Executive Summary

Nearly all New York school districts need permission from voters before undertaking major construction projects or equipment purchases.¹ These ballot questions are typically presented as part of board-member elections and the annual budget approval referendum held statewide on the third Tuesday in May.

State law, however, allows district officials to hold these votes effectively at any point in the year—even sometimes within a few weeks of regularly scheduled elections—at what’s known as a “special meeting.”² School districts have used more than 400 special meetings since 2011 to approve nearly $9 billion in spending, including about $2 billion last year. 

The use of special meetings appears to have exploded in 2022, as districts used them 87 times to present voters with $2.6 billion in proposed spending, winning voter approval on 85 out of 101 questions. Fifty-one special meetings were held in 2019. 

The data show these elections generally have lower voter participation than May school board and budget elections, which nearly always have lower turnout than November general elections. Turnout in some cases has been less than a quarter of the number of ballots cast in that year’s May election—which, statewide, this year averaged just 8 percent of eligible voters. 

Lower turnout also appears to increase the likelihood measures will pass: in cases where turnout exceeded May levels, a proposal was three times as likely to fail (30 percent versus 10 percent) as it was in cases where turnout stayed at or below. To be sure, many factors influence when a measure will be put before voters, but the lower turnout of special meetings has undoubtedly created an incentive for project proponents who, after all, are specifically seeking a green light. 

Special meetings are especially common in December and most of last year’s were conducted in that month. It appears state lawmakers could reduce the need for these special meetings by making it easier for districts to piggyback on the November general election ballot. 

Arbitrarily scheduled, low-turnout special votes are a disservice to the democratic process that state officials should restrict or eliminate. The votes now presented at special meetings should be held on the same date as budget votes and board elections, or on Election Day in November. 

To make this feasible, lawmakers should amend the Election Law to eliminate the obstacles that prevent school boards from accessing the Election Day ballot. The Legislature here can make a fair trade that results in greater certainty for both voters and school officials. 

Background 

By state law, almost all of New York’s 673 school districts need voter approval for major spending decisions, including setting their annual budgets and, among other things, undertaking significant construction projects and equipment purchases. 

The Legislature requires all districts to hold budget votes on the same Tuesday in May—a strategy for minimizing voter confusion and increasing turnout. (The exceptions are New York’s five most populous cities, New York City, Buffalo, Rochester, Yonkers and Syracuse). School districts routinely present propositions to voters on the same ballot as the budget referendum. 

But while the dates for budget approvals are fixed, school districts have the option to schedule votes, known as “special meetings,” to authorize additional spending and debt at any time during the year without any additional permission from state officials.  

Since 2011, school boards have presented voters with at least 480 propositions using at least 410 special meetings, that is, outside the regularly scheduled May or November elections. All told, voters have signed off on $8.9 billion in added spending using the special meeting process since that time. 

These are partial figures because the State Education Department does not monitor the use of special meetings, leaving the authors to scour news accounts, board minutes and other school records to recreate a record of these votes. 

It bears noting that state law treats school district votes, which are run by the school district, entirely separately from local government elections, which are run by local boards of elections. School districts are often unable to place questions on the November ballot on which federal, state and local government candidates appear, partially because school district boundaries are rarely coterminous with county-drawn election districts. This limits school officials’ options to holding special meeting or waiting up to 12 months for the next opportunity for a regularly scheduled school vote. 

Voters are accustomed to voting in May and November elections, but turnout at special meetings tends to be lower. That means project proponents must persuade fewer people to support their proposal, creating an incentive for scheduling votes at times when turnout will be lower. 

As the Middletown Times Herald-Record editorial page remarked in a 2014 editorial titled “Every day is election day in NY school districts”: 

You can’t blame schools for trying to improve their odds. But you can blame a state Legislature that could require that these elections take place in a more convenient fashion, coupled with other public decisions to ensure the maximum participation.³ 

As a practical matter, this gives school boards only one regular chance, in May, to present questions to voters. Faced with the state’s lengthy approval process for construction projects, and that districts often aim to complete construction projects during summer vacation, districts must choose between postponing a project by a full year or taking it to voters in a special meeting. 

At the same time, some districts may be using their powers in a manner that the Legislature did not intend. 

For instance, after Carmel school district voters in December 2017 rejected a pair of capital spending proposals, the board went back to them in October 2018—after the May vote and just weeks before the general election. That proposal was shot down too, and the district called its third special meeting in just 16 months. Voters approved the plan in March 2019—six weeks before the May vote. 

In Albany—where district officials can place questions on the November ballot—voters in November 2015 rejected a proposal for a new, amenity-laden high school with an eye-popping price-tag—$196 million.⁴ Undaunted, and rather than wait until May, officials scheduled a revote for February on a slightly modified proposal at which they, by their own admission, planned on voter turnout being considerably lower. A subsequent investigation by the district found the revote was plagued by irregularities, including polling places that didn’t open on time and a shortage of ballots in an area of the city where opposition had been especially strong three months prior. The plan passed narrowly.⁵

More Votes, Fewer Voters 

In 2022, districts presented voters with 101 questions at 87 special meetings held outside the May school votes or the November general election. Eighty-five measures were approved as voters signed off on, among other things, nearly $2 billion in spending. 

That’s considerably higher than the 51 special meetings held in 2019, the last year before the novel coronavirus pandemic. 

Most 2022 special meetings were held in December (figure 1). Each proposal put before voters is different, but the month’s popularity stems in part from superintendents looking to get projects approved before crafting district budgets in January and February.

Figure 1.

Source: Empire Center data

The State Education Department does not publish voter enrollment statistics, so district-level turnout cannot be readily calculated. However, the number of ballots cast in a district’s special meeting can be compared to the number cast in that year’s May school budget vote. 

Turnout in 2022 special meetings averaged 82 percent of May levels and ranged from just 9 percent to 330 percent. Looking at the low end of the scale, and considering May turnout was, statewide, just over 9 percent of eligible voters, it is likely there were multiple special meetings in which less than 5 percent of eligible voters cast ballots. 

Figure 2.

Source: Empire Center calculations, New York State Education Department

Fewer Voters Means Better Odds 

The number of ballots cast in each special meeting since 2011 averaged 93 percent of the number cast in that year’s May vote. However, that figure is skewed by a small number (23) of special meetings in which turnout at least doubled. 

Turnout was less than half of May levels in 25 percent (104) of the special meetings, and less than a quarter of May levels in 27 of them. 

Given how low May turnout has been in recent years, this means school districts are routinely hearing from less than 5 percent of voters on questions with significant fiscal ramifications for the community. 

The greatest difference in turnout between May and the special meeting was observed in Westchester’s Briarcliff Manor Union Free School District, where nearly six times as many people turned out for a December 2018 building renovation referendum than had voted on the budget seven months prior.⁶ Opponents cited declining enrollment and the $35 million cost of the project, which was defeated. 

At the other extreme, a February 2022 bus-purchase proposal in the North Country’s Ausable Valley schools passed overwhelmingly—among the 47 voters who turned out.⁷ The school budget vote three months later drew 524 people to the polls. 

The likelihood of success for a ballot proposition appears linked to turnout. Looking at all special meetings since 2011, when turnout was at or below May levels, more than 90 percent of proposals passed. However, that passage rate sank below 70 percent when turnout exceeded May levels. Put another way, the likelihood of defeat triples if more voters turn out for a special meeting. (Table 1)

Table 1. Special Meetings Since 2011 – Failure Rate (Question) and Turnout

Source: Empire Center calculations

Recommendations 

As a starting point, the State Education Department should scrutinize the practice and measure voter participation levels (as it does for school budget votes).  

Every facet of school district operations is ultimately controlled by the state Legislature, and the use of special meetings has previously faced legislative scrutiny. The state Senate each year between 2013 and 2017 passed legislation⁸ that would have required school districts to hold these measures on one of three uniform dates during the year, including the date of May school budget votes. However, bill sponsor Senator John Bonacic retired in 2018, and the measure has not been reintroduced. 

The Legislature should stop permitting special meetings. The opportunities and incentives for abuse are too large. 

Instead, the Legislature should make it easier for school districts to place questions on the November ballot by proscribing a process for cooperation between local boards of elections and school districts. Granting this access for schools will result in more people weighing in on ballot propositions and allow more participation—and greater confidence—in New York’s strong system of local control.

Endnotes

1. Education Law §416
2. Education Law §2007
3. “It’s always election day for NY school districts,” Times Herald-Record, 28 Jan 14. 
4. Bump, Bethany, “Albany rejects high school proposal,” Times Union, 12 Nov 15. timesunion.com/local/article/Ballot-counters-close-in-on-decision-in-Albany-6627156.php
5. “Counsel Vote Review Report – Preliminary Report,” Albany CSD, 25 Feb 16. 
6. Wilson, Colleen, ”Briarcliff voters reject $34 million bond for school upgrades,” Lohud, 14 Dec 18. lohud.com/story/news/education/2018/12/13/briarcliff-schools-bond-vote-2018/2289004002
7. “Minutes of the Special Meeting,” Ausable Valley CSD, 16 Feb 22. 
8. For example, Senate Bill No. 1002 of 2017

Data Notes 

Authors reviewed news reports, district websites, and district social media accounts. The analysis is based on special meetings for which the vote date, outcome and total number of votes cast could be readily found. An additional 68 special meetings were identified; however, vote totals were not readily available. Of those, 23 passed, 3 failed, and outcome information was not readily available for 42.