms of the polling place agreements. They Family-Focused Events involved electioneering activity inside the polling places and within fifty feet of the entrances to the polling places. One clear example was the “Support the Baltz Bear” sign that the Baltz principal placed behind a table at the main entrance to the school, despite knowing that voters often used that entrance.398 Another example took place at Richardson Park, where the principal recruited parents to tell other parents to vote “YES” and encourage them to bring other “positive voters” to the polls.399 He also instructed his teachers to “mingle” with parents “to make sure they voted” and to “say how passing the referendum will help children” because “[h]aving someone ask if they voted yet and telling the importance of the vote and how little time it will take to make a difference for students is important.”400 398 JX 162; Tr. 133-34, 136 (Penoyer). Red Clay has argued that the sign was more than 100 feet from the actual voting room and was not near the entrance that the Department of Elections officially designated for use by voters. Tr. 607 (Amman); JX 323. The Red Clay agreements and the Department of Elections’ custom and practice prohibited signage and other electioneering activities within fifty feet of any door to the polling place, i.e., the school building, that could be used by voters. 399 Tr. 500-01, 505, 507-08 (Mathis); JX 108. JX 140; accord Tr. 505, 507 (agreeing that “it is harder to say no to someone who is directly asking you, essentially, ‘Is your child important enough to take 5 minutes to vote?’”). 400 105 Red Clay did not present any evidence suggesting that these were the actions of two rogue principals. Ammann himself approved the placement of the Baltz sign using an overly narrow interpretation in which electioneering only was prohibited within fifty feet of an entrance specifically designated for use by voters.401 Recruiting parents to promote the tax increase was part of the Steering Committee’s plan to drive pro-referendum communications through parent volunteers.402 From a big picture standpoint, the whole purpose of holding seventy-five Family-Focused Events was to support Red Clay’s preferred outcome. The Family-Focused Events also interfered with the voting process by hindering the ability of elderly and disabled residents to access the polls. Thousands of students and their family members attended the Family-Focused Events, with many of the evening events drawing hundreds of people.403 The Family-Focused Events resulted in packed parking lots that prevented elderly and disabled residents from finding accessible parking. The plaintiffs presented unrebutted testimony about parking congestion at seven different schools.404 They also presented unrebutted testimony that the congestion 401 Tr. 606 (Ammann). 402 See Tr. 195, 198-99 (Nash). 403 JX 301. 404 Tr. 8-10 (Young testifying about North Star); Tr. 27-28, 34, 53-54 (Hudson testifying about Marbrook); Tr. 89-90 (O’Neill testifying about Marbrook); Tr. 73-74 (Pickering testifying about H.B. duPont and Skyline); Tr. 97-101 (Fitzpatrick testifying about Linden Hill); Tr. 148-53 (Schnell testifying about McKean); Tr. 163-69 (Boyle testifying about Brandywine Springs). 106 prevented would-be voters from casting ballots.405 Despite having employees at every school that was a polling place, Red Clay did not present any persuasive evidence to the contrary. Red Clay’s own attendance estimates supported the plaintiffs’ contentions. For example, Richardson Park’s parking lot could hold about 150 to 200 cars.406 Approximately 300 people attended the bingo and movie night,407 or enough to fill the parking lot. Other evening events similarly drew hundreds of people.408 Red Clay did not fulfill its obligation to monitor the parking spaces designated for voters. Superintendent Daugherty testified that Red Clay did not monitor who was using the parking spaces.409 Ammann admitted that he did not instruct the principals to monitor the parking situation.410 The plaintiffs proved that parking problems prevented some elderly and disabled residents from voting.411 Tr. 8-11 (Young); Tr. 89-90 (O’Neill). This decision does not rely on Pickering’s testimony about his parents’ decision not to vote, which is hearsay. Tr. 73 (Pickering). 405 406 Tr. 488 (Mathis). 407 JX 301; Tr. 499 (Mathis). 408 See JX 301. 409 Daugherty Dep. 205. 410 Tr. 632 (Ammann testifying that he did not speak with the principals about monitoring parking access). As discussed in the Statement of Facts, Ammann testified that he instructed the custodial staff to monitor the parking lots, but there is no contemporaneous evidence that he gave that instruction, and Ammann uses email as a primary means of communication. 411 Tr. 8-11 (Young); Tr. 88-89 (O’Neill). 107 An election is not “free and equal” when one side holds partisan events in the polling places on election day, when the events subject voters to electioneering within the polling places themselves, and when the events interfere with the ability of elderly and disabled residents to access the polls. Consequently, regardless of whether or not the FamilyFocused Events operated as impermissible rewards for voting, the Family-Focused Events violated the Elections Clause. 3. The Family-Focused Events As A Delaware Tradition In a final effort to justify the Family-Focused Events, Red Clay called its former superintendent, Robert Andrzejewski, as an expert witness. Andrzejewski currently serves as superintendent of the Christina School District and has presided over six referendums. He opined that it is “standard practice for referendum campaigns across the State of Delaware [to] include get-out-the-vote activities and voting by eligible students.”412 As one example, Andrzejewski cited an event held in the Smyrna School District called “I Love the Smyrna School District Day.” He explained that on that day, the entire community is invited to a celebration of Smyrna schools, at the same time of which the referendum ballots are cast. The voting, the polls are open beyond the activity. However, that’s a significant activity that that district uses to generate excitement around their schools, and hopefully the people come out to support their community.413 412 Tr. 515. 413 Id. at 522-23. 108 On cross-examination, however, Andrzejewski conceded that the Smyrna day event was not held at a school that served as a polling place.414 It therefore did not provide an inducement for families to come to a polling place and vote, nor did it operate as an impediment to voting. As his second example, Andrzejewski cited an event held in the Caesar Rodney School District called “Rider Pride Day.” It was on homecoming day, and it was reported that thousands of people came out to celebrate the school and homecoming and everything the same day that the polls were open for a referendum vote. 415 Once again, on cross-examination, Andrzejewski conceded that the Rider Pride Day event was not held at a school that served as a polling place.416 It therefore did not provide an inducement for families to come to a polling place and vote, nor did it create an impediment to voting. Red Clay did not provide any evidence of other school districts that have held widespread events on the day of a special election, in the schools that served as polling places, that were designed to draw parents and families of students to the polls. The plaintiffs submitted evidence that the Brandywine School District held a referendum after the Special Election and did not hold Family-Focused Events.417 414 Id. at 534-35. 415 Id. at 523. 416 Id. at 535-36. 417 See JX 291; JX 292; JX 293; JX 294; JX 295; JX 296; JX 297; JX 298. 109 Andrzejewski testified persuasively that other Delaware school districts hold events to support their referendums. His testimony did not establish that other Delaware school districts run top-down campaigns that involve holding widespread events that will appeal to parents in the schools that serve as polling places on the day of the referendum and which create impediments for the elderly and disabled. This decision concludes that Red Clay violated the Elections Clause by holding the seventy-five Family-Focused Events on the day of the Special Election in the schools that served as polling places. The analysis in this decision does not affect the ability of a school district to hold pro-referendum events on the day of the election in locations that are not polling places. It also does not affect the ability of a school district to hold pro-referendum events in locations that are polling places on days other than election day. D. Red Clay’s Government Campaign Speech Red Clay’s other electoral interventions all involve government campaign speech. Brennan remains the controlling Delaware Supreme Court precedent on government campaign speech during a school referendum. In that decision, the Delaware Supreme Court permitted a school district to engage in limited advocacy in favor of a referendum, but held that the “expenditure of public funds in support of one side” must remain “within reasonable limits.”418 The Delaware Supreme Court also held that the school district’s speech should not venture “beyond [a] factual presentation” to the point of “overstatement 418 104 A.2d at 790. 110 and emotional appeals.”419 The Dismissal Ruling argued for a more relaxed application of the principle for purposes of broadly directed campaign speech that is factually accurate and which openly identifies its source. Red Clay’s government campaign speech exceeded these limitations, whether under Brennan or the more relaxed approach recommended by the Dismissal Ruling. Red Clay engaged in a full-scale political campaign that included extensive campaign speech targeted at both identified groups and individuals. At the same time, Red Clay took steps to avoid communicating with the electorate as a whole so that the election would not become a debate. Red Clay’s disproportionate efforts to mobilize supportive voters meant that the Special Election was not “free and equal” as required by the Elections Clause. 1. The Brennan Decision The Brennan case concerned steps that the Mount Pleasant School District took to promote a school bond referendum. Voters who opposed the tax increase challenged the district’s activities. In describing the district’s conduct, the Delaware Supreme Court first recounted examples of broadly directed campaign speech: During the year 1953 and particularly in September and early October there were held various meetings of Parent-Teacher Associations of certain of the public schools, including the schools of the Mount Pleasant District. At one or more of these meetings there were discussions of the [referendum], by 419 Id. 111 certain school officials. Publicity was given prior to these meetings by statements in the public press, and by circulars . . . .420 The high court characterized these activities as “entirely appropriate and desirable.”421 The Delaware Supreme Court next described what appears to have been instances of targeted campaign speech: In addition, a letter . . . addressed to “Parents and Patrons” on the letterhead of the Mount Pleasant School District, was prepared and sent out by the Superintendent of Schools. This letter is almost wholly factual in its nature . . . . It does contain one or two statements indicating that the proposed building program should in the judgment of the Superintendent of Schools, be supported by the District. In addition a memorandum or circular urging . . . approval . . .was prepared and circulated, and also a circular containing sketches of children and showing the need for immediate action to provide more class rooms, equipment, etc., couched in such language as to appeal to the natural desire of parents for good education for their children.422 The Delaware Supreme Court cautioned against the level of rhetoric in these communications: It does seem to us that some of the publicity urging support of the [referendum], such as the appeals “to get out the vote”, [sic] and the statement that the voters’ approval of the program is “a must” if they are to continue to have a good school system, went somewhat beyond the factual presentation which would have been more decorous in an official statement of the case.423 420 Id. 421 Id. 422 Id. 423 Id. 112 Although it is not fully clear from the opinion, the Delaware Supreme Court appears to have been concerned about both the extent of the district’s advocacy and its appeal to parents as an identifiable group. In ruling on the propriety of these communications, the Delaware Supreme Court largely followed decisions from other jurisdictions that had limited the government’s ability to engage in partisan campaign speech during a referendum. The leading decision was Citizens to Protect Public Funds v. Board of Education,424 an influential opinion written by future United States Supreme Court Justice William Brennan while serving on the New Jersey Supreme Court. In Citizens, a school board decided to issue bonds to finance the expansion of several school buildings. The bond issuance required a favorable vote. The school board spent public funds to print and disseminate an eighteen-page booklet urging voters to “Vote Yes,” with an additional page providing a list of bleak consequences in response to the question, “What Will Happen If You Don’t Vote Yes?”425 A group that opposed the bond issue alleged that the school district had violated New Jersey law. The New Jersey Supreme Court held that the school board’s advocacy was improper, characterizing it as unfair to citizens with different views: [T]he board made use of public funds to advocate one side only of the controversial question without affording the dissenters the opportunity by means of that financed medium to present their side, and thus imperiled the 424 98 A.2d 673 (N.J. 1953). 425 Id. at 674. 113 propriety of the entire expenditure. The public funds entrusted to the board belong equally to the proponents and opponents of the proposition, and the use of the funds to finance not the presentation of facts merely but also arguments to persuade the voters that only one side has merit, gives the dissenters just cause for complaint. The expenditure is then not within the implied power and is not lawful in the absence of express authority from the Legislature. . . . . . . We are persuaded, however, that simple fairness and justice to the rights of dissenters require that the use by public bodies of public funds for advocacy be restrained within those limits in the absence of a legislative grant in express terms of the broader power.426 Despite rejecting the school board’s ability to engage in advocacy, the New Jersey Supreme Court did not insist that the government remain silent. The court recognized that voters need information, that a government body is well positioned to provide it, and that “it is not only the right but perhaps the duty of the body to endeavor to secure the assent of the voters.”427 The court suggested that a government agency could serve this need by spending money to inform voters. For example, it could make “reasonable expenditures” to purchase 426 Id. at 677-78. Notably, by the time of the decision, the referendum had passed, and the court regarded the issue of the booklet as moot. The court nevertheless decided that the legal issue was important enough to warrant addressing. See id. at 676. 427 Id. at 676-77. In this respect, the Citizens court followed an earlier New Jersey decision that had affirmed a ruling permitting government campaign speech. See, e.g., City Affairs Comm. v. Bd. of Comm’rs, 41 A.2d 798, 800 (N.J. Sup. Ct. 1945) (“We think municipalities may, within their discretion and in good faith, present their views for or against proposed legislation or referendum to the people of questions which in their judgment would adversely affect the interests of their residents. To accomplish this purpose we think they may incur expenditures by the publication of pamphlets, circulars, newspaper advertisements or radio addresses and that to do so is a proper governmental function.”), aff’d, 46 A.2d 245 (N.J. 1946). The Citizens court noted that the limitations it imposed on government campaign speech had not been suggested by the earlier ruling. 98 A.2d at 678. 114 radio advertisements that presented both sides’ views on the building expansion. 428 But using government money to try to convince voters to “Vote Yes” was not fair to those who wanted to “Vote No.”429 In Brennan, the Delaware Supreme Court held that “[a] Board of Education charged with the duty of managing the public schools in its district, maintaining them in good condition, and providing for necessary improvements and expansion of the school system, 428 98 A.2d at 676. 429 Id. at 677. Another leading example of this approach is Stanson v. Mott, 551 P.2d 1 (Cal. 1976) (en banc), which relied heavily on Citizens. The Mott decision held that a government agency could not engage in advocacy in connection with a bond referendum, but could expend funds “to provide the public with a ‘fair presentation’ of relevant information” relating to the issue under consideration. Id. at 11. The Mott court recognized the potential difficulties “in attempting to distinguish improper ‘campaign’ expenditures from proper ‘informational’ activities,” but observed that with respect to some items, “the distinction is rather clear.” Id. [T]hus, the use of public funds to purchase such items as bumper stickers, posters, advertising “floats,” or television and radio “spots” unquestionably constitutes improper campaign activity, as does the dissemination, at public expense, of campaign literature prepared by private proponents or opponents of a ballot measure. On the other hand, it is generally accepted that a public agency pursues a proper “informational” role when it simply gives a “fair presentation of the facts” in response to a citizen’s request for information or, when requested by a public or private organization, it authorizes a[n] agency employee to present the department’s view of a ballot proposal at a meeting of such organization. Id. at 11-12 (citations omitted). The Mott court suggested that when the line was unclear, “the determination of the propriety or impropriety of the expenditure depends upon a careful consideration of such factors as the style, tenor and timing of the publication; no hard and fast rule governs every case.” Id. at 12 (footnote omitted). 115 is not prevented from using reasonable publicity to bring the issues before the voters.”430 The school district thus could engage in factual, broadly directed campaign speech. Relying on Citizens, however, the Delaware Supreme Court indicated that the district’s more partisan campaign speech went too far: The expenditure of public funds in support of one side of the matter should be kept within reasonable limits. The temptation to any public official to overemphasize the importance and the urgency of a project which he sincerely believes is necessary is understandable; but overstatement and emotional appeals in circulars and other similar matter prepared and distributed at public expense are to be avoided.431 The Delaware Supreme Court viewed the more targeted communications in Brennan as a form of excessive advocacy. On the facts of Brennan, however, the Delaware Supreme Court declined to set aside the election. In reaching this outcome, the Brennan opinion stressed the involvement of a competing campaign that actively opposed the district’s efforts: At the same time that the Board was circularizing the voters an organization of taxpayers opposed to the referendum was likewise conducting a campaign and was also circularizing the voters. It is a fair inference that upon the whole the issue was fairly presented and debated. 432 430 104 A.2d at 790. 431 Id. 432 Id. at 790-91. 116 The Delaware Supreme Court concluded that there was “no showing whatever that the voters were improperly influenced in any way.”433 The high court therefore allowed the election result to stand. The Dismissal Ruling’s Proposed Update Of Brennan 2. The Dismissal Ruling suggested that the Delaware Supreme Court as currently constituted could be inclined to revisit Brennan and relax the restrictions on advocacy articulated in that case. My reasoning boiled down to the societal developments that have taken place during the intervening sixty-four years: I suspect that in today’s news-and-advertisement-saturated age, the Delaware Supreme Court would expand the [limitations set out in Brennan]. The high court could well regard contemporary voters as more media savvy and less trusting of government than denizens of a seemingly simpler era culturally stereotyped by Leave It To Beaver and Father Knows Best. The electorate of sixty years ago might be seen as imbued with a post-World War II confidence in state institutions and held together by a nascent Cold War fear of communism which, along with other cultural factors, generated a greater respect for institutional authority and hence a greater obligation on the part of institutional actors not to abuse voters’ trust. The initial round of shocks to that outlook included the countercultural awakening of the 1960s and Watergate-era revelations about the misuse and abuse of government power. Each ensuing decade brought shocks and scandals of its own. A more worldly and skeptical twenty-first century electorate could well be viewed as less susceptible to being misled by officials deploying urgent language and making emotional appeals. I also suspect that the Delaware Supreme Court would be concerned about whether litigants could mount challenges to government campaign speech too readily by arguing about largely subjective factors like word choice and tone. It seems to me that if presented with the issue, the high court would expand the degree of deference given to a school district in crafting its speech. A more flexible inquiry would require at a minimum that government 433 Id. 117 statements be both accurate (in the sense of factually supported) and truthful (in the sense that the speaker subjectively believed them). Facts and circumstances like tone and timing would continue to play a role, but more important factors would include (i) whether the communication identified itself transparently as coming from a government source, so that the recipient could take into account the speaker’s goals and interests, (ii) the degree to which recipients could avoid the message or access other points of view, and (iii) the vulnerability of the audience, such as whether it comprised children or youth. A government speaker still might go too far, but it would require more extreme communications.434 The Dismissal Ruling nevertheless followed Brennan for purposes of analyzing Red Clay’s motion to dismiss.435 3. Too Much Partisan Advocacy The plaintiffs did not continue after the pleading stage to object to the tone of Red Clay’s campaign speech. They focused on challenging the targeted nature of Red Clay’s speech. In essence, they contended that Red Clay violated the Elections Clause by running a partisan “stealth campaign” directed primarily at the groups and individuals whom Red Clay believed would support the tax increase. The Brennan decision recognized that a government can violate the Elections Clause by engaging too vigorously in partisan advocacy. The question of “how much” is invariably fact-specific. As Brennan demonstrates, it can also depend on the extent of an organized opposition. When an organized opposition engages in partisan advocacy, the government has greater freedom to respond with partisan advocacy of its own. 434 Id. at 852-53 (citations omitted). 435 Id. at 853-54. 118 The plaintiffs proved at trial that Red Clay violated the Elections Clause because of the profound imbalance between Red Clay’s intense engagement with targeted groups and its relatively minimal efforts to engage with the electorate as a whole. Red Clay’s effort to avoid engaging with other identifiable groups, such as the elderly, accentuated the imbalance. This is a fact-specific holding. Red Clay launched its campaign on November 6, 2014, when the core team held its initial meeting with the members of the Steering Committee and the Red Clay principals.436 The campaign stressed focusing on favorable groups while avoiding other voters.437 One of the speakers gave an example of a senior citizen on a fixed income as a “NO” voter with whom there was “no point” in engaging.438 The Steering Committee co-chair wrote to others on the core team that she understood the campaign would “not . . . reach out to [retired] folks as they could bring out the no vote.”439 Red Clay spent November, December, and January priming its favored groups with communications about the referendum. During this period, each school engaged in a telephone campaign to locate its goal number of parents who would vote “YES.” The script 436 See JX 28 (Superintendent Daugherty mandating attendance by principals); Tr. 114 (Baltz principal testifying that she understood the meeting was mandatory for all Red Clay principals); Tr. 489 (Richardson Park principal testifying that he only attended the meeting because it was mandatory). 437 See JX 25 at D2555 (campaign strategy slide listing one of four key planks as “[w]ork on the Yes not the No”). 438 Tr. 675 (Floore). 439 JX 310 at D8610. 119 for the calls was explicitly pro-referendum and instructed callers to ask parents to vote for the tax increase.440 The callers only gave polling information to “YES” voters.441 Red Clay followed up with the “YES” voters “to remind the ‘yes’ voters to vote,” but did “not call back folks that said they are voting ‘no.’”442 Red Clay also targeted parents of Red Clay students with other communications, such as letters, flyers, emails, and materials stuffed inside student report cards. Promotional efforts involved attaching “VOTE” stickers to elementary school children and having district employees and parents pass out “push cards” at the drop-off and pickup lines on the day of the Special Election.443 Red Clay schools also used their automated alert system to remind parents to come to the schools on the day of the Special Election.444 Red Clay focused on parents because those voters were most likely to support a tax increase.445 When Red Clay voluntarily looked beyond parents, it focused on other 440 See JX 311 (“Can we count on you supporting Red Clay schools on February 24th?”). See id. (telling “YES” voters but not others “that polls are open from 10 a.m. to 8 p.m. and you can vote at any Red Clay school”); Tr. 209-11 (Nash). 441 442 JX 313. 443 Tr. 213-15 (Nash); Tr. 490 (Mathis); Tr. 628-31 (Ammann); JX 227; JX 247. 444 Tr. 121-22 (Penoyer). See, e.g., Tr. 497-98 (Richardson Park principal stating, “I believe that if [parents] came out, they would vote yes”); Tr. 351 (Johnson acknowledging, in connection with the events Red Clay held at all polling place schools during the Special Election, that she “wanted parents in the schools on the day of the referendum because [she] wanted them to vote” and she believed that “very few parents would oppose the tax increase”); Tr. 537 (Red Clay expert and former superintendent Andrzejewski agreeing that “the purpose of 445 120 supportive groups, such as parents of future Red Clay students, Red Clay graduates, and college educators who worked with student teachers.446 As part of its one-sided campaign, Red Clay communicated in ways that masked the official origins of its speech. Nash created a Facebook account called “Red Clay Parents for Students” and a Twitter handle called “RedClayParents.”447 She ghostwrote much of the content with no attribution to herself or the district.448 She also offered to ghostwrite letters to the editor.449 The lack of attribution was intentional: Nash felt “it was more real and would resonate more [with] voters if it came from parents . . . .”450 Without accurate disclosure of source and authorship, recipients could not fairly evaluate the interests of the speaker or the content of the communication. Compared to the effort that Red Clay made to reach groups that would favor the tax increase, Red Clay did relatively little to reach out to the electorate as a whole. That does not mean Red Clay did nothing. In August 2014, when the School Board first considered the family friendly events in Red Clay was to get likely yes voters into the polling places”); Tr. 180-81 (Nash agreeing that Red Clay “was targeting or prioritizing positive voters” with certain campaign efforts). The Statement of Facts collects further evidence and makes findings regarding this point. 446 JX 50; JX 104; JX 120. 447 JX 25 at D2558. 448 Tr. 197-99 (Nash); JX 106. 449 JX 103; Tr. 199-201 (Nash). 450 Tr. 194 (Nash). 121 holding the Special Election, it held a workshop that was open to the public.451 In September, a community advisory board discussed the potential referendum during a meeting to that was open to the public, the School Board held a separate meeting to receive input from members of the community, and the School Board again discussed the matter during a portion of its meeting that was open to the public.452 In October, when the School Board formally resolved to hold the Special Election, it did so during a meeting that was open to the public.453 But these meetings did little to publicize the Special Election, and they were qualitatively and quantitatively different than the affirmative outreach to favorable groups that Red Clay began the following month. After launching its campaign in November 2014, Red Clay continued to do relatively little to inform the electorate as a whole. In December, Red Clay sent an edition of the Red Clay Record to all of the residents and businesses in the district.454 Red Clay also identified the fact that the Special Election was taking place on the district website and on the district Facebook page, and it aired some videos on the district’s educational cable channel.455 And Red Clay representatives attended various meetings with school and community groups, although the groups largely comprised constituencies that supported 451 Tr. 562-64 (Ammann). 452 Id. at 564-67. 453 Id. at 568-69. 454 JX 281; Tr. 329 (Johnson). 455 Tr. 225-26 (Nash); Tr. 330-31 (Johnson); Tr. 577, 583-86 (Ammann). 122 Red Clay.456 This level of outreach did not come close to Red Clay’s efforts to contact favorable voters. Equally important, Red Clay’s internal documents reveal that the district tried affirmatively during this period to avoid communicating broadly or with groups that might oppose the tax increase.457 The Red Clay team avoided contact with “random retirees,”458 delayed sending letters to the editor that “might ‘wake up’ the ‘no’ voters,”459 resisted appearing on a radio show that was perceived as having an “anti-vote” audience,460 and stayed away from other communication channels that might trigger a public debate.461 As a result, it was not until February 2015 that wider segments of the community became meaningfully aware of the Special Election. In February, after the Department of 456 See Tr. 590-91 (Ammann) (discussing attendance at a Red Clay school resource fair that drew Red Clay families); id. at 592-94 (discussing meeting with Delaware Decision Makers business group, New Castle County Chamber of Commerce, and state legislators). See JX 25 at D2555 (describing Red Clay’s strategy as “[w]ork on the Yes not the No”); Tr. 203-04 (Nash). 457 458 JX 309; see Tr. 352-54 (Johnson). 459 JX 124; accord Tr. 201 (Nash). Even after the News Journal ran a story about the referendum in early February 2015, the team considered waiting until anti-referendum letters to the editor appeared. See Tr. 201-03 (Nash); JX 103; JX 124. Once they did, the team debated whether “the ‘pro’ letters would wake up even more seniors??” JX 134; Tr. 215-16 (Nash). 460 Tr. 219-20 (Nash); JX 135. 461 See Tr. 201-02 (Nash testifying that Red Clay stayed away from certain forums to avoid alerting “no” voters); Tr. 217-19 (Nash testifying that the Steering Committee decided against having residential yard signs to avoid alerting “no” voters to the Referendum). 123 Elections asked whether Red Clay intended to send out any district-wide communications, Red Clay circulated another edition of the Red Clay Record.462 Unlike the December edition, the February edition generated a response. 463 Red Clay also published the statutorily required notices in the News Journal once a week for the four weeks before the Special Election.464 And the School Board held another meeting that was open to the public at which it discussed the referendum.465 These communications remained qualitatively and quantitatively abbreviated compared to Red Clay’s engagement with favorable groups. Under Brennan, Red Clay violated the Elections Clause by conducting a fully partisan, one-sided campaign. If Red Clay had faced active opposition that was engaging in similar tactics, then its targeted campaign speech might not have violated the Elections Clause. In this case, however, Red Clay spent three months mobilizing its supporters before a public debate began to emerge.466 462 JX 280. 463 JX 132 (Johnson telling the Red Clay team that objections from seniors had been “coming in since the red clay record landed”); see JX 134 (parent informing Nash that “there were 2 [letters to the editor] in today about seniors and school tax”); id. (Nash responding to question about whether “‘pro’ letters to the editor would wake up even more seniors” by stating, “I think they [the seniors] are awake”). 464 Tr. 577-78 (Ammann); JX 271; JX 189, at 2. 465 JX 115; Tr. 587 (Ammann). 466 Once it did, then I believe Red Clay had greater license to respond. For example, there is evidence that just before the Special Election, an opposition group made 5,000 to 6,000 automated calls urging voters to oppose the tax increase. JX 329, at 27. But for other complicating factors in the case, the existence of those calls would have justified Red 124 Although the outcome in this case seems clear under Brennan, I do not believe that the Elections Clause should require that a school district treat all portions of the electorate identically during a referendum. Envision an alternative scenario in which Red Clay had openly identified the source of all of its government campaign speech, taking that issue off the table. If Red Clay had not conducted its telephone campaign to identify individual “YES” voters, and as long as Red Clay sent out a communication like the December edition of the Red Clay Record to all district residents early in the process, then I personally would not view Red Clay’s government campaign speech as constitutionally problematic. That view arguably conflicts with a strict reading of Brennan. For purposes of this decision, the issue is irrelevant, because the combination of the Family-Focused Events and Red Clay’s fully partisan campaign prevented the Special Election from being “free and equal.” III. THE REMEDIAL CALCULUS The plaintiffs have shown that Red Clay violated the Elections Clause. The next issue is the appropriate remedy. One remedy the plaintiffs seek is a permanent injunction prohibiting Red Clay from engaging in similar electoral interventions in the future. Other potential remedies hinge on whether Red Clay’s interventions were sufficiently serious to warrant declaring the Special Election invalid. If so, then a range of potential remedies becomes available. Options include an order requiring the holding of a new special election, a permanent injunction against Red Clay continuing to collect the increased taxes Clay’s use of its School Messenger system and other school resources during the days before the vote. 125 that were approved in the Special Election, a requirement that Red Clay refund the moneys it collected in the incorrect belief that the Special Election was valid, or various combinations of relief. A. An Injunction Against Further Unconstitutional Conduct The plaintiffs seek a permanent injunction barring Red Clay from repeating in any future elections the actions that this court has found unconstitutional. A permanent injunction against future conduct is not warranted simply because a court has found past conduct illegal. “[T]he court must presume that [parties] will respect any decision rendered by any competent court of this State.”467 As Chief Justice Strine explained while serving as a Vice Chancellor, a request for injunctive relief against future violation of a statute “both trivializes equity’s role and implicitly suggests that the most powerful expression of a societal prohibition—an express statute forbidding conduct—is somehow insufficient without an ‘us, too’ from the judicial branch.”468 In this case, the societal prohibition is a guarantee protected by the Declaration of Rights. 467 Christiana Town Ctr., LLC v. New Castle Ctr., 2003 WL 21314499, at *3 (Del. Ch. June 6, 2003); accord Del. Bldg. & Constr. Trades Council v. Univ. of Del., 2014 WL 2218730, at *2 (Del. Ch. May 29, 2014) (holding that injunctive relief was not warranted where a plaintiff “merely contends that, because the Defendants have purportedly not complied with [a] statute in the past, they will continue this alleged pattern of noncompliance” after a court order); Reeder v. Del. Dep’t of Ins., 2006 WL 510067, at *16 (Del. Ch. Feb. 24, 2006) (Strine, V.C.) (“There is no justification on this record for an injunction requiring the Insurance Department to do what it must do in any event—comply with applicable statutory constraints on its behavior.”). 468 State ex rel. Brady v. Pettinaro Enters., 870 A.2d 513, 536-37 (Del. Ch. 2005) (Strine, V.C.). 126 Courts will only enjoin future wrongful conduct where the facts “show[] reason to apprehend a threat of future violation of judicially-determined rights and duties.”469 The plaintiffs assert that Red Clay will repeat the conduct that led to this case whenever it needs to conduct another referendum. I do not share the plaintiffs’ concern. I believe that Red Clay will abide by a final decision rendered by the judiciary. B. Whether To Invalidate The Special Election The other forms of relief that the plaintiffs seek all hinge on whether the Special Election was invalid. The fact that a plaintiff has proven electoral violations, even constitutional violations, does not lead inexorably to invalidating the election. A further remedial calculus is necessary. In this case, a multi-factor analysis counsels against invalidating the Special Election. Courts are understandably reluctant to invalidate elections.470 As the Delaware Superior Court has observed, “It is the duty of the Court, if possible, to sustain elections which have resulted in a full and fair expression of the public will.”471 In Brennan, the Delaware Supreme Court observed that “minor irregularities in the conduct of an election unaccompanied by fraud or unfair dealing, and not affecting the result, will not void an 469 McMahon v. New Castle Assocs., 532 A.2d 601, 606 (Del. Ch. 1987) (Allen, C.). 470 See generally Stephen F. Huefner, Remedying Election Wrongs, 44 Harv. J. on Legis. 265, 283-85 (2007) [hereinafter Election Wrongs]; Kenneth W. Starr, Federal Judicial Invalidation as a Remedy for Irregularities in State Elections, 49 N.Y.U. L. Rev. 1092, 1105-08 (1974) [hereinafter Irregularities in State Elections]. 471 State ex rel. Stabler v. Whittington, 290 A.2d 659, 662 (Del. Super. 1972). 127 election otherwise valid.”472 Decisions from other jurisdictions similarly stress that the judiciary should exercise caution before setting aside an election.473 The parties have not identified a governing standard for this court to apply and have battled over competing formulations from other jurisdictions. Given the state of the law, this comes as little surprise. “Federal courts have struggled to articulate an easily applied test to determine when the court should invalidate a state election and order a special election.”474 State authorities use varying formulations.475 The best that can be said is this: “Rather than utilizing a general rule, each case must be considered individually. There is no all-encompassing list of factors which a court must consider.”476 From my review of election decisions and related scholarly literature, several factors appear salient. These include the nature and scope of the misconduct, the actor’s intent, the 472 104 A.2d at 789. See, e.g., Bortner v. Woodbridge, 736 A.2d 104, 113 (Conn. 1999) (“[A]lthough a court undoubtedly has the power to order a new election . . . , the court should exercise caution and restraint in deciding whether to do so. A proper judicial respect for the electoral process mandates no less.”); Putter v. Montpelier Pub. Sch. Sys., 697 A.2d 354, 357 (Vt. 1997) (“Voiding an election and ordering a new one represents one of the more extreme remedial measures available to a court sitting in equity.”). 473 474 Gjersten v. Bd. of Election Comm’rs for Chicago, 791 F.2d 472, 478 (7th Cir. 1986). See Election Wrongs at 283 (“Courts have resorted to this remedy in a variety of circumstances, and many courts have wrestled with whether and when they have authority to order a new election.”); Note, Developments in the Law: Elections, 88 Harv. L. Rev. 1111, 1302-05 (1975) (examining varying standards that are used) [hereinafter Election Law Developments]. 475 476 Gjersten, 791 F.2d at 479. 128 clarity of the effect on the election, and the consequences for the parties involved.477 In this case, after weighing these factors, I do not believe that Red Clay’s conduct warrants invalidating the Special Election. 1. The Nature Of The Misconduct One factor that courts frequently consider is the nature of the misconduct. Flagrantly unconstitutional actions and blatant violations of election laws are more likely to lead to the invalidation of an election.478 At the other end of the spectrum are cases See, e.g., id. (noting relevant interests include “the integrity of the electoral system[,] . . . the necessities of the process of governing[,] . . . [and whether] the unconstitutional practice had a significant impact on the particular election”); Hendon v. N.C. State Bd. of Elections, 710 F.2d 177, 182 (4th Cir. 1983) (“Whether the irregularity amounts to a constitutional claim depends on its severity, whether it was intentional or more of a negligent failure to carry out properly the state election procedures, and whether it erodes the democratic process.” (citation omitted)); Putter, 697 A.2d at 357 (“[C]ourts have focused on several key considerations, including the nature and severity of the . . . violation, the probability that it actually affected the election result, the presence or absence of culpable intent, and the harm to the organic process of the election.” (citations omitted)). 477 478 Putter, 697 A.2d at 358. The clearest examples are the foundational federal cases that invalidated elections in which state governments excluded voters based on race. See, e.g., Hadnott v. Amos, 394 U.S. 358, 367 (1969) (directing district court “to require the state and local officials promptly to conduct a new election” after finding that state officials struck candidates from ballot on basis of race); Toney v. White, 488 F.2d 310, 316 (5th Cir. 1973) (en banc) (invalidating election where voters were purged on basis of race); Bell v. Southwell, 376 F.2d 659, 664 (5th Cir. 1967) (holding that invalidating local election was warranted “[c]onsidering the gross, spectacular, completely indefensible nature of this state-imposed, state-enforced racial discrimination”). See generally Irregularities in State Elections at 1097 (noting that the federal judiciary’s historic unwillingness to interfere in state elections “was finally interred during the 1960’s with the full advent of reapportionment and civil rights litigation”). A less-extreme Delaware example involved a municipal election where city-appointed election officials refused to serve, at which point a private individual “assumed the office of inspector, appointed two judges of elections, and the election was held.” Hampson v. State ex rel. Buckson, 233 A.2d 155, 158 (Del. 1967). The Delaware Supreme Court explained that it was “immaterial that the election 129 involving “minor irregularities.”479 Although misconduct that violates a constitutional guarantee is typically more problematic than a statutory or regulatory violation, the existence of a constitutional violation does not “ipso facto work any fundamental change in judges’ equitable methodology.”480 Depending upon the circumstances, courts have may have been conducted honestly. If those persons conducting it had no lawful right to do, it is a nullity, and its results must be set aside.” Id. 479 Brennan, 104 A.2d at 789. Most of the Delaware precedents involve this level of misconduct. See, e.g., id. (declining to invalidate school tax referendum in which residents received separate ballots for voting for or against the proposal, rather than single ballot containing both options, as required by then-existing state law); State ex rel. Wahl v. Richards, 64 A.2d 400, 406-07 (Del. 1949) (declining to invalidate General Assembly election in which officials in one district improperly counted ballots that were not signed by both clerks, as required by then-existing state law); Spencer v. Smyrna Bd. of Educ., 547 A.2d 614, 618 (Del. Super. 1988) (declining to invalidate school tax referendum with eighteen-vote discrepancy between vote totals from initial vote count, recount, and polling lists); Whittington, 290 A.2d at 662 (declining to invalidate municipal election where local officials failed to use voting machines, as required by then-existing state law); State ex rel. Green v. Holzmueller, 5 A.2d 251, 256 (Del. Super. 1939) (declining to invalidate municipal election where city officials and Board of Elections permitted voting by proxy, which was not authorized by city’s charter); McComb v. Dutton, 122 A. 81, 83 (Del. Super. 1923) (declining to invalidate school tax referendum in which election officers opened and closed at hours specified in statute but using Daylight Savings Time rather than Standard Time); Kelley v. Mayor & Council of Dover, 300 A.2d 31, 36 (Del. Ch. 1972) (declining to invalidate special election for annexation where city did not have number of election officials required by city charter). 480 Irregularities in State Elections at 1099; see Gjersten, 791 F.2d at 478 (“Although federal courts have the power to invalidate elections held under constitutionally infirm conditions, the courts need not exercise this power in the case of all elections held pursuant to unconstitutional statutes.” (internal citations omitted)); Hamer v. Campbell, 358 F.2d 215, 222 (5th Cir. 1966) (“This action does not mean that we necessarily would set aside every election in which a substantial number of citizens have been denied the right to vote.”); Putter, 697 A.2d at 357 (“Invalidation of an election requires more than merely a claim of election irregularity, even one of constitutional dimensions.” (citing Irregularities in State Elections at 1099)); Abrahams v. Superior Court 131 A.2d 662, 671 130 sometimes allowed an election to stand even though it was tainted by unconstitutional conduct.481 The primary consideration appears again to be the clarity of the violation. Courts have declined to invalidate elections “when the validity of the subsequently invalidated governmental practice or policy [was] subject to ‘rational disagreement.’”482 In one pertinent example, the Supreme Court of Vermont rejected a complaint that sought to invalidate a school referendum in which the school board published and distributed newsletters, leaflets, and stickers in support of a budget and bond proposal.483 Despite acknowledging that the plaintiff’s constitutional claim “finds support in a number of judicial authorities,” the court observed that the theories were “relatively novel, the supporting authorities comparatively sparse, and the rules defining the scope of permissive conduct currently unsettled; the issue, in short, is one about which reasonable minds may easily disagree.”484 The court concluded that ordering a new election would be (Del. 1957) (declining to invalidate municipal election canvassed pursuant to a state law requiring judicial canvassing despite holding that law was unconstitutional). See, e.g., Bowes v. Ind. Sec’y of State, 837 F.3d 813, 821 (7th Cir. 2016); Gjersten, 791 F.2d at 479; Putter, 697 A.2d at 359. 481 482 Irregularities in State Elections at 1101; see, e.g., Allen v. State Bd. of Elections, 393 U.S. 544, 571-72 (1969) (declining to invalidate election after finding that state had erroneously interpreted Section 5 of the Voting Rights Act). 483 Putter, 697 A.2d at 359. 484 Id. at 358. 131 disproportionate absent “any basis for a finding that the disputed materials were outrageously illegal,” among other considerations.485 In this case, Red Clay’s electoral interventions violated the Elections Clause. They were constitutional violations, and hence of serious magnitude, but Red Clay did not cross clearly delineated, brightly marked lines. This court has issued two lengthy decisions analyzing the Family-Focused Events. Although this court has concluded that the FamilyFocused Events prevented the Special Election from being “free and equal,” reasonable minds could disagree. Both decisions relied for support on the Anti-Bribery Clause in the Delaware Constitution and comparable statutory prohibitions, but there was not previously a judicial ruling addressing targeted inducements to vote like the Family-Focused Events. Likewise, although both decisions relied on the Electioneering Statute, the Dismissal Ruling found that statute ambiguous in two respects,486 and this decision relied in significant part on evidence of Delaware custom and practice to apply it to the facts. Under the circumstances, it was not unreasonable to believe that the Family-Focused Events could be constitutionally permissible. The case for Red Clay’s extensive and targeted campaign speech is somewhat harder, because the Delaware Supreme Court held in Brennan that “expenditure of public 485 Id. at 359. In a more recent decision, the Supreme Court of Vermont followed Putter and refused to invalidate an election where a town distributed campaign materials that advocated its positions on various issues. See Daims v. Town of Brattleboro, 148 A.3d 185, 190-91 (2016). 486 See Dismissal Ruling, 122 A.3d at 854-56. 132 funds in support of one side” must remain “within reasonable limits” and that a school district’s speech should not venture “beyond [a] factual presentation” to the point of “overstatement and emotional appeals.”487 But as the Dismissal Ruling explained, more than sixty years have elapsed since Brennan, and a strong argument can be made that its strictures should be relaxed for referendums in which the state has an affirmative position on a particular policy issue.488 Under the circumstances, it was not unreasonable to believe that Red Clay’s campaign speech was constitutionally permissible. In light of these considerations, Red Clay’s electoral misconduct, although unconstitutional in nature, was neither flagrant nor extreme. This factor counsels against invalidating the Special Election. 2. The Scope Of The Misconduct A second factor that courts frequently consider is the scope of the misconduct. When the misconduct is isolated or sporadic, courts are unlikely to invalidate the 487 104 A.2d at 790. 488 See Dismissal Ruling, 122 A.3d at 852-54; cf. Mercier v. Inter-Tel (Del.), Inc., 929 A.2d 786, 808-09 (Del. Ch. 2007) (Strine, V.C.) (“Here’s a news flash: directors are not supposed to be neutral with regard to matters they propose for stockholder action.”); In re MONY Gp., Inc. S’holder Litig., 853 A.2d 661, 675-76 (Del. Ch. 2004) (“[O]nce a board of directors deems a merger agreement favorable, it may employ various legal powers to achieve a favorable outcome on a shareholder vote required to approve that agreement.”). My suggestion in the Dismissal Ruling only addressed referendums, not candidate elections. 133 election.489 When the misconduct is widespread or systematic, courts are more likely to determine that the election was void.490 489 See, e.g., Soules v. Kauaians for Nukolii Campaign Comm., 849 F.2d 1176, 1184 (9th Cir. 1988) (“[The plaintiffs’] claims of lax security and the delivery of 60 absentee ballots by persons other than the voter, do not bring into question the fundamental fairness of the conduct of the special election.”); Pettengill v. Putnam Cty. R-1 Sch. Dist., 472 F.2d 121, 122 (8th Cir. 1973) (dismissing challenge to local school board election that alleged 116 absentee ballots were counted illegally); Johnson v. Hood, 430 F.2d 610, 611, 613 (5th Cir. 1970) (dismissing challenge to local election that alleged “arbitrary and capricious” rejection of ten ballots); Ron Barber for Congress v. Bennett, 2014 WL 6694451, at *7 (D. Ariz. Nov. 27, 2014) (“Here, Plaintiffs’ claims are not basely broadly on the fairness of the terms and procedures of the election; rather they focus on individual and infrequent pollingplace irregularities and verification procedures.”); Rizzo v. Bizzell, 530 So.2d 121, 128 (Miss. 1988) (“When the percentage of illegal votes is small[], even though the winning margin is less than the number of illegal votes, a special election may not be required.”); Anderson v. Ivy, 955 N.E.2d 795, 803 (Ind. Ct. App. 2011) (refusing to invalidate election where candidate’s violations of state anti-electioneering law “occurred in only two out of six precincts” and for only “a few hours in the morning and early afternoon”). 490 See, e.g., Bonas v. Town of N. Smithfield, 265 F.3d 69, 72-73, 75 (1st Cir. 2001) (holding that school committee members elected in 1997 to four-year terms could not delay next school committee election by year, and stating, “If the [city’s] decision is allowed to stand, every resident of North Smithfield will be deprived of his or her right to vote for the affected offices. In our judgment, such across-the-board disenfranchisement betokens an utter breakdown of the electoral process”); Bennett v. Yoshina, 140 F.3d 1218, 1226 (9th Cir. 1998) (“We have drawn a distinction between ‘garden variety’ election irregularities and a pervasive error that undermines the integrity of the vote.”); Marks v. Stinson, 19 F.3d 873, 887 (3d Cir. 1994) (upholding district court order invalidating election tainted by “massive absentee ballot fraud, deception, intimidation, harassment and forgery”); Griffin v. Burns, 570 F.2d 1065, 1077 (1st Cir. 1978) (“There is precedent for federal relief where broad-gauged unfairness permeates an election . . . .”); Krieger v. City of Peoria, 2014 WL 4187500, at *5 (D. Ariz. Aug. 22, 2014) (“This fundamental unfairness is more than isolated. . . . [T]he defective ballots in this case were mailed to approximately one-half of the voters.”); Hoblock v. Albany Cty. Bd. of Elections, 341 F. Supp. 2d 169, 177 (N.D.N.Y. 2004) (granting preliminary injunction prohibiting certification of election results after finding “pervasive and fundamental unfairness in the election”), aff’d in relevant part, 422 F.3d 77 (2d Cir. 2005); Ex parte Vines, 456 So.2d 26, 28 (Ala. 1984) (ordering new election where one of four voting machines in city race failed to register any votes); Miller v. Picacho Elementary Sch. Dist. No. 33, 877 P.2d 277, 279 (Ariz. 1994) (“This is not a case 134 In this case, Red Clay’s electoral interventions were pervasive. Red Clay held seventy-five Family-Focused Events in the twenty-three schools that served as polling places. Using Red Clay’s own low-end estimate, at least 6,383 people attended the FamilyFocused Events, a figure that does not include attendees at twenty-two events where the attendance was listed as “Unknown.”491 That figure is just a dozen votes short of the 6,395 votes in favor of the referendum. A more realistic estimate for the Family-Focused Events is that they brought between 9,023 and 9,233 people to the polling places, well in excess of the number of “YES” votes. Red Clay also engaged in extensive and intensive government campaign speech. The widespread nature of Red Clay’s conduct counsels in favor of invalidating the Special Election. of mere technical violation . . . . District employees with a pecuniary interest in the override’s passage delivered [absentee] ballots to electors whom they knew. . . . These were substantive irregularities.”); Bolden v. Potter, 452 So.2d 564, 567 (Fla. 1984) (“The fraud in this instance was not inconsequential. It was blatant and corrupt and it permeated a substantial part of this absentee-election process.”); Barbour v. Gunn, 890 So.2d 843, 84748 (Miss. 2004) (“In the case sub judice, the problems are not ‘technical;’ an entire subprecinct was not allowed to vote.”); McNally v. Tollander, 302 N.W.2d 440, 445 (Wis. 1981) (“[T]he number of voters who were denied ballots in the present case was very substantial. Some 2,500 voters, approximately forty percent of the electorate, were denied ballots.”). 491 JX 301. 135 3. The Actor’s Intent A third factor that courts frequently consider is the actor’s intent.492 If the actor has acted intentionally, then a court is more likely to invalidate the election.493 If the violation resulted from negligence or inadvertence, then a court is less likely to invalidate the election.494 492 Hendon, 710 F.2d at 182 (noting significance of whether failure to comply with state election law “was intentional or more of a negligent failure”); Bolden, 452 So.2d at 566 (stating one factor courts must consider is “the presence or absence of fraud, gross negligence, or intentional wrongdoing”); McCavitt v. Registrars of Voters, 434 N.E.2d 620, 630 (Mass. 1982) (“We think there is an important distinction between fraud, intentional wrongdoing, and mere failure to follow directions.”); Putter, 697 A.2d at 357 (noting that courts often consider “the presence or absence of culpable intent”). See, e.g., Bonas, 265 F.3d at 75 (“Here, . . . the decision to dispense with an election was deliberate.”); Duncan v. Poythress, 657 F.2d 691, 704 (5th Cir. Unit B 1981) (“We . . . can imagine no claim more deserving of constitutional protection than the allegation that state officials have purposely abrogated the right to vote”); Bolden, 452 So.2d at 567 (holding that “clear fraud and intentional wrongdoing” that “tainted the entire absentee voting procedure” warranted invalidating a local school board election); Pabey v. Pastrick, 816 N.E.2d 1138, 1154 (Ind. 2004) (“The magnitude, pervasiveness, and widespread effect of the deliberate series of actions found in this case leads to but one conclusion. The Pastrick campaign certainly knew or consciously intended that the results of their conduct would . . . profoundly undermine the integrity of the election . . . .”); Tollander, 302 N.W.2d at 444 (“[T]he record shows that the election statutes were intentionally ignored by public officials . . . .”). 493 See, e.g., Gold v. Feinberg, 101 F.3d 796, 802 (2d Cir. 1996) (“[H]uman error in the conduct of elections does not rise to the level of a Fourteenth Amendment constitutional violation actionable under § 1983 in the absence of willful action by state officials intended to deprive individuals of their constitutional right to vote.”); Hendon, 710 F.2d at 182 (“There is no indication that the failure was other than simple negligence on the part of election officials.”); Gamza v. Aguirre, 619 F.2d 449, 454 (5th Cir. 1980) (“There was no allegation of intent to violate their constitutional rights or of deliberate deprivation of their right to vote.”); Hamer v. Ely, 410 F.2d 152, 156 (5th Cir. 1969) (“[T]he attitude of Sunflower’s Election Commissioners may have been shoddy, but it does not justify . . . voiding a state election.”); Putter, 697 A.2d at 359 (“[T]he claimed constitutional 494 136 In this case, Red Clay intended to intervene in the election in the manner that it did. Red Clay developed and implemented a top-down strategy to prevail in the Special Election. Red Clay intended to hold the Family-Focused Events and to use them to bring parents to the polls. Red Clay also intended to reach groups that it believed would favor the referendum through targeted government campaign speech. In these senses, Red Clay’s conduct was intentional. Red Clay did not, however, intend to violate the law. The trial record convinced me that the Red Clay team believed that they were acting lawfully and doing what was both necessary and appropriate for the district to prevail in the Special Election. Primarily through Ammann, Red Clay worked with the Department of Elections and sought to comply with its instructions. More broadly, Red Clay tried to follow what its administrators understood the law to be, even if they pushed the edges of that understanding. This decision has found that the Red Clay administrators did not consciously discriminate against elderly or disabled voters. infringement provides no basis for a finding that the Board willfully and knowingly violated plaintiff’s constitutional rights.”); Adair Cty. Bd. of Elections v. Arnold, 2015 WL 5308132 (Ky. Ct. App. Sept. 11, 2015) (refusing to invalidate municipal election despite “incompetence and incomprehensible carelessness” of local officials). Of course, other factors may result in relief being granted notwithstanding a lack of intent. See Hoblock v. Albany Cty. Bd. of Elections, 422 F.3d 77, 98 (2d Cir. 2005) (holding that issuance of preliminary injunction against certifying election result was warranted “when election officials refuse to tally absentee ballots that they have deliberately (even if mistakenly) sent to voters”); Krieger, 2014 WL 4187500, at *4 (“wrongful intent is not required” where there is fundamental unfairness in the election process). 137 In short, the Red Clay team acted in good faith. At worst, they were mistaken about the scope of the Elections Clause. That description could apply to my decisions as well, because only the Delaware Supreme Court can issue an authoritative interpretation of that provision. Red Clay’s effort to comply with the law, at least as they understood it, supports not invalidating the Special Election. 4. The Clarity Of The Effect On The Election A fourth factor that courts examine is the extent to which the misconduct tainted the result. Demonstrating the interrelatedness of the factors, the scope and seriousness of the misconduct often affect how a court takes this consideration into account. When parties argue about specific ballot irregularities or misconduct at particular polling locations, then courts frequently examine whether the plaintiff can show that the outcome of the election would have been different. If it is clear that the wrongdoing could not have affected the outcome, then that fact is dispositive.495 For example, the Delaware See, e.g., Bennett, 2014 WL 6694451, at *8 (“Even if all 133 votes are counted, it is undisputed that Martha McSally wins the election because she leads by a margin of 161 votes at this time.”); Abbott v. Hunhoff, 491 N.W.2d 450, 452 (S.D. 1992) (“[I]f it is possible through the exercise of due diligence to show for whom the illegal votes were cast, [the plaintiff] must show that but for the illegal votes he would have prevailed.” (citation and internal quotation marks omitted)); Taylor v. Armentrout, 632 S.W.2d 107, 118 (Tenn. 1991) (“The existence of two uncertain votes, when the referendum passed by six votes, does not justify voiding the entire election.”); Files v. Hill, 594 S.W.2d 836, 839-40 (Ark. 1980) (“The difference amounted to 1,731. Assuming, but not deciding, that the votes of the 1,522 persons named who said they offered to vote for Files but were unable to do so were counted as legal votes, Munson would still have 209 more votes . . . .”); Baggett v. State Election Bd., 501 P.2d 817, 824 (Okla. 1972) (“If competent evidence can be introduced establishing that in spite of the illegal ballots cast, it may be determined with mathematical certainty which candidate received the majority of the legal votes cast, the 495 138 Superior Court refused to invalidate an election based on minor discrepancies between the vote tally and the number of certified ballots because “even if the plaintiffs are given the benefit of the doubt . . . the tax referendum still would have passed.”496 Conversely, courts will usually invalidate an election when there is hard proof that the outcome would have been different but for the conduct in question.497 When the misconduct is more widespread and more serious, courts have not required mathematical certainty that the result would have been different. In those cases, it may be sufficient to show that the misconduct made it impossible to determine how the election otherwise would have turned out. For example, the Delaware Supreme Court has observed that “[w]hen illegal ballots have been voted in an election district in such numbers as to affect the result, or at least to make it uncertain, . . . there are cases where justice State Election Board should issue its certificate of election.”). See generally Election Wrongs at 280. 496 Spencer, 547 A.2d at 618. 497 See, e.g., Hadnott, 394 U.S. at 367 (ordering new election in county where voting totals indicated that candidates would have prevailed if they had not been struck from the ballot); Perkins v. Matthews, 336 F. Supp. 6, 11 (S.D. Miss. 1971) (“In [a particular] Ward[,] Glynn L. Cook, the winner at large, received 343 votes while his opponent, Sam Young, received 747 votes. Had the election been held in the Ward only, as in 1965, Sam Young would presumably have been elected. . . . [T]he election of Mr. Cook, from the City at large, must be set aside.”); Qualkinbush v. Skubisz, 826 N.E.2d 1181, 1207 (Ill. App. Ct. 2004) (affirming trial court and invalidating election after the deduction of thirty-nine fraudulent absentee ballots changed the outcome). See generally Election Law Developments at 1317 (“All jurisdictions will overturn the results of an election and install a new winner when it can be demonstrated that but for the violation, the announced winner would have lost the election.”). 139 requires that the entire vote of that election district be rejected in making the count.”498 The Delaware Superior Court has similarly stated that invalidating an election is warranted “where there is such uncertainty arising from the reception of supposedly illegal votes as to make it impossible to ascertain the true expression of the opinion of the voters . . . . [I]mpossibility is the test.”499 Put differently, mathematical proof that electoral misconduct changed or could not have changed the result is sufficient for a court to issue a ruling, but it is not necessary: “While the ‘outcome’ test provides a sensible guideline for determining when federal judicial invalidation of an election might be warranted, it is not a principle requiring mathematical certainty.”500 If the misconduct has been widespread, serious, and intentional, then courts have been willing to invalidate elections if the misconduct rendered the outcome uncertain.501 In those situations, definitive proof is likely impossible to 498 Richards, 64 A.2d at 406 (emphasis added). 499 Holzmueller, 5 A.2d at 255 (internal quotation marks omitted). 500 Griffin, 570 F.2d at 1080 (citations omitted); accord Bolden, 452 So.2d at 567 (“Once substantial fraud or corruption has been established to the extent that it permeated the election process, it is unnecessary to demonstrate with mathematical certainty that the number of fraudulently cast ballots actually affected the outcome of the election.”). See, e.g., Bowes, 837 F.3d at 819 (holding that “if plaintiffs show a ‘reasonable possibility’ that [and unconstitutional law] affected the outcome of an election, that may demonstrate a ‘significant impact’” such that a court may invalidate an election (citations omitted)); Stinson, 19 F.3d at 889 (“If the district court finds a constitutional violation, it will have authority to order a special election, whether or not it is able to determine what the results would have been in the absence of that violation.”); Henderson v. Graddick, 641 F. Supp. 1192, 1204 (M.D. Ala. 1986) (“Because the right to vote is so important, the possibility that the results of an election were changed as a result of illegal votes is enough 501 140 achieve.502 Thus, “where there is substantial wrongdoing in an election, the effects of which are not capable of quantification but which render the apparent result an unreliable indicium of the will of the electorate, courts have frequently declined” to uphold the to justify ordering a new election.”); Coal. for Educ. in Dist. One v. Bd. of Elections, 370 F. Supp. 42, 57, 58 (S.D.N.Y. 1974) (invalidating school board election; concluding that a “substantial violation of constitutional rights” by local officials “could very well have modified the outcome of the election”), aff’d, 495 F.2d 1090 (2d Cir. 1974); Whitley v. Cranford, 119 S.W.3d 28, 35 (Ark. 2003) (“[A]n election may be voided when the outcome is uncertain.”); Arras v. Reg’l Sch. Dist. No. 14, 125 A.3d 172, 184 (Conn. 2015) (an election may be invalidated where “there were substantial violations” of election law and “as a result of those violations, the reliability of the result of the election is seriously in doubt”); Jackson v. Maley, 806 P.2d 610, 620 (Okla. 1991) (explaining that an election may be invalidated where irregularities are “of such a character in either quality or quantity to prove the outcome of an election cannot be determined”); Buonanno v. DiStefano, 430 A.2d 765, 770 (R.I. 1981) (“[T]he contestant [must] show that the irregularities were sufficiently large in number to establish the probability that the result would be changed . . . .”); Armentrout, 632 S.W.2d at 119 (an election may be invalidated if the outcome is “rendered incurably uncertain”). See, e.g., Cranford, 119 S.W.3d at 35 (“[W]here the wrongs are so serious that they render the election results uncertain or doubtful, there is no way for the trial court to determine who won and who lost the election.”); Pabey, 816 N.E.2d at 1151 (“When as here an election is characterized by a widespread and pervasive pattern of deliberate conduct calculated to cast unlawful and deceptive ballots, the election results are inherently deceptive and unreliable.”); Election Wrongs at 281 (“[W]hen the tainted votes cannot be specifically identified, as is often the case, the proper remedy is less clear.”). 502 141 election.503 As one court observed, “an altered outcome should be found readily when there is a serious violation and close election.”504 Stinson, 19 F.3d at 887; accord Buonanno, 430 A.2d at 771 (“A new election is a remedy frequently utilized by courts when a cloud of doubt encircles the original election results.”); see, e.g., Griffin, 570 F.2d at 1080 (upholding district court decision invalidating local election; concluding that district court “could infer that it was more likely than not that a very significant proportion of those voting by absentee ballot would have gone to the polls had such ballots not been available”); Southwell, 376 F.2d at 662 (“[W]e do not think the Court could justify denial of effective, present relief because of any assumed inability to demonstrate that the outcome would have been different.”); Pabey, 816 N.E.2d at 1151 (ordering new election after pervasive absentee voter fraud made it “impossible to determine the candidate who received the highest number of legal votes”). 503 504 Griffin, 570 F.2d at 1080 (citation and internal quotation marks omitted). 142 Based on the trial record that the parties created, this factor is difficult to assess. In my view, this case cried out for serious statistical analysis,505 informed by relevant social science literature.506 Obvious issues to consider would have included: 505 Scholars have argued that courts should use statistical models when deciding election challenges. See Election Law Developments at 1324-28 (endorsing the use of statistical models by courts where possible to “facilitate more rational and consistent decisionmaking”); Michael O. Finkelstein & Herbert E. Robbins, Mathematical Probability in Election Challenges, 73 Colum. L. Rev. 241, 241-42 (1973) (criticizing courts for subjective approach to probability in election challenges and proposing statistical model); see also Bernard Harris, Election Recounting, 42 Am. Stat. 66, 67 (1988) (constructing model and applying it to hypothetical election); James E. Ward III, The Probability of Election Reversal, 54 Mathematics Mag. 256, 257-59 (1981) (constructing model and applying it to challenged municipal election in Brunswick, Maine); Tom Downs et al., Probability in a Contested Election, 32 Am. Stat. 122, 123-24 (1978) (constructing model and applying it to challenged municipal election in Flint, Michigan). Statistical analysis is not a cure-all, but it can be particularly helpful when assessing the relative likelihood of events, where the simple heuristics that the human mind deploys when making day-to-day decisions often lead to error. See generally Daniel Kahneman, Thinking Fast and Slow (2011). Although judges deciding election contests generally rely on their own subjective assessments of probability, at least one court has considered statistical evidence when determining whether to invalidate an election. See Green v. Reyes, 836 S.W.2d 203, 206 (Tex. App. 1992) (discussing statistical evidence but declining to find it dispositive in light of conflicting expert testimony). 506 There appears to be an extensive body of social science literature that could have informed an expert’s analysis of issues pertinent to this case. See, e.g., Randall Reback, Buying Their Votes? A Study of Local Tax-Price Discrimination, 53 Econ. Inquiry 1451 (2014) (finding presence of an aging population correlates with decreased school revenue, unless elderly homeowners receive state-financed reductions in their local tax prices); Huan Gong & Cynthia L. Rogers, Does Voter Turnout Influence School Bond Elections?, 81 S. Econ. J. 241 (2014) [hereinafter School Bond Elections] (concluding that targeted voter mobilization strategies have potential to influence school bond election outcomes); Sarah F. Anzia, Election Timing and the Electoral Influence of Interest Groups, 73 J. Pol. 412 (2011) (using school district elections to analyze whether interest groups with large stake in electoral outcome can influence low-turnout elections); Craig S. Maher & Mark Skidmore, Voter Response to Referenda Seeking to Exceed Revenue Limits, 29 Pub. Budgeting & Fin. 71 (2009) (analyzing factors that contribute to passage of school bond referendums); Marc Meredith, The Strategic Timing of Direct Democracy, 21 Econ. & Pol. 143 The base rates at which parents and seniors have participated historically in referendums and other sufficiently comparable elections. The rates at which parents and seniors participated in the Special Election. Tests of statistical significance to evaluate whether any difference in rates of participation were inconsistent with random chance. The base rates at which parents and seniors support school referendums. The rates at which parents and seniors supported the tax increase in the Special Election. Because there were (i) twenty-five different polling places, (ii) different numbers of parents and seniors who voted at each polling place, and (iii) different levels of support for the tax increase at each polling place, it would have been a simple matter for a properly trained expert to prepare scatter plots and conduct regressions to identify any statistically significant correlations. With this type of information, an expert could have offered informed opinions about whether and to what degree Red Clay’s electoral interventions affected the outcome of the Special Election, either by changing the result or rendering the unaffected outcome sufficiently uncertain. But no one did any of this. 159 (2009) (examining how scheduling of referendum affects voter turnout and likelihood of passage); Ronald G. Ehrenberg et al., Why Do School Budget Referenda Fail?, 26 Educ. Evaluation & Pol’y Analysis 111 (2004) (analyzing factors that contribute to passage versus failure of school bond referendums); Endogenizing the Median Voter (examining how school districts schedule referendums to shape electorate and increase chances of referendum passing); see also Henry E. Brady & John E. McNulty, Turning Out to Vote: The Costs of Finding and Getting to the Polling Place, 105 Am. Pol. Sci. Rev. 115 (2011) (examining how creating low-intensity impediment to voting by changing locations of polling places influences voter turnout by increasing cost of voting for subsets of electorate). Neither side cited any of this literature, much less presented an expert who could explain its implications. This decision therefore merely notes its existence. I have not attempted to take it into account in rendering this decision. 144 The factual record at trial established that the Family-Focused Events and Red Clay’s targeted campaign speech brought parents to the polls in large numbers. The community witnesses testified that they saw polling places that were filled with students and their families. By Red Clay’s own estimate, at least 6,383 people attended the FamilyFocused Events; more realistically there were more than 9,128 attendees. 507 The low-end attendance figure is just a dozen votes short of the 6,395 total votes recorded in favor of the referendum and the more realistic figure substantially exceeds it. With 5,515 votes against, the winning margin was only 880 votes, or approximately 7% of the residents who voted. With such a narrow margin of victory, one possible inference is that the Family- 507 JX 301. 145 Focused Events swung the result.508 But that is only an intuition.509 Many of the parents who attended the Family-Focused Events likely would have voted anyway. Each community witness testified that the voting lines were short, suggesting that the FamilyFocused Events may not have substantially increased the number of votes. Without more 508 See, e.g., School Bond Elections at 247 (concluding that targeted voter mobilization strategies have potential to influence school bond election outcomes); Samuel Issacharoff, Collateral Damage: The Endangered Center in American Politics, 46 Wm. & Mary L. Rev. 415, 427 (2004) (“Particularly in low profile elections, or whenever turnout is low, [a strategy targeting a partisan base] is especially effective since fewer votes will be necessary in the low turnout race to produce a victorious election.” (citation omitted)); James Adams & Samuel Merrill III, Voter Turnout and Candidate Strategies in American Elections, 65 J. Pol. 161, 170 (2003) (noting that when turnout is low, “each candidate is motivated to appeal to his own partisan constituency since he can affect his supporters’ turnout decisions but not the decisions of the rival party’s supporters”); Samuel Issacharoff, Private Parties with Public Purposes: Political Parties, Associational Freedoms, and Partisan Competition, 101 Colum. L. Rev. 274, 307 (2001) (noting that in an election with low turnout, “activist-fueled get-out-the-vote drives may prove as effective in pulling out a close election as concerted appeals to the center”); see also Michael Peress, Securing the Base: Electoral Competition Under Variable Turnout, 148 Pub. Choice 87, 103 (2011) (observing that while targeting swing voters is the optimal strategy in large elections, this rule may not hold in local elections where “there is more room for variable turnout to have a substantial effect”); Thomas M. Holbrook & Scott D. McClurg, The Mobilization of Core Supporters: Campaigns, Turnout, and Electoral Composition in United States Presidential Elections, 49 Am. J. Pol. Sci. 689, 691 (2005) (“[I]ndependents need to be persuaded and mobilized, while partisans mainly need to be mobilized. Therefore, campaigns have strategic incentives to target their mobilization efforts on partisans out of fear that a core will stay home without the mobilization effort and that a broader canvass would bring the wrong voters to the polls.”). There is a suggestion in the record that it was Floore’s intuition as well. At trial, she expressed concern about whether Red Clay could prevail in a referendum if it could not hold Family-Focused Events or engage in targeted communications with Red Clay parents, but she stopped short of agreeing that they were dispositive. See Tr. 719-22. 509 146 meaningful analysis, it is difficult to assert with confidence that Red Clay’s actions made the difference. One question is whether Red Clay’s interventions actually affected turnout. The plaintiffs tried to provide some numerical analysis of this issue by comparing the rates at which parents voted in the Special Election with the rates at which other groups voted. Using data that Red Clay eventually produced,510 the plaintiffs calculated that parents of Red Clay students voted at between twice and 3.8 times the rate of other registered voters. The plaintiffs reasoned as follows: There were 93,905 active registered voters in Red Clay on the date of the Special Election.511 There were 19,793 parents of Red Clay students on the date of the Special Election.512 510 During discovery, Red Clay initially failed to produce the voter information that the plaintiffs requested. Red Clay only produced partial “call lists” from four different schools, which identified the parents of the children on the lists. JX 33; JX 34; JX 63; JX 111. Of the 2,660 parents, 757 (or 28.5%) were registered voters who voted in the referendum. JX 279. After the close of discovery, Red Clay moved to exclude the plaintiffs’ expert witnesses on the theory that they had not relied on meaningful data. In response, the plaintiffs pointed to Red Clay’s failure to produce it, and the court observed that under settled principles of Delaware law, Red Clay’s failure to produce evidence would permit the court to draw a negative inference about what the evidence would show. See Kahn v. Lynch Commc’n Sys., Inc., 638 A.2d 1110, 1119 n.7 (Del. 1994) (“[T]he production of weak evidence when strong is, or should have been, available can lead only to the conclusion that the strong would have been adverse.”); accord Smith v. Van Gorkom, 488 A.2d 858, 878 (Del. 1985); Chesapeake Corp. v. Shore, 771 A.2d 293, 301 n.7 (Del. Ch. 2000) (Strine, V.C.). Afterwards, Red Clay produced additional information about the electorate. See JX 305; JX 306. 511 JX 207; Dkt. 144. 512 JX 305; Dkt. 144. 147 There were 3,985 parents who voted in the Special Election.513 Of the parents who voted, 3,677 were registered voters, or 92.3%.514 Assuming that the same percentage held true for all parents in the Red Clay school district, then there were 18,269 active registered voters among those parents (92.3% of 19,793). Consequently, 20.13% of the registered parents voted in the Special Election. (3,677/18,269). A total of 11,300 registered voters voted in the Special Election. Because 3,677 of the voters were parents, there were 7,623 voters who were not. A total of 75,636 registered voters were not parents of Red Clay students (93,90518,269). Consequently, 10.08% of the registered voters who were not parents or guardians of Red Clay students voted in the Special Election (7,623/75,636). Registered parents voted at approximately twice the rate of other registered voters.515 These calculations were based on Red Clay’s identification of registered voters, which relied on identical matches between the names on the registered voter list and the list of Red Clay parents. The plaintiffs showed that Red Clay’s insistence on exact matches underreported the number of Red Clay parents who voted by excluding some voters who 513 JX 305; JX 306. 514 JX 305 (3,677/3,985 = 92.3%). 515 The calculations are based on registered voters—although non-registered voters may vote in school referenda—because the total number of unregistered voters in Red Clay is unknown. The difference is likely immaterial: All but 610 of the 11,910 voters in the referendum were registered. JX 161. 148 were obviously the same person. Based on the plaintiffs’ independent review of the call list data, they calculated that parents voted at 3.8 times the rate of other voters.516 The plaintiffs’ calculations indicate that parents voted at a higher rate in the Special Election than other eligible voters, but that by itself is not dispositive. Perhaps parents always vote in referendums at higher rates than other voters. The other key piece of data is the base rate at which parents vote absent interventions. There is also always random variation in any sample, so another key input would be a test of statistical significance to determine whether or not any difference in the Special Election was the result of random chance. No one provided any of this. The plaintiffs did show that Red Clay’s electoral interventions suppressed turnout by seniors.517 Ironically, the plaintiffs established this point through the testimony of Professor Ed Ratledge, one of the defendant’s experts. Ratledge teaches public policy at See JX 279. Based on Red Clay’s initial call lists, produced at JX 33, JX 34, JX 63, and JX 111, the plaintiffs calculated that 28.5% of registered parents voted. Applying the plaintiffs’ percentage to the total number of parents in the district (19,793) indicates that 5,641 registered parents voted. Subtracting this number from the 11,300 registered voters who voted yields 5,659 non-parent and guardian registered voters. The resulting voting rate for non-parent and guardian registered voters is 7.5% (5,659/75,636). That rate is approximately one fourth of the 28.5% voting rate for parents (28.5%/7.5% = 3.8). 516 “Suppression” is a one-sided term. The percentage of the senior vote in the electorate depended on the number of non-senior voters, so it could have been affected if other voters participated at higher levels. If parents turned out in greater numbers than usual, that necessarily would lower the percentage of seniors, even if the rate at which seniors participated remained constant. No one attempted to untangle these effects. Because the testimony on this issue was framed in terms of “suppression,” I have used that terminology. 517 149 the University of Delaware and directs the University’s Center for Applied Demography and Survey Research. He initially opined that the Family-Focused Events did not suppress the elderly vote. His opinion, however, relied on comparing the percentage of seniors who voted in the Special Election (25.9%) with the percentage of seniors who voted in the 2012 presidential election (26.5%). He conceded when examined that the 2014 midterm election was a better comparable.518 In the 2014 midterm election, approximately 33% of the voters were 65 years old or older.519 Ratledge agreed that the 7% gap between the percentage of seniors who participated in the Special Election and the percentage of seniors who participated in the 2014 midterm election supported an inference of suppression. 520 The plaintiffs introduced data on senior turnout in other elections that could support a similar inference. In the 2010 midterm election, approximately 32% of the voters were 65 years 518 Tr. 260-61, 265-66. Ratledge agreed that the 2014 mid-term election was more comparable to the Special Election than the 2012 presidential election, both because it was closer in time and because the mid-term elections receive less publicity and have less widespread turnout than presidential elections. See id. at 260-66. He offered no plausible explanation for initially ignoring the mid-term data, saying that he simply did not recall why he chose to use the less comparable 2012 presidential election. See id. at 261 (Ratledge: “I just don’t recall.”). 519 Id. at 267, 315-16. 520 Id. at 279-80. 150 old or older.521 In recent Red Clay school board elections, the percentage of older voters has ranged from 29.73% to 31.65%.522 Ratledge’s initial opinion was based on an unreliable comparison that seemed cherry-picked to support the outcome his client wanted him to reach. His opinions based on the more comparable 2014 midterm election supported the plaintiffs’ case. Here again, however, the plaintiffs did not take the further steps that would be necessary to show that the degree of suppression likely affected the outcome of the Special Election or called the result into question. What was missing was some evidence about the rates at which seniors support referendums generally or supported the Red Clay tax increase. With that information, the plaintiffs might have been able to show that suppressing the senior vote by 7% made a statistically significant difference. Without it, they have only one input for a multi-variable equation. Implicitly recognizing the gap in the evidence, the plaintiffs argued based on anecdotal descriptions of the voting patterns that parents of Red Clay students supported the tax proposal to a greater extent than seniors. They noted the following: At all thirteen polling places where 37% or more of the voters were parents, the tax increase was approved. 521 Id. at 315-16. 522 Data from Red Clay Board of Education elections shows the percentage of voters 65 or older in those elections to have been 31.63% in 2012, 31.65% in 2013, 30.42% in 2014, and 29.73% in 2015. See JX 6; JX 8; JX 13; JX 187; Dkt. 155, at 25 n.11. 151 In the twelve polling places where fewer than 37% of the voters were parents, the tax increase was approved at only three locations and defeated at nine. In the six polling places where 30% or more of the voters were seniors, the tax increase was rejected. In the nineteen polling places where seniors comprised less than 30% of the electorate, the tax increase was approved at sixteen and defeated at three.523 These observations are interesting, but I have no idea whether these figures are consistent with historical trends or fall within the variation produced by random chance. No one provided a cross-tabulation that compared the percentages of parents or seniors at a polling location with the voting results. No one presented any other tests of statistical significance. No one tried to use the polling location data to show how the number of parents or seniors in the voting population correlated with the number of “YES” votes. No one explained why drawing the line at 37% electoral participation by parents is meaningful versus 30% for seniors. In lieu of statistical analysis informed by social science research, the plaintiffs tendered two veteran campaigners: Senator Peterson and Representative Hudson. Both were qualified by virtue of experience to opine on election issues. Both opined that Red Clay sought to use the Family-Focused Events and targeted campaign speech to intervene in the election. Both opined that Red Clay’s electoral interventions had the desired effect and determined the outcome of the Special Election. 523 JX 194; JX 305; Tr. 460 (Agne). 152 Senator Peterson is now retired. At the time of the Special Election, she represented the 9th State Senate District, which includes portions of Red Clay. She served as a state senator for fourteen years. Before that, she served as President of the New Castle County Council. She has been elected to public office six times and has helped manage eight other campaigns. She resides in Red Clay and voted in the Special Election.524 Senator Peterson explained that in any election, “you want to get the people out to vote who will support the outcome you want,” while trying to avoid having the people who oppose you turn out.525 She testified that someone seeking to prevail in the Special Election “would do everything in their power to get as many of the parents and guardians to the polls as possible, . . . because their children would be the beneficiaries of the outcome.”526 She opined that the Family-Focused Events served that purpose by drawing parents to the polls.527 She also opined that the parking problems disproportionately affected older voters, because parents of Red Clay students would have less difficulty physically accessing the polls.528 She concluded that “the outcome [of Red Clay’s efforts] was a disproportional 524 Peterson Dep. at 3-7. 525 Id. at 27-28. 526 Id. at 29-30. 527 Id. at 24. 528 Id. at 30-33. 153 number of likely supporters of the referendum were able to vote as compared with those who would not be as likely to support the referendum.”529 Representative Hudson currently represents the 12th State House District, which includes much of Red Clay. She has served as a state legislator for twenty-two years, having been elected to that position eleven times. Before serving as a representative, she served as the New Castle County Prothonotary, then an elected position, for five years. She lives in Red Clay and voted in the Special Election.530 Representative Hudson testified that successful campaigns “put the most effort” into “people that are open-minded and may consider you, if you appeal to them.” 531 She explained that at the state and local level, “it’s better just to ignore” individuals who are likely to vote against: “You would hope that they weren’t paying attention to even know there was an election. . . . [I]t’s better to just ignore that group and hope that maybe they won’t even come out to vote.”532 Representative Hudson testified that in her experience, “parents of the children do tend to support referendums,” while the elderly are “the most vocal group in [the] district . . . [and] tend to not want an increase in their school taxes.”533 She opined that by holding the Family-Focused Events, Red Clay drew parents with 529 Id. at 40-41. 530 Tr. 25-26 (Hudson). 531 Id. at 42-43. 532 Id. at 43. 533 Id. at 44, 45. 154 children to the polls and “increase[d] the likely event that more people will vote positive.”534 Although their opinions were informed by their considerable experience, Senator Peterson and Representative Hudson ultimately offered intuitive assessments about the outcome of the election. They did not support their intuitions with data or step-by-step analyses. They did not attempt to estimate the magnitude of the effect. Their testimony amounted to a version of “I know it when I see it.” To rebut Senator Peterson and Representative Hudson’s testimony, Red Clay retained Karl Agne, the founding partner of a strategic consulting firm who has extensive experience in campaign polling, including referendums. Agne’s central point was that we cannot know, without more detailed analysis, what happened in the Special Election.535 Agne repeatedly pointed out the obvious fact that Senator Peterson and Representative Hudson had not supported their opinions with data.536 He also described instances in the 534 Id. at 66; accord JX 210, at 2. The plaintiffs spin Agne’s agnosticism in their favor by reframing it as a failure to contest Senator Peterson’s or Representative Hudson’s opinions, but Agne’s point was broader: No one can know what took place without more detailed analysis. See, e.g., Tr. 449-50 (Agne stating that he had “no idea how seniors might have voted in [the referendum] one way or another”); see also id. at 447 (Agne agreeing that he “basically had no idea whether or not” the referendum passed because of Red Clay’s “get-out-thevote activities and their effect on voters and potential voters”); id. (Agne agreeing that did not have an opinion “on whether or not parents and guardians were more likely to support the tax increase than other voters”). 535 536 Id. at 413-14. 155 voting results that appeared inconsistent with their opinions, but those anecdotal examples were neither more nor less persuasive than what the plaintiffs cited.537 Because of his technical expertise, Agne could have been the most helpful witness in the case. He had the ability to examine the social science literature and perform a sophisticated, scientific, and data-driven analysis of the Special Election. He could have determined whether Red Clay’s interventions had a statistically significant effect and untangled the various influences on voting, including the extent of any correlation between independent variables like parent/guardian status and voter age and dependent variables like turnout and the number of “YES” votes. If he had conducted such an analysis, it likely would have been dispositive. Both Agne and Ratledge had the ability to prepare scatterplots and regressions that would have provided more meaningful information about the voting patterns. In fact, the plaintiffs attempted at trial to introduce simple scatterplots and regressions through Agne, but those analyses came too late. The fact that Red Clay did not have their qualified and capable experts conduct these obvious types of analyses makes me think that they might have supported the plaintiffs’ position. But it was ultimately the plaintiffs’ burden of proof. My intuition makes me 537 For example, Agne noted that at the two polling locations that were not schools, and hence did not hold Family-Focused Events, voters passed the referendum. He also noted that at three schools where elderly voters constituted roughly the same percentage of all voters, the spread in support for the referendum ranged from 31 points in favor to 11 points opposed. Tr. 418-20. These are interesting anecdotal observations, but without more meaningful analysis, it is impossible to know whether they are artifacts of random chance, part of a larger and statistically significant pattern, or legitimately disconfirmatory. 156 suspect that Red Clay’s interventions affected the outcome of the Special Election, but intuition and suspicion are not substitutes for evidence and analysis. Because of the gap in the evidentiary record, I am not able to reach any conclusions about Red Clay’s actions and the outcome of the Special Election. This factor counsels against invalidating the vote. 5. The Balancing of the Hardships Because invalidating an election is an equitable remedy, a balancing of the hardships is always relevant. “In every instance in which a court of equity is asked to issue an equitable remedy, it must concern itself with the effects upon others of its action.”538 “[H]owever inartfully, equity must try to right the wrongs with adequate remedies that destroy no party in the process.”539 A court may conclude under the circumstances that it is “more important for a contested election to be resolved conclusively than that it be resolved perfectly.”540 In this case, invalidating the Special Election would lead to potentially catastrophic consequences for Red Clay. That determination would mean that Red Clay has received approximately $26.3 million in taxes since July 1, 2015, that it was not entitled to get. Red 538 Merrill Lynch, Pierce, Fenner & Smith, Inc. v. Price, 1989 WL 108412, at *2 (Del. Ch. Sept. 13, 1989) (Allen, C.). 539 Cantor Fitzgerald, L.P. v. Cantor, 2000 WL 307370, at *29 (Del. Ch. Mar. 13, 2000), overruled on other grounds, Scion Breckenridge Managing Member, LLC v. ASB Allegiance Real Estate Fund, 68 A.3d 665, 686 (Del. 2013). 540 Election Wrongs at 293. 157 Clay’s CFO testified credibly at trial that there is no possible way Red Clay could reimburse taxpayers for that amount.541 She believed that even if Red Clay were given twenty years to pay it back, Red Clay would not be able to shoulder the burden.542 To avoid this problem, the plaintiffs have represented that they do not seek in this action to force Red Clay to disgorge or otherwise pay back the taxes it has collected, nor do they seek to recover damages in that amount.543 But that representation is not binding on other Red Clay taxpayers. A different group of plaintiffs could file suit and seek a classwide damages award, relying on principles of collateral estoppel from this decision. There is also another route that individual taxpayers could invoke. By statute, “[l]ocal county school taxes paid through error or by mistake may be refunded by the school district to which the taxes were paid . . . .”544 Red Clay has been collecting taxes since July 1, 2015, in the erroneous belief that the increases were validly approved by voters in the Special Election. Individual taxpayers could demand that Red Clay refund their money and sue Red Clay if it did not comply. At present, Red Clay has no procedures for handling or defending those types of requests. Red Clay has no sense of how many taxpayers would seek to recover the overpayments and has not established any reserves to 541 Tr. 688-93 (Floore describing consequences to Red Clay of having to repay funds). 542 Id. at 688, 693. 543 See Dkt. 155 at 55; Tr. 697-700 (plaintiffs’ counsel confirming representation). 544 14 Del. C. § 1921. 158 cover the potential claims.545 Invalidating the Special Election would create a host of unknown complications if taxpayers attempted to invoke their statutory right. Invalidating the Special Election also would affect Red Clay’s ability to collect taxes going forward. On July 1, 2017, Red Clay is scheduled to begin receiving the $0.35 tax increase. If the Special Election is invalid, then Red Clay should not be able to collect those amounts. But Red Clay has already budgeted for the year on the assumption that it would have those funds as operating income, and it has hired teachers and other personnel in reliance on those moneys. If Red Clay could not collect those amounts, it would start the year with an $18.5 million budget deficit.546 Faced with these consequences, Red Clay’s only choice would be to hold another special election and ask its residents to ratify the tax increase.547 To my mind, effectively forcing Red Clay to hold another election would be unfair, because if the system for financing public schools operated as it should, then Red Clay would not have been forced to hold to hold a referendum in the first place. As this decision has explained at length, Delaware school districts only have to hold regular referendums because no one is currently updating the property tax assessments.548 The Delaware Code requires that “[a]ll property 545 See Tr. 681-86 (Floore). 546 Id. at 687-88. 547 Id. at 693-94. 548 See Part I.A, supra. 159 subject to assessment shall be assessed at its true value in money,”549 and the Delaware Supreme Court has held that this concept is the same as “fair market value.”550 But in New Castle County, property values remain pegged to valuations from 1983. 551 If those valuations were kept current, as the Delaware Code appears to contemplate and as seems to have been the practice until 1983, then the underlying tax base would rise as property values increased. School districts would not have to call referendums just to keep up. In this case, the primary reason that Red Clay held the Special Election was just to keep up. True, Red Clay contemplated some new initiatives, but what forced Red Clay to appeal to its residents was the relentless grind of inflation, coupled with a steadily increasing student population. If Red Clay’s underlying property values had been reassessed each year at current levels, then Red Clay’s operating revenues would have grown with those reassessments. Red Clay has a legitimate interest in not being forced to hold a referendum again when, if the system functioned properly, Red Clay might never have needed to hold the referendum in the first place. In addition to Red Clay’s interests, there are the interests of the district’s residents. The vast majority of residents did not vote. Over eleven thousand residents did. Decisions in which courts have considered invalidating elections teach that a court should not focus 549 9 Del. C. § 8306(a). 550 Teachers Ins., 669 A.2d at 102. JX 25 at D2548 (“Local taxes are collected by New Castle County and are fixed based on 1983 assessed property values.”); Tr. 645 (Floore). 551 160 exclusively on the interests of those who were unable to participate or whose votes were excluded; the court also should consider the interests of those who took the time to participate by properly casting ballots.552 In this case, opponents of the referendum had the opportunity to go to the polls. Red Clay’s interventions did not prevent anyone from voting. They made it more difficult, as a practical matter, for elderly and disabled resident to vote, but they did not actually bar anyone from participating. The interventions also did not have the likely effect of changing anyone’s vote. They primarily operated by encouraging already supportive voters to get to the polls. Voiding the Special Election would provide those who chose not to vote in the last election with an opportunity to vote in a new election, but at the expense of invalidating the votes of everyone who made the effort. In my view, the potential negative consequences to Red Clay and the voters who participated in the Special Election outweigh the importance of giving voters who chose not to participate a new opportunity to vote. The balancing of hardships therefore does not favor invalidating the Special Election. 6. Weighing the Factors As with other multi-factor balancing tests, the factors cited in this decision are not exclusive, and the analysis is not intended to be a mathematical exercise.553 The tests rather 552 See, e.g., Gjersten, 791 F.2d at 479; Bortner, 736 A.2d at 112-13. 553 Cf., e.g., Bell Helicopter Textron, Inc. v. Arteaga, 113 A.3d 1045, 1051-60 (Del. 2015) (balancing non-exclusive list of factors to be considered when determining choice of law under the Restatement (Second) of Conflict of Laws (Am. Law. Inst. 1971)); 161 guide the court when weighing various interests. In this case, the pervasiveness of Red Clay’s conduct weighs in favor of invalidating the Special Election. All of the other factors counsel to varying degrees against invalidating the Special Election. I therefore conclude that notwithstanding Red Clay’s constitutional violations of the Elections Clause, the Special Election should not be invalidated.554 IV. CONCLUSION In this case, the high-stakes pursuit of a laudable end caused well-intentioned people to resort to improper means. No one should infer from this decision that anyone involved in the Red Clay campaign acted in bad faith or with an ill motive. They sought to achieve what I regard as an unmitigated public good: adequate funding for our state’s public schools. Unfortunately, Delaware’s current system for conducting property tax Martinez v. E.I. DuPont de Nemours & Co., 86 A.3d 1102, 1104-05 (Del. 2014) (identifying non-exclusive list of factors to be considered when conducting forum non conveniens analysis); Tumlinson v. Advanced Micro Devices, Inc., 81 A.3d 1264, 1269 (Del. 2013) (identifying a “non-exhaustive list of factors” that a trial court may consider when evaluating the reliability of expert testimony under Daubert v. Merrell Dow Pharm., Inc., 509 U.S. 579 (1993)); In re Poliquin, 49 A.3d 1115, 1134-35 (Del. 2012) (identifying “non-exhaustive list of aggravating factors” that may be considered when imposing attorney discipline); Sugarland Indus., Inc. v. Thomas, 420 A.2d 142, 149 (Del. 1980) (identifying non-exhaustive list of multiple factors to be considered when awarding attorneys’ fees under common fund and common benefit doctrines). I would reach the same conclusion if this decision had considered the plaintiffs’ federal claims and held that Red Clay’s conduct also violated the Due Process Clause and the Equal Protection Clause. Whether Red Clay’s interventions violated those provisions is not an issue governed by clearly established law or precedent, and Red Clay’s administrators believed in good faith that they were complying with the law. The other factors also remain the same. 554 162 assessments puts school districts in the untenable position of having to ask residents to raise their taxes on a regular basis, typically once every four years. Forced to operate within that system, the Red Clay administrators knew that they needed to prevail in the Special Election. When humans are placed under great pressure to achieve a particular outcome, they look for ways to make it happen. History teaches that high-pressure situations often lead to behavior that comes close to a line or crosses it. In this case, the line was the Elections Clause, which requires that all elections in Delaware be “free and equal.” In its zeal to prevail in the Special Election, Red Clay violated the Elections Clause by holding seventy-five Family-Focused Events, in the twenty-three school buildings that served as polling places, on the day of the Special Election. The Family-Focused Events operated as targeted rewards for the families of Red Clay students and drew them to the polls en masse. An election is not “free and equal” when the government provides targeted rewards for voting to a group it believes will support its favored position. Although not purposeful, the effect of holding the Family-Focused Events in the polling places, on the day of the election, was to interfere with the ability of elderly and disabled residents to vote. The many attendees of the Family-Focused Events crowded the school parking lots. Several of the evening events drew hundreds of people to schools with parking lots that could not accommodate the influx. Some elderly and disabled residents who tried to vote gave up when they could not find accessible parking spaces. Others undoubtedly inferred from the overflowing lots that the voting lines were unmanageably long. Red Clay was obligated to monitor the parking situation at the polling places and 163 ensure that designated parking spots were available for voters. Red Clay knew that the Family-Focused Events would create parking problems and took some steps to prepare, but Red Clay did not fulfill its monitoring obligation, and because of the Family-Focused Events, effective monitoring was not feasible. An election in which the government obstructs the ability of elderly and disabled voters to access the polls is not “free and equal.” Red Clay also violated the Elections Clause by engaging in an election campaign that went far beyond the limited advocacy permitted by governing Delaware Supreme Court precedent. Although I personally favor loosening the restrictions our law places on government campaign speech that is factually accurate, broadly directed to the electorate as a whole, and openly identifies its source, the extent of Red Clay’s government campaign speech went too far. Rather than engaging from the outset in broadly directed communications that would have presented Red Clay’s side as part of an open and vigorous debate about an important policy issue, Red Clay spent three months priming its base of favorable voters, while avoiding communications with groups that might be opposed. Red Clay used tactics that individual candidates and private groups regularly deploy, but a government’s ability to engage in similar advocacy is limited. Red Clay engaged in the equivalent of a full-scale political campaign, which Delaware Supreme Court precedent does not permit. Taken as a whole, Red Clay’s electoral interventions violated the Elections Clause. Notwithstanding that conclusion, I am not convinced that Red Clay’s actions warrant invalidating the Special Election. Precedent from many jurisdictions teaches that electoral violations, even constitutional ones, do not lead ineluctably to a new election. A court must 164 balance multiple considerations before invalidating an election. In this case, that balancing weighs in favor of allowing the Special Election to stand. 165 Appendix A This appendix identifies the school tax referendums in Delaware since 1980. All information is drawn from publicly available sources. The list is comprehensive but not exhaustive. Some sources referred to other referendums for which publicly available data were not readily available. The Special Election is highlighted. The appendix includes three types of referendums: capital, operating, and transfer. A capital referendum seeks approval for a tax increase to fund the construction or renovation of a school, classroom, or similar project. An operating referendum seeks approval for a tax increase to fund a school district’s operating budget. A transfer seeks authority to transfer surplus capital funds to the operating budget or to retain them in a reserve. Date Feb. 12, 1980 District Seaford Milford Woodbridge Capital Type Capital Operating Transfer Transfer Transfer Result Approved, 1,516 to 540 Approved, 1,262 to 754 Approved, 810 to 155 Approved, 811 to 117 Approved, 682 to 428 Mar. 25, 1980 Mar. 25, 1980 May 20, 1980 May 24, 1980 Delmar Transfer Not Available Oct. 22, 1980 Operating Defeated, 46,740 to 4,851 Oct. 29, 1980 New Castle County Smyrna Transfer Not Available Nov. 15, 1980 Indian River Operating Defeated, 2,559 to 1,397 June 9, 1981 Appoquinimink Operating Defeated, 899 to 285 Nov. 10, 1981 Caesar Rodney Transfer Approved, 682 to 106 Dec. 8, 1981 Appoquinimink Operating Defeated, 932 to 536 Source Mark Matthews, Seaford Voters OK School Tax Increase, The Morning News, Feb. 13, 1980, at 18. Budgets Up, But Taxes Aren’t, The Morning News, Mar. 26, 1980, at 16. Parents: More Taxes, More School Nurses, The Morning News, May 21, 1980, at 20. Delmar To Vote on Tax Transfer, The Morning News, May 23, 1980, at 18. Steve Goldberg, Voters Trounce Tax Proposal, The Morning News, Oct. 23, 1980, at A1. Smyrna Schools Seeking Tax Shift, The Evening Journal, Sept. 19, 1980, at A19. Indian River District Denies School Tax Hike, The Sunday News Journal, Nov. 16, 1980, at C3. Janine Jaquet, Property Tax Referendum Is Defeated, The Morning News, June 10, 1981, at C1. Caesar Rodney District Oks Tax Transfer, The Evening Journal, Nov. 11, 1981, at C1. David L. Preston, Appoquinimink Voters Veto Tax Hike; Greenhouse Project May Be Casualty, The Evening Journal, at C8. Date Mar. 9, 1982 District Capital Type Transfer Result Approved, 939 to 549 Mar. 16, 1982 Seaford Operating Defeated, 953 to 906 Mar. 30, 1982 Milford Operating Approved, 1,256 to 992 May 11, 1982 Smyrna Operating Defeated, 763 to 450 May 15, 1982 Indian River Capital Approved, 1,936 to 1,179 June 8, 1982 Appoquinimink Operating Approved, 1,360 to 1,215 Oct. 12, 1982 Seaford Transfer Approved, 1,066 to 126 Nov. 20, 1982 Delmar Oct. 25, 1983 Smyrna Operating Transfer Operating Defeated, 200 to 89 Approved, 175 to 113 Approved, 898 to 667 Nov. 10, 1983 Cape Henlopen Dec. 14, 1983 Capital Capital Operating Operating Approved, 1,857 to 659 Approved, 1,532 to 865 Approved, 1,018 to 840 Mar. 6, 1984 Christina Operating Defeated, 4,874 to 2,602 Mar. 28, 1984 Indian River May 8, 1984 May 8, 1984 Mar. 21, 1985 Lake Forest Caesar Rodney Laurel Mar. 28, 1985 Milford Operating Transfer Operating Operating Operating Transfer Operating Approved, 2,298 to 1,756 Approved, 2,433 to 1,613 Approved, 942 to 928 Approved, 1,561 to 713 Approved, 984 to 615 Approved, 1,202 to 402 Approved, 1,260 to 969 Source Jane Brooks, Capital District Voters Support School Repairs, The Evening Journal, Mar. 10, 1982, at B8. Kevin Feeley, Seaford Votes Down Tax Hike, The Morning News, Mar. 31, 1982, at B1. Kevin Feeley, Voters OK Tax Hike in Milford, The Morning News, Mar. 17, 1982, at B1. Nathan Gorenstein, Posts Filled on State School Boards; Voters Reject Smyrna Tax Hike, The Morning News, May 12, 1982, at D1. Voters OK Tax Hike, The News Journal, May 16, 1982, at B10. Dan Piper, Appoquinimink Tax Rises, The Evening Journal, June 9, 1982, at A1. Seaford School District Voters OK Tax Transfer, The Morning News, Oct. 13, 1982, at B1. Robin Brown, Delmar Votes to Hike Funds for Schools, The News Journal, Nov. 21, 1982, at B3. Tax Rise OK’d In Smyrna, The Morning News, Oct. 26, 1983, at B1. Henlopen Voters OK Tax Hikes, The Morning News, Nov. 11, 1983, at B4B. Jane Brooks, Dover Area OKs School-Tax Boost, The Morning News, Dec. 15, 1983, at A1. Laurie Hays, Christina Tax Hike Rejected, The Evening Journal, Mar. 7, 1984, at A1. Dennis Friedel, Indian River OKs Tax Hike, The Morning News, Mar. 29, 1984, at A1. Jane Brooks & Rowan Scarborough, 2 Kent Districts OK Hikes, The News Journal, May 9, 1984, at B3. Molly Murray, Laurel Voters OK Tax Increases, The News Journal, Mar. 23, 1985, at A4. Jerry Hager, Milford Votes Tax Hike to Aid School Date District Type Transfer Operating Result Approved, 1,519 to 698 Approved, 1,073 to 504 May 14, 1985 Seaford May 18, 1985 Woodbridge Delmar Operating Transfer Operating Approved, 785 to 391 Approved, 964 to 196 Approved, 249 to 128 Nov. 9, 1985 Nov. 14, 1985 Cape Henlopen Operating Approved, 1,299 to 1,044 Nov. 16, 1985 Lake Forest Capital Defeated, 1,388 to 691 Nov. 19, 1985 Smyrna Operating Approved, 933 to 488 Dec. 10, 1985 Capital Operating Approved, 936 to 793 Dec. 11, 1985 Appoquinimink Operating Approved, 944 to 744 Feb. 4, 1986 Caesar Rodney Operating Approved, 1,136 to 680 Mar. 18, 1986 Indian River Operating Approved, vote tally not available Oct. 7, 1986 Christina Operating Capital Approved, 4,827 to 3,412 Approved, 5,143 to 3,100 Mar. 31, 1987 Brandywine Operating Defeated, 6,841 to 6,313 Source System, The Morning News, Mar. 29, 1985, at A1. Seaford Voters OK Tax Hikes for Schools, The Morning News, May 15, 1985, at A1. Molly Murray, School Tax Hike Approved, The Morning News, May 21, 1985, at B3. Molly Murray, Delmar Voters OK Tax Hike, The News Journal, Nov. 10, 1985, at C6. Carolyn Lewis, Property Tax Hike Approved, The Morning News, Nov. 15, 1985, at B1. Molly Murray, Lake Forest Rejects Tax Hike for Schools, The News Journal, Nov. 17, 1985, at C1. Pattie Sewell, Smyrna School Voters OK Tax Increase, The Morning News, Nov. 20, 1985, at B2. Jane Brooks, Capital District to Get Tax Hike, The Morning News, Dec. 11, 1985, at B6 Voters OK Higher Tax for Schools, The Morning News, Dec. 12, 1985, at B2. Phil Milford, Tax Boost Approved, The News Journal, Feb. 5, 1986, at B1; Molly Murray, Rodney School Post Is Filled, The Morning News, Apr. 16, 1986, at B4. Dennis Friedel, Indian River Seeks Tax Hike for Schools, The News Journal, Feb. 13, 1986, at B1; Michael Jackson, Indian River OKs Tax, The News Journal, Mar. 19, 1986, at B1. Sandy Dennison, Christina District’s Board to Seek 17% Increase in Taxes, Evening Journal, Aug. 13, 1986, at A1; Voters Approve Christina Tax Increase, The Morning News, Oct. 8, 1986, at B8C. Sandy Dennison, Brandywine District Says No to Tax Increase, The Evening Journal, Apr. 1, 1987, at A1, A4. Date May 19, 1987 District Seaford Type Capital Oct. 27, 1987 Brandywine Operating Nov. 10, 1987 Lake Forest Operating Result Approved, vote tally not available Approved, vote tally not available Approved, 11,938 to 6,589 Defeated, 1,383 to 1,071 Capital Defeated, 1,257 to 1,215 Operating Nov. 10, 1987 Woodbridge Capital Defeated, 1,009 to 635 Mar. 29, 1988 Capital Operating Approved, 1,634 to 1,209 Apr. 12, 1988 Smyrna Operating Transfer Approved, 1,450 to 1,443 Approved, 2,213 to 698 May 17, 1988 Lake Forest Operating Approved, 2,068 to 1,051 Capital Approved, 2,193 to 1,351 May 24, 1988 Caesar Rodney Operating Approved, 1,051 to 543 Apr. 13, 1989 Laurel Capital Defeated, 777 to 684 May 9, 1989 Christina Capital Approved, 3,274 to 1,629 Mar. 7, 1990 Laurel Capital Approved, 1,272 to 874 Apr. 4, 1990 Red Clay Operating Defeated, 6,569 to 6,101 Source Carolyn Lewis, Seaford Voters OK Bond Issue, The Evening Journal, May 20, 1987, at B3. Sandy Dennison, Brandywine Voters OK School Tax Hike, The Evening Journal, Oct. 28, 1987, at A1, A4. Lake Forest Tax Vote, The Evening Journal, Sept. 21, 1987, at A3; Voters Kill Tax Hike, The Evening Journal, Nov. 11, 1987, at B1. Molly Murray, Woodbridge Schedules Referendum, The Morning News, Oct. 14, 1987, at B1; School Plan Rejected, The Evening Journal, Nov. 11, 1987, at B1. Capital Voters OK Hike in School Taxes, The Evening Journal, Mar. 30, 1988, at B1. Nancy Kesler, Shaky Smyrna School District Ekes Out Tax Hike Vote Victory, The Morning News, Apr. 13, 1988, at A1, A12. Rhonda Graham, Lake Forest Voters Approve Increase In School Taxes, The Morning News, May 18, 1988, at B1. Higher Property Taxes OK’d, The Evening News, May 25, 1988, at B2 Carolyn Lewis, Voters Say “No” to Laurel’s $6.5 Million School Bond Bill, The News Journal, Apr. 14, 1989, at B4. Sandy Dennison, Christina School District Votes 3 Percent Tax Hike, The News Journal, May 10, 1989, at A1. Nancy E. Lynch, School Bond Issue Approved, The News Journal Delmarva, Mar. 8, 1990, at B2A. Sandy Dennison, Proposal Loses By Less Than 500 Date District Type Result May 8, 1990 Christina Capital Defeated, 5,083 to 3,452 May 8, 1990 Capital Capital Approved, 1,424 to 524 Oct. 4, 1990 Red Clay Operating Defeated, 9,858 to 7,007 Oct. 11, 1990 Appoquinimink Operating Defeated, 1,505 to 1,267 Dec. 6, 1990 Christina Apr. 9, 1991 Milford Operating Approved, 6,244 to 4,920 (17 cents); Capital (1.5 cents) Capital Approved, 1,164 to 798 Apr. 16, 1991 Appoquinimink Operating Approved, 1,791 to 1,572 May 2, 1991 Red Clay Operating Oct. 22, 1991 Appoquinimink Transfer Approved, 10,814 to 8,585 Approved, 866 to 348 Nov. 7, 1991 Indian River Capital Defeated, 3,060 to 2,883 Capital Defeated, 3,547 to 3,022 Nov. 7, 1991 Smyrna Capital Approved, 717 to 642 Nov. 9, 1991 Woodbridge Capital Approved, 1,207 to 443 Source Votes, The News Journal, Apr. 5, 1990, at A1. Eric Ruth, Christina Tax Hike Is Rejected, The News Journal, May 9, 1990, at A1, A4. Jeff Montgomery, Capital Tax Hike Approved, The News Journal Delmarva, May 9, 1990, at B2A. Eric Ruth, Red Clay Tax Hike Defeated, The News Journal, Oct. 5, 1990, at A1. Tom Curley, Appoquinimink Voters Reject School Tax Hike, The News Journal Delmarva, Oct. 13, 1990, at A4A. Eric Ruth, Christina Voters OK 24 Percent Tax Increase, The News Journal, Dec. 7, 1990, at A1, A19. Milford Voters OK Tax Hikes, The News Journal Delmarva, Apr. 10, 1991, at B1. Tom Curley, Appoquinimink Voters OK Tax Hike, The News Journal, Apr. 17, 1991, at B1. Sandy Dennison, Red Clay Voters Approve 29% Tax Increase, The News Journal, May 3, 1991, at A1, A4. Tom Curley, Appoquinimink Voters OK Bond Money for Redding, The News Journal Delmarva, Oct. 23, 1991, at B1. Bruce Pringle, Schools Building Plans Rejected by Indian River Residents, The News Journal Delmarva, Nov. 8, 1991, at A1, A16. Kanchalee Svetvilas, Smyrna Voters OK Plans for School Work, The News Journal Delmarva, Nov. 8, 1991, at B1. Eddy J. Parker, Schools Need Renovation, The News Journal, Nov. 5, 1991, at A7; Nan Clements & Molly Date District Type Result Mar. 12, 1992 Cape Henlopen Operating Defeated, 1,433 to 1,226 May 14, 1992 Indian River Capital Approved, 3,636 to 3,609 May 23, 1992 Laurel Transfer Approved, 197 to 55 May 30, 1992 Delmar Not available Oct. 6, 1992 Colonial Not available Operating Mar. 9, 1993 Capital Capital Approved, 1,060 to 165 May 4, 1993 Colonial Operating Approved, 7,082 to 5,228 May 8, 1993 Cape Henlopen Capital Approved, 839 to 225 May 11, 1993 Appoquinimink Capital Approved, 805 to 623 Nov. 2, 1993 Brandywine Operating Capital Defeated, 9,827 to 8,866 Approved, 8,518 to 8,476 Defeated, 3,900 to 2,988 Source Murray, Woodbridge Voters OK $10 Million Capital, The News Journal, Nov. 10, 1991, at C3. Bruce Pringle, Cape District Residents To Vote On Tax Hike, The News Journal Delmarva, Mar. 2, 1992, at A3; Chris Donahue & Bruce Pringle, Henlopen Tax Increase Is Rejected, The News Journal Delmarva, Mar. 13, 1992, at B1. Bruce Pringle, Review Mixed for Indian River Expansion Plan, The News Journal Delmarva, Mar. 28, 1992, at A3; Indian River OKs School Building, The News Journal, May 15, 1992, at B4. Laurel Voters OK Debt Transfer, The News Journal, May 24, 1992, at B6. Special Notice, The News Journal, May 15, 1992, at C8. Sandy Dennison, Colonial Asks Voters to Approve Tax Hike, The News Journal, Oct. 4, 1992, at B7; Referendum Report, The News Journal Crossroads, Oct. 15, 1992, at 16. School Spending Ok’d, The News Journal, Mar. 10, 1993, at B1. Sandy Dennison, Colonial Tax Hike Wins OK, The News Journal, May 5, 1993, at A1. Cris Barrish & Nan Clements, Cape Residents Approve Tax Hike to Fix Schools, The News Journal, May 9, 1993, at B7. Sue Denny, Appoquinimink Lowers Tax Proposal, The News Journal, Apr. 8, 1993, at B2; Sandy Dennison, Runoff Looms for Lake Forest Board, The News Journal, May 12, 1993, at B3. Sandy Dennison, Brandywine Voters Split on Tax Hikes, The News Journal, Nov. 3, 1993, at A1; Sandy Date District Type Result Nov. 9, 1993 Nov. 9, 1993 Caesar Rodney Capital Capital Capital Approved, 1,208 to 905 Approved, 1,117 to 747 Nov. 18, 1993 Cape Henlopen Operating Not available Mar. 8, 1994 Mar. 8, 1994 Mar. 29, 1994 Lake Forest Milford Seaford Capital Operating Operating Defeated, 1,506 to 897 Defeated, 1,996 to 649 Defeated, 1,268 to 1,128 May 3, 1994 Brandywine Operating June 14, 1994 Appoquinimink Transfer Approved, 14,579 to 10,669 Approved, 306 to 109 Feb. 7, 1995 Red Clay Operating Defeated, 9,296 to 6,586 Feb. 14, 1995 Appoquinimink Operating Defeated, 1,632 to 777 Mar. 7, 1995 Christina Capital Approved, 4,800 to 3,162 Mar. 28, 1995 Seaford Operating Defeated, 1,711 to 1,353 Apr. 11, 1995 Apr. 11, 1995 May 9, 1995 Caesar Rodney Cape Henlopen Appoquinimink Operating Transfer Operating Defeated, 1,559 to 1,034 Approved, 1,529 to 930 Defeated, 2,259 to 2,119 Source Dennison, School Officials in Retreat, The News Journal, June 26, 1993, at A1. Robert Moore, School Tax Hikes Approved in Kent, The News Journal Kent & Sussex, Nov. 10, 1993, at A1. Bruce Pringle, Cape Henlopen School Tax Hike Up for Vote Again, The News Journal Kent & Sussex, Nov. 15, 1993, at A3. Kim Hoey, School Funding Plans Rejected, The News Journal Kent & Sussex, Mar. 9, 1994, at B1. Nan Clements, Seaford Sinks New School Tax, The News Journal Kent & Sussex, Mar. 30, 1994, at A1. Sandy Dennison, Vote Gives Brandywine New Life, The News Journal, May 4, 1994, at A1. Appoquinimink Fund Transfer OK’d, The News Journal, June 15, 1994, at B3. Esteban Parra, Red Clay Tax Boost Rejected, The News Journal, Feb. 8, 1995, at A1. Joy Gwillim, Appoquinimink Tax Hike Fails, The News Journal, Feb. 15, 1995, at A1. Esteban Parra, Christina Tax Hike Proposal Supported, The News Journal Crossroads, Dec. 22, 1994, at 10; Eric Ruth, Christina Tax Hike OK’d, The News Journal, Mar. 8, 1995, at B1. Chris Donahue, Tax Increase Rejected in Seaford Vote, The News Journal Kent & Sussex, Mar. 29, 1995, at A1. Rodney District: No Tax Hike, The News Journal Kent & Sussex, Apr. 12, 1995, at B1. Amy Knowles, Appoquinimink: No Tax Hike, The News Journal Kent & Sussex, May 10, 1995, at B1. Date May 24, 1995 District Red Clay Type Operating Feb. 6, 1996 Seaford Capital Result Approved, 12,145 to 10,644 Approved, 855 to 771 Feb. 15, 1996 Indian River Operating Approved, 4,360 to 3,331 Mar. 12, 1996 Colonial Capital Defeated, 3,292 to 703 Mar. 12, 1996 Capital Capital Approved, 963 to 602 Apr. 4, 1996 Caesar Rodney Capital Approved, 1,477 to 1,288 Apr. 27, 1996 Appoquinimink Operating Defeated, 1,750 to 1,429 Dec. 3, 1996 Colonial Capital Defeated, 2,504 to 1,266 Mar. 8, 1997 Delmar Capital Approved, 708 to 267 Mar. 11, 1997 Lake Forest Capital Operating Defeated, 1,271 to 1,260 Defeated, 1,407 to 1,123 Mar. 18, 1997 Red Clay Capital Mar. 25, 1997 Milford Operating Defeated, vote tally not available Approved, 2,196 to 1,314 May 6, 1997 Lake Forest Capital Operating Defeated, 2,186 to 1,874 Defeated, 2,399 to 1,647 Source Esteban Parra, Red Clay Passes Tax Hike, The News Journal, May 25, 1995, at A1. Bruce Pringle, Seaford Voters OK School Funds, The News Journal Kent & Sussex, Feb. 7, 1996, at A1. Tax Hike Referendum Set For Indian River District, The News Journal Kent & Sussex, Dec. 18, 1995, at A5; Esteban Parra, District Adds 45 New Teachers, The News Journal Kent & Sussex, Sept. 4, 1996, at B1. Nan Clements, Colonial Proposal Defeated, The News Journal, Mar. 13, 1996, at A1. Robert Moore, Capital Voters OK Tax Hike, The News Journal Kent & Sussex, Mar. 13, 1996, at A1. James Merriweather, Caesar Rodney Voters OK Tax Hike, The News Journal Kent & Sussex, Apr. 5, 1996, at A1. Mike Billington & Edward L. Kenney, School Tax Proposal Defeated, The News Journal, Apr. 28, 1996, at A1. Nan Clements, School Tax Increase Voted Down, The News Journal, Dec. 4, 1996, at A1. Nan Clements, Delmar Votes to Build School, The News Journal, Mar. 9, 1997, at B2. James Merriweather, Lake Forest Tax Hikes Defeated, The News Journal Kent & Sussex, Mar. 12, 1997, at A1. Beverly James Coleman, Red Clay Tax Hike Rejected, The News Journal, Mar. 19, 1997, at A1. Kim Hoey, Milford School Tax Hike Approved, The News Journal Kent & Sussex, Mar. 26, 1997, at B1. James Merriweather & Patricia V. Rivera, Lake Forest Voters Kill Tax Hike, The News Journal, May 7, 1997, Date District Type Result May 13, 1997 Appoquinimink Capital Defeated, 1,590 to 1,194 Mar. 10, 1998 Red Clay Capital Defeated, 7,838 to 6,652 Capital Approved, 7,980 to 6,566 Operating Approved, 7,700 to 6,781 Mar. 19, 1998 Capital Operating Defeated, 1,744 to 1,009 Mar. 31, 1998 Woodbridge Operating Defeated, 739 to 419 Apr. 21, 1998 Lake Forest Capital Defeated, 1,361 to 1,129 Apr. 28, 1998 Caesar Rodney Capital Defeated, 1,644 to 1,240 May 12, 1998 Appoquinimink Capital Operating Approved, 2,359 to 2,341 Approved, 2,321 to 2,218 June 2, 1998 Red Clay Capital Defeated, 5,449 to 3,396 Mar. 11, 1999 Lake Forest Capital Approved, 2,025 to 1,343 Mar. 23, 1999 Milford Capital Approved, 1,536 to 1,258 Mar. 30, 1999 Capital Capital Approved, 1,683 to 1,147 Source at B5. Stephen Chrzanowski, Appoquinimink School District Referendum Rejected, The News Journal, May 14, 1997, at B1. Allison Taylor & Esteban Parra, Red Clay Wins Split Decision, The News Journal, Mar. 11, 1998, at A1, A12. James Merriweather, Tax Hike in Capital Is Rejected, The News Journal Kent & Sussex, Mar. 20, 1998, at B1. Lynn Parks, After 10 Years, Track Is Back at Woodbridge High School, The News Journal Delmarva Crossroads, Apr. 15, 1998, at 1. James Merriweather, Lake Forest Tax Hike Is Rejected, The News Journal Kent & Sussex, Apr. 22, 1998, at B1. James Merriweather, Tax Hike Is Rejected by Voters, The News Journal Kent & Sussex, Apr. 29, 1998, at B1. Allison Taylor & Sean O’Sullivan, Appoquinimink Voters Approve School Tax Hike, The News Journal, May 13, 1998, at A1. Edward L. Kenney, Red Clay Tax Hike Is Rejected, The News Journal, June 3, 1998, at B1. Lake Forest Voters OK Tax Increase, The News Journal, Mar. 12, 1998, at B2. James Merriweather, Tax Hike OK’d in Milford, The News Journal Kent & Sussex, Mar. 24, 1999, at A1. James Merriweather, Capital Tax Hike Approved, The News Journal Kent & Sussex, Mar. 31, 1999, at A1. Date Nov. 23, 1999 District Caesar Rodney Type Capital Result Approved, 2,068 to 1,643 Nov. 29, 1999 Colonial Capital Approved, 2,418 to 1,395 Feb. 26, 2000 Smyrna Feb. 29, 2000 Woodbridge Capital Operating Capital Operating Approved, 1,493 to 1,104 Approved, 1,478 to 1,132 Defeated, 1,022 to 958 Defeated, 1,034 to 917 Mar. 7, 2000 Cape Henlopen Capital Approved, 2,920 to 1,322 Mar. 23, 2000 Indian River Capital Defeated, 3,526 to 3,425 May 6, 2000 Woodbridge May 9, 2000 Appoquinimink Capital Operating Transfer Defeated, 1,124 to 889 Defeated, 1,126 to 843 Approved, 1,730 to 335 May 23, 2000 Indian River Capital Approved, 4,909 to 3,528 June 3, 2000 Lake Forest Operating Defeated, 934 to 541 Feb. 13, 2001 Lake Forest Operating Approved, 2,138 to 1,565 Apr. 3, 2001 Christina Capital Capital Defeated, 3,072 to 2,864 Defeated, 3,482 to 2,400 Apr. 10, 2001 Red Clay Capital Defeated, 6,271 to 5,532 Source James Merriweather, Caesar Rodney Tax Boost Approved, The News Journal, Nov. 24, 1999, at B2. Stephen Sobek, Colonial Voters OK New School, The News Journal, Nov. 30, 1999, at A1, A11. Chip Guy, District’s School-Tax Hike OK’d, The News Journal, Feb. 27, 2000, at B1. Molly Murray, Woodbridge Defeats School Referendum, The News Journal Kent & Sussex, Mar. 2, 2000, at B1. Molly Murray, Voters Pass Plan to Build Two Schools, The News Journal Kent & Sussex, Mar. 8, 2000, at B1. Patricia V. Rivera, Indian River Rejects a Tax Hike, The News Journal Kent & Sussex, Mar. 24, 2000, at A1, A9. Molly Murray, District Residents Reject 2nd Tax-Hike Referendum, The News Journal, May 7, 2000, at B1. Stephanie L. Arnold, School Building Plan OK’d, The News Journal, May 10, 2000, at A1. Patricia V. Rivera, Indian River’s Spending Plan Wins Voters’ Approval, The News Journal, May 24, 2000, at B6. James Merriweather, Lake Forest District Says No to Tax Hike, The News Journal, June 4, 2000, at B5. James Merriweather, Lake Forest Voters OK a Tax Hike, The News Journal Kent & Sussex, Feb. 14, 2001, at B1. Michele Besso, Christina Schools Face Vote, The News Journal, Apr. 3, 2001, at B1, B5; Michele Besso, Voters Defeat Christina Proposal, The News Journal, Apr. 4, 2001, at A1, A7. Stephen Sobek, Red Clay’s Plans Are Rejected, The Date District Type Result May 3, 2001 Indian River Capital Operating Capital Defeated, 2,112 to 1,874 Defeated, 2,222 to 1,740 Defeated, 2,137 to 1,814 May 8, 2001 Woodbridge Capital Operating Approved, 1,077 to 1,016 Defeated, 1,016 to 976 May 19, 2001 Delmar Transfer Approved, 133 to 2 May 31, 2001 Brandywine Capital Capital Approved, 7,554 to 2,313 Approved, 7,120 to 2,822 Mar. 6, 2002 Red Clay Capital Approved, 4,986 to 4,473 Mar. 26, 2002 Woodbridge Operating Defeated, 876 to 682 Apr. 9, 2002 Seaford Transfer Approved, 731 to 175 Apr. 18, 2002 Christina Capital Approved, 4,291 to 1,979 Capital Approved, 3,505 to 2,054 Apr. 23, 2002 Brandywine Operating Approved, 6,971 to 2,297 May 9, 2002 Delmar Transfer Approved, 85 to 5 May 23, 2002 Caesar Rodney Transfer Approved, 1,564 to 284 Source News Journal, Apr. 11, 2001, at A1, A12. Molly Murray, Indian River Voters Go to Polls Thursday, The News Journal Kent & Sussex, May 2, 2001, at B1-2; Molly Murray, Schools Plan Gets a No Vote, The News Journal Kent & Sussex, May 4, 2001, at B1. Molly Murray, Woodbridge Referendum to Go to Voters Again, The News Journal, May 7, 2001, at B5; Molly Murray, Residents Approve Tax Hike, The News Journal Kent & Sussex, May 9, 2001, at B1. Delmar District Votes to Transfer Money, The News Journal, May 20, 2001, at B3. Kate Bailey, Brandywine Sets Referendum On School Repairs, The News Journal, Mar. 23, 2001, at B1; Stephen Sobek, School Overhauls Approved, The News Journal, June 1, 2001, at A1. Stephen Sobek, Red Clay Tax Hike Gets Voter Approval, The News Journal, Mar. 7, 2002, at A1, A7. Mike Billington, Defeated Tax Hike Means New Woodbridge School Won’t Open, The News Journal Kent & Sussex, Mar. 28, 2002, at B3. Chip Guy, Seaford Voters OK Debt-Service Money Transfer, The News Journal Kent & Sussex, Apr. 10, 2002, at B1. Michele Besso, Christina Voters OK Tax Increase, The News Journal, Apr. 19, 2002, at A1, A16. Stephen Sobek, Voters OK Tax Boost for Schools, The News Journal, Apr. 24, 2002, at A1, A12 Delmar Voters Pass School’s Tax Proposal, The News Journal Kent & Sussex, May 12, 2002, at B3. James Merriweather, Voters Approve Tax Plan, The Date District Type Result June 1, 2002 Woodbridge Operating Approved, 1,802 to 1,202 Mar. 18, 2003 Indian River Transfer Approved, 2,179 to 174 Apr. 10, 2003 Christina Nov. 12, 2003 Appoquinimink Jan. 27, 2004 Woodbridge Operating Operating Capital Operating Operating Transfer Transfer Approved, 5,530 to 2,270 Approved, 5,334 to 2,431 Approved 3,363 to 1,450 Approved 3,087 to 1,711 Approved 2,885 to 1,939 Approved, 3,689 to 1,105 Approved, 547 to 378 Feb. 7, 2004 Delmar Transfer Approved, 194 to 28 Feb. 26, 2004 Red Clay Operating Approved, 4,281 to 3,658 Mar. 30, 2004 Indian River Capital Operating Approved, 1,960 to 1,482 Approved, 1,953 to 1,544 Oct. 12, 2004 Capital Operating Defeated, 1,883 to 1,444 Feb. 26, 2005 Smyrna Capital Approved, 1,114 to 429 Mar. 8, 2005 Capital Operating Approved, 2,491 to 1,784 Source News Journal Kent & Sussex, May 24, 2002, at B1. Molly Murray, Woodbridge Schools Face a Crucial Vote, The News Journal, May 31, 2002, at B5; Molly Murray, Woodbridge District Tax Hike Passes, The News Journal, June 2, 2002, at A1. Molly Murray, Indian River Athletic-Fields Plan Passes, The News Journal Kent & Sussex, Mar. 19, 2003, at B3. Melissa Tyrrell, Voters OK Christina Tax Increase, The News Journal, Apr. 11, 2003, at B1. Melissa Tyrrell, District’s Voters OK Plans, The News Journal, Nov. 13, 2003, at A1, A4. Murali Balaji, Voters OK Woodbridge Upgrades, The News Journal, Jan. 28, 2004, at B1. Melissa Tyrrell, Voters Approve Delmar School Expansion, The News Journal, Feb. 8, 2004, at B3. Mike Billington & Michele Besso, Red Clay OKs More Funds, The News Journal, Feb. 27, 2004, at B1. Molly Murray, School Tax Hike Approved, The News Journal Kent & Sussex, Mar. 31, 2004, at B1. James Merriweather, Capital District Rejects Tax Hike, The News Journal Kent & Sussex, Oct. 13, 2004, at A1. Chris Barrish, Smyrna Vote OKs Tax Hike, The News Journal, Feb. 27, 2005, at B1. James Merriweather, Voters OK Tax Increase for Capital Schools, The News Journal Kent & Sussex, Mar. 9, 2005, at B1. Date Mar. 22, 2005 District Milford Type Operating Result Approved, 1,882 to 1,097 Apr. 12, 2005 Seaford Capital Approved, 593 to 380 Apr. 12, 2005 Lake Forest Transfer Approved, 950 to 66 May 5, 2005 Laurel Transfer Approved, 533 to 34 May 24, 2005 Brandywine Transfer Capital Capital Operating Capital Operating Approved, 5,371 to 1,994 Approved, 4,995 to 1,904 Approved, 4,908 to 2,450 Approved, 5,018 to 2,321 Approved, 2,467 to 936 Approved, 2,251 to 1,140 Capital Operating Operating Capital Defeated, 2,268 to 2,229 Defeated, 2,376 to 2,118 Defeated, 2,311 to 2,191 Approved, 997 to 421 Sept. 27, 2005 Capital Dec. 6, 2005 Caesar Rodney Dec. 15, 2005 Cape Henlopen Jan. 26, 2006 Christina Capital Operating Defeated, 4,798 to 2,921 Defeated, 4,798 to 2,921 Feb. 25, 2006 Smyrna Capital Defeated, 1,231 to 1,108 Source James Merriweather, Milford Voters Approve Tax Increase, The News Journal Kent & Sussex, Mar. 23, 2005, at B3. Chip Guy, Seaford Voters Pass Bond Referendum, The News Journal Kent & Sussex, Apr. 13, 2005, at B1. Lake Forest OKs Funds Transfer, The News Journal Kent & Sussex, Apr. 13, 2005, at B3. Murali Balaji, Laurel District Voters OK Tax Transfer, The News Journal Kent & Sussex, May 6, 2005, at B3. Edward L. Kenney, Brandywine Voters OK School Upgrades, The News Journal, May 25, 2005, at B1-2. Edward L. Kenney, Colonial Bond Questions Win Voter Approval, The News Journal, Sept. 28, 2005, at B1. James Merriweather, CR Rejects Tax Hikes By Slim Margins, The News Journal Kent & Sussex, Dec. 7, 2005, at B1. Molly Murray, Cape Vote on Tax Hike Is Thursday, The News Journal Kent & Sussex, Dec. 13, 2005, at B1; Molly Murray, Cape Henlopen Voters OK Tax Hike for Air Conditioning, The News Journal, Dec. 16, 2005, at B5. Edward L. Kenney, Christina Voters Question Second School Bond, The News Journal, Jan. 14, 2006, at B12; Edward L. Kenney, Christina Voters Defeat Referendum, The News Journal, Jan. 27, 2006, at A1, A6. Jeff Montgomery, Smyrna’s Voters Reject Tax Date District Type Result Feb. 28, 2006 Seaford Operating Approved, 943 to 603 Mar. 4, 2006 Caesar Rodney Operating Capital Defeated, 3,207 to 2,554 Defeated, 3,117 to 2,721 Mar. 16, 2006 Cape Henlopen Capital Operating Approved, 3,178 to 2,733 Defeated, 3,012 to 2,805 Mar. 28, 2006 Indian River Operating Operating Approved, 2,315 to 1,971 Approved, 2,378 to 1,885 Approved, 2,259 to 2,006 Approved, 482 to 386 Apr. 26, 2006 Woodbridge Capital Capital Oct. 10, 2006 Smyrna Capital Defeated, 1,628 to 1,556 Dec. 12, 2006 Appoquinimink Capital Capital Approved, 2,572 to 1,320 Approved, 2,298 to 1,457 Dec. 14, 2006 Lake Forest Capital Approved, 731 to 696 Feb. 15, 2007 Milford Capital Approved, 1,215 to 1,052 Source Increase, The News Journal Kent & Sussex, Feb. 26, 2006, at B1, B3. Patricia V. Rivera, Seaford School District Voters OK Referendum, The News Journal Kent & Sussex, Mar. 2, 2006, at B3. James Merriweather, CR District to Vote on Tax Hike in Mar., The News Journal Kent & Sussex, Jan. 18, 2006, at B1-2; Voters Put the Squeeze on School Kids, Midstate Living, Apr. 2006, at 11. Molly Murray, Cape Voters Approve Tax to Build New High School, The News Journal, Mar. 17, 2006, at B3. Patricia V. Rivera, Indian River Voters OK Tax Hike for Raises, Repairs, The News Journal Kent & Sussex, Mar. 29, 2006, at A1-2. Woodbridge School District Voters Approve Referendum, The News Journal Kent & Sussex, Apr. 27, 2006, at B3. Andrew Tangel, Smyrna School Building Plan Defeated, The News Journal Kent & Sussex, Oct. 11, 2006, at B1. Andrew Tangel, Appoquinimink Referendum Dec. 12, Midstate Living, Dec. 2006, at 33; Andrew Tangel, Appoquinimink Voters Say Yes to Tax Increase, The News Journal, Dec. 13, 2006, at B1, B3. J.L. Miller, Lake Forest Voters OK Bond Issue to Upgrade Schools, The News Journal, Dec. 15, 2006, at B5. James Merriweather, Milford Approves Tax for Schools, The News Journal Kent & Sussex, Feb. 16, 2007, at B1. Date Mar. 3, 2007 District Smyrna Type Capital Mar. 7, 2007 Capital Capital Result Approved, vote tally not available Defeated, vote tally not available Approved, 1,553 to 1,133 Apr. 24, 2007 Brandywine Capital Operating Defeated, 4,865 to 4,330 Defeated, 4,800 to 4,332 May 3, 2007 Caesar Rodney May 17, 2007 Red Clay Capital Operating Operating Operating Approved, 1,705 to 1,694 Defeated, 1,971 to 1,424 Defeated, 2,010 to 1,388 Defeated, 6,220 to 4,822 June 4, 2007 Brandywine Operating Approved, 7,584 to 6,305 Nov. 6, 2007 Christina Capital – Limited Plan Capital – Full Plan Operating Approved, 6,600 to 2,608 Operating Operating Defeated, 1,830 to 1,467 Defeated, 1,735 to 1,552 Capital Feb. 28, 2008 Red Clay May 22, 2008 Indian River Source Diane Mouskorie, Smyrna Tries Again, Midstate Living, Dec. 2006, at 32; Andrew Tangel, One Win, One Loss for Smyrna School Plan, The News Journal, Mar. 4, 2007, at B1. James Merriweather, $13 Million Bond OK’d for Capital District, The News Journal Kent & Sussex, Mar. 8, 2007, at B1. Brandywine School Board Sets Date to Hold Operating Referendum, The News Journal, Mar. 13, 2007, at B3; Edward L. Kenney, Voters Reject Referendum for Brandywine, The News Journal, Apr. 25, 2007, at B1. James Merriweather, Caesar Rodney Voters Approve Building Plan, The News Journal Kent & Sussex, May 4, 2007, at B1-2. Edward L. Kenney, Red Clay Hopes for Best, Braces for Worst in Vote, The News Journal, May 16, 2007, at B3; Edward L. Kenney, Red Clay Voters Defeat Operating Referendum, The News Journal, May 18, 2007, at B1. Edward L. Kenney, Brandywine Voters Say Yes to Plan, The News Journal, June 5, 2007, at B1-2. Edward L. Kenney, Christina District’s Proposal Approved, The News Journal, Nov. 7, 2007, at B1-2. Approved, 6,786 to 1,916 Approved, 8,550 to 7,414 Edward L. Kenney, Voters Approve Red Clay Spending, The News Journal, Feb. 29, 2008, at A1, A8. Molly Murray, Indian River District Voters Reject Proposals, The News Journal Kent & Sussex, May 23, Date District Cape Henlopen Type Operating Transfer Transfer Transfer Result Defeated, 1,887 to 1,389 Defeated, 1,257 to 955 Defeated, 1,603 to 603 Defeated, 1,769 to 1,650 June 12, 2008 Cape Henlopen Feb. 24, 2009 Dec. 8, 2009 Appoquinimink Transfer Approved, 2,229 to 409 Mar. 30, 2010 Indian River Transfer Approved, 602 to 84 Mar. 31, 2010 Capital Capital Approved, 1,729 to 1,469 Mar. 31, 2010 Laurel Capital Capital Approved, 1,612 to 1,114 Defeated, 1,444 to 1,241 Apr. 13, 2010 Seaford Capital Approved, 475 to 222 May 25, 2010 Christina Operating Approved, 3,049 to 3,023 Oct. 4, 2010 Laurel Capital Capital Approved, 1,214 to 1,200 Defeated, 1,265 to 1,128 Mar. 29, 2011 Woodbridge Capital Approved, 964 to 653 Source 2008, at B3. Molly Murray, Cape Voters Reject Proposals, Take Tax Cut, The News Journal, June 13, 2008, at B3. Molly Murray, Cape District Voters Reject Money Shift, The News Journal Kent & Sussex, Feb. 25, 2009, at B3. Edward L. Kenney & Ira Porter, Voters OK Plan for Two New Schools, The News Journal, Dec. 9, 2009, at B1. Molly Murray, Indian River Voters Approve Referendum, The News Journal, Mar. 31, 2010, at B1, B3. James Merriweather, Capital Voters OK School Bond Issue, The News Journal, Apr. 1, 2010, at B1-2. Dan Shortridge, Laurel Sets New Date for Schools Vote, The News Journal Kent & Sussex, Feb. 26, 2010, at B3; Dan Shortridge, Laurel Voters Say ‘No’ to New School Facilities, The News Journal, Apr. 1, 2010, at B1. Dan Shortridge, Seaford Voters OK School Referendum, The News Journal, Apr. 14, 2010, at B1. Edward L. Kenney, Christina Voters Narrowly Pass Referendum, The News Journal, May 26, 2010, at B12. Dan Shortridge, Laurel Rolls Out Smaller ‘Plan B’ Referendum, The News Journal, Aug. 26, 2010, at B3; Dan Shortridge, Laurel Schools Project Grinds Out Passing Vote, The News Journal, Oct. 5, 2010, at B2. Dan Shortridge, New High School Approved, The News Journal, Mar. 30, 2011, at B1, B3. Date May 10, 2011 District Colonial Type Transfer Result Defeated, 319 to 239 May 18, 2011 Seaford Capital Approved, 404 to 331 Mar. 28, 2012 Red Clay Capital Capital Approved, 5,398 to 4,552 Approved, 6,675 to 3,494 Mar. 28, 2012 Brandywine Operating Approved, 4,814 to 4,126 Dec. 4, 2012 Lake Forest Operating Approved, 753 to 734 Jan. 29, 2013 Indian River Feb. 28, 2013 Feb. 28, 2013 Colonial Appoquinimink May 9, 2013 Appoquinimink Capital Operating Operating Operating Operating Operating Approved, 2,695 to 1,239 Approved, 2,588 to 1,341 Defeated, 2,484 to 925 Defeated, 4,633 to 3,301 Defeated, 5,008 to 2,668 Approved, 4,637 to 3,023 June 4, 2013 Colonial Operating Approved, 3,005 to 2,938 Feb. 22, 2014 Smyrna Capital Operating Approved, 930 to 571 Approved, 958 to 569 Source Ira Porter, District Voters Fill Board Seats Throughout State, The News Journal, May 11, 2011, at B1. Notice of Special Election to Authorize the District to Issue Bonds to Fund School Capital/ Renovation, The News Journal, Apr. 25, 2011, at C9; Seaford Voters OK Tax Increase, The News Journal, May 20, 2011, at B2. Matthew Albright & Saranac Hale Spencer, School Votes Splinter Citizens, The News Journal, Mar. 20, 2016, at 1A, 8A. Terri Sanginiti, Voters OK School-Tax Hike, The News Journal, Mar. 29, 2012, at B1-2. Robin Brown, Lake Forest Residents OK $1.2M Referendum, The News Journal, Dec. 5, 2012, at B2. James Fisher, Voters OK New Spending, The News Journal, Jan. 30, 2013, at B1. Nichole Dobo, Referendums Defeated, The News Journal, Mar. 1, 2013, at B1-2. Appoquinimink to Have Second Referendum, The News Journal, Mar. 28, 2013, at B1; Matthew Albright & Saranac Hale Spencer, School Votes Splinter Citizens, The News Journal, May 20, 2016, at 1A, 8A. Matthew Albright, Colonial Tax Hike Passes By Slim Margin, The News Journal, June 5, 2013, at A1. Notice of Special Election in the Smyrna School District, The News Journal, Jan. 23, 2014, at D3; Matthew Albright & Saranac Hale Spencer, School Votes Splinter Citizens, The News Journal, May 20, 2016, at 1A, 8A. Date Feb. 27, 2014 District Seaford Type Operating Result Not available Mar. 27, 2014 Milford Capital Operating Defeated, 1,020 to 842 Defeated, 1,074 to 769 Apr. 2, 2014 Cape Henlopen Capital Approved, 3,597 to 2,410 May 28, 2014 Lake Forest Capital Approved, 622 to 363 Feb. 24, 2015 Feb. 24, 2015 Red Clay Christina Mar. 10, 2015 Delmar Operating Operating Operating Operating Capital Approved, 6,395 to 5,515 Defeated, 6,076 to 2,119 Defeated, 6,348 to 1,826 Approved, 130 to 78 Approved, 139 to 67 May 5, 2015 Milford Capital Defeated, 2,084 to 1,777 May 27, 2015 Christina Operating Defeated, 5,968 to 5,074 Oct. 6, 2015 Oct. 17, 2015 Milford Caesar Rodney Operating Operating Approved, 1,621 to 1,282 Approved, 1,939 to 1,171 Mar. 23, 2016 Cape Henlopen Capital Approved, 2,947 to 1,031 Source Matthew Albright, Seaford Seeks Tax Increase, The News Journal, Feb. 25, 2014, at B1. Matthew Albright & Saranac Hale Spencer, School Votes Splinter Citizens, The News Journal, May 20, 2016, at 1A, 8A. Matthew Albright, Cape Henlopen Voters OK Tax Hike for Schools, The News Journal, Apr. 3, 2014, at A4. Matthew Albright & Saranac Hale Spencer, School Votes Splinter Citizens, The News Journal, May 20, 2016, at 1A, 8A. Matthew Albright, Red Clay Says Yes; Christina Says No, The News Journal, Feb. 25, 2015, at A1, A4. Matthew Albright & Saranac Hale Spencer, School Votes Splinter Citizens, The News Journal, May 20, 2016, at 1A, 8A. James Fisher, Milford Hoping Voters Re-Assess Referendum, The News Journal, May 4, 2015, at 3A; Matthew Albright & Saranac Hale Spencer, School Votes Splinter Citizens, The News Journal, May 20, 2016, at 1A, 8A. Matthew Albright, Christina Voters Reject Tax Increase; Budget Cuts, Layoffs Likely Ahead, The News Journal, May 28, 2015, at 1A. Matthew Albright & Saranac Hale Spencer, School Votes Splinter Citizens, The News Journal, May 20, 2016, at 1A, 8A. Jon Bleiweis, ‘Our Time Is Now’: Cape Sets Referendum Date, The News Journal, Jan. 24, 2016, at 8A; Saranac Hale Spencer, Two Tax Hikes OK’d; 1 Fails, The News Journal, Mar. 24, 2016, at 1A, 5A. Date Mar. 23, 2016 Mar. 23, 2016 District Christina Brandywine Type Operating Operating Result Approved, 6,770 to 6,625 Defeated, 3,892 to 3,729 May 17, 2016 Brandywine Operating Approved, 9,500+ to 5,780 Nov. 22, 2016 Indian River Operating Defeated, 3,351 to 3,321 Dec. 20, 2016 Appoquinimink Feb. 28, 2017 Colonial Operating Capital Capital Operating Approved, 5,152 to 2,496 Approved, 5,506 to 2,102 Defeated, 2,733 to 2,193 Defeated, 2,961 to 2,067 Mar. 2, 2017 Indian River Capital Approved, 7,095 to 5,394 Source Saranac Hale Spencer, Christina Makes Its Third Attempt at Referendum, The News Journal, Mar. 8, 2016, at 1A, 4A; Saranac Hale Spencer, Two Tax Hikes OK’d; 1 Fails, The News Journal, Mar. 24, 2016, at 1A, 5A. Saranac Hale Spencer, Brandywine Referendum Passes in Landslide, The News Journal, May 18, 2016, at 2A. Gray Hughes, Indian River Referendum Reaction Is Mixed, The News Journal, Nov. 3, 2016, at 7A; Gray Hughes, Referendum Fails By Just 30 Votes, The News Journal, Nov. 23, 2016, at 2A. Jerry Smith, $268M Referendum Easily Passes, The News Journal, Dec. 21, 2016, at 1A, 5A. Jessica Bies, Colonial School District Referendum Fails, Delaware Online (Feb. 28, 2017, 9:34 PM), http://www.delawareonline.com/story/news/education/ 2017/02/28/colonial-school-district-referendum-fails/ 98508124/. Gary Hughes, Indian River Referendum Passes by Almost 2,000 Votes, Delmarva Now (Mar. 2, 2017 8:35 PM), http://www.delmarvanow.com/story/news/ local/delaware/2017/03/02/indian-river-referendumpasses-more-than-1k-votes/98592544/. Appendix B Abbreviation in Decision School Name A.I. duPont Alexis I. duPont High School N/A Alexis I. duPont Middle School Baltz Austin D. Baltz Elementary School Brandywine Springs Brandywine Springs Elementary School Cab Calloway Cab Calloway School of the Arts Central The Central School Conrad Conrad Schools of Science Dickinson John Dickinson High School Forest Oak Forest Oak Elementary School H.B. duPont Henry B. duPont Middle School Heritage Heritage Elementary School Highlands Highlands Elementary School N/A William C. Lewis Dual Language Elementary School Linden Hill Linden Hill Elementary School Marbrook Marbrook Elementary School McKean Thomas McKean High School Mote Anna P. Mote Elementary School North Star North Star Elementary School Richardson Park Richardson Park Elementary School Richey Donald J. Richey Elementary School Shortlidge Evan G. Shortlidge Academy Skyline Skyline Middle School Stanton Stanton Middle School Warner Warner Elementary School