Examining Elections Past

The secret to passing your tax or bond measure may be found in the data of earlier votes by Don E. Lifto and J. Bradford Senden

In promoting its primetime hit “Crossing Jordan,” NBC touts costar Jill Hennessy as “a sexy, smart and fearless Boston medical examiner with a penchant for going beyond the call of duty to investigate crimes.” A growing fan base of forensic groupies tune in weekly to watch Jordan “channel her inner anger toward piecing together complex murder cases that have been hidden, shoved aside—or conveniently forgotten,” as the network puts it.

The clues, of course, first must be coaxed from the dead and reassembled to understand what actually happened.

Albeit not as sexy, school leaders also must be smart and fearless as they conduct their forensic studies—in this context from the examination table of school elections past.

Research and practice have yet to yield a modus operandi in K-12 education that always produces winners on school bond and tax measures. Whether it’s bricks and mortar or requests for more operating money, each election type and context are unique with no guarantee that a set of campaign strategies—even if successful in one district—won’t fail in your community. If successful campaigns were not such a delicate balance of science and art, the formula for success would have long since been discovered, resulting in significantly more school districts finding success at the polls. Notwithstanding this reality, both research and successful practice suggest the best way to start planning your next successful bond or operating levy is to take a much closer look at your last.

Rearview Mirror

Most school districts squander a key strategic opportunity when they fail to collect, analyze and archive valuable data after school finance elections—equally important following successful or losing campaigns.


The most obvious data analysis, although seldom done well, is to understand exactly who participated in a recent finance election as compared to earlier events. How did the campaign effort in support of the ballot question influence the electorate? How did various demographic groups vote relative to their proportionate share of the voter file and past voting habits? To what extent did targeted supporters show up from various precincts or attendance areas? These are examples of the questions that can be probed in a post-election analysis, yielding critical information for school leaders planning future campaigns.

The idea that a school district should look back at its last election when planning the next one is based on the concept that one of the best predictors of future voter behavior are the past actions of those voters. Therefore, understanding the key factors and behaviors in past local elections (for example, older voters, female voters or voters living in a certain area) will help a district plan a successful election.

Voter Files

To complete a post-election analysis, several steps need to be taken. First the district should acquire an electronic listing of all of the district’s registered voters. Even in states with same-day registration, an understanding of who might vote in an upcoming election begins with a determination of who voted in the past.


The simplest way to do this is to identify the five most recent elections conducted within the school district. Voters who have cast ballots in at least four of these elections are assigned to the “Very Active Voter” category, those who have voted two or three times to the “Active Voter” category, and all other voters to the “Less Active” or “New Voter” category. This allows one to gain a historical perspective of the voting patterns unique to different types of elections helping the district target campaign efforts in similar elections in the future.

Voter files vary in the amount of additional information they provide. Generally, a voter file yields information about each voter’s gender, age and date of registration. These facts allow a district to divide voters into various groups. For example, registration date is useful because it serves as a good proxy for length of residence in a community. Data from an elections office can be expanded to include geographic information, census data and, when possible, ethnic coding in the voter file.

Before beginning a post-election evaluation, two important pieces of information should be added to the voter file. Parents should be identified in the file by using a list of those who resided in the school district at the time of the election. After all, these are the parents who could have participated in the election.

In addition, every individual identified as supporting the most recent district proposal should be marked in the voter file. This information is extremely critical and, far too often, is poorly stored or discarded soon after the polls have closed. Every election should archive an electronic file of its identified supporters containing sufficient information to enable it to be easily linked to a new copy of the voter file for a future election campaign.

Assessing Turnouts

At this point, the voter file has been fully annotated and the population that voted in the election to be analyzed can be isolated. Once this population is isolated, all of its demographic characteristics are counted and compared to the population of all voters in the district.


These counts allow the district to see, for example, if male or female voters are over- or underrepresented in the population that cast ballots. If women make up 52 percent of all voters in the district but research finds that women made up 56 percent of the population that voted, the district knows that its election attracted more women than men. Other counts enable the district to refine this view. Were these older or younger women? Of the male voters in the district, which age or registration groups failed to get to the polls?

The most important counts look at the parents and the identified campaign supporters. Each group needs to be examined in two ways. If parents represent 23 percent of the population of registered voters, are they over- or underrepresented in the Election Day voting population? If parents are underrepresented in the population that voted, something was wrong in the planning or execution of the campaign or the proposal was out of alignment as parents are the only group of voters that a school tax election can depend on for support.

Even when the parents are overrepresented in the population that voted, the district needs to look at the degree to which the potential vote within the parent population was maximized. If parents represent 23 percent of the voter file, but comprise 34 percent of the population that voted, an initial conclusion could be that parents participated strongly in the election. But if only 43 percent of all registered parents actually cast a ballot, the initial conclusion needs modification. Leaving that many parents home on Election Day means the campaign in support of the district’s proposal failed to maximize impact from the parent vote. Such a situation can cost the district on Election Day.

The same applies to voters identified by the campaign as supporters. Many techniques can be used to develop a list of tax supporters, but all rely on campaign volunteers asking community members if they will support the district’s tax proposal. After Election Day, look at how well this population performed. Are identified supporters over- or underrepresented in the population that voted? What percentage of the population of supporters cast ballots?

Even if supporters are overrepresented in the election as compared to their proportion of the voter file, the campaign may not be satisfied with the actual percentage casting ballots. In our experience, we have seen results ranging from a dismal 33 percent of supporters participating up to an outstanding 95 percent of identified supporters casting ballots.

Finally, a post-election analysis should examine how the election being evaluated compares to other recent elections. In most communities, elections develop specific characteristics. A November election to elect a governor or the president will attract a much younger, more male population than a primary or school board election. The latter generally attracts a population that is older, more female and more likely to vote in any election held in the area. Comparing the demographic characteristics of the election a district just held to others held during the last few years will help in planning the district’s next successful election.

Planning Ahead

According to Merriam Webster’s Collegiate Dictionary, an autopsy is a “critical examination, evaluation or assessment of someone or something past.” The purpose of looking back, of course, is to determine what happened and why.


The foundation for success at the ballot box begins with a similar examination of what happened during prior elections conducted within the district. Like the medical examiners in “Crossing Jordan,” successful school leaders need to piece together disconnected clues about voter behavior, demographic tendencies and turnout under various conditions.

Reconstructing this forensic puzzle, in combination with other steps in a comprehensive planning model, provides the foundation for an effective campaign and guides the critical steps of voter identification and a get-out-the-vote effort—key ingredients to achieving success in your next finance election.

Don Lifto, a consultant on school finance, is superintendent of Northeast Metro 916 Intermediate School District, 3300 Century Ave. North, White Bear Lake, MN 55110. E-mail: dlifto@nemetro.k12.mn.us. Bradford Senden is the managing partner at the Center for Community Opinion in San Ramon, Calif. They are co-authors of School Finance Elections: A Comprehensive Model for Success (Scarecrow Press).