Every year, thousands of governments, school districts, and special districts ask their voters to approve a ballot measure. Our team are experienced experts at designing and managing ballot measure survey research projects. This article explains everything you should include in a professional ballot measure survey. We hope this information will help your organization achieve voter approval on Election Day.
Determining the appropriate respondents for a ballot measure survey is not a straightforward matter. While registered voters are typically the ones who ultimately decide the outcome of such measures, not all registered voters actually cast their ballots. Thus, it is advisable to limit the survey sample to likely voters. Occasionally, ballot measure surveys may also target adult residents who may not be registered voters, potentially out of concern that a voter-exclusive survey could alienate this demographic. However, this issue can be addressed in the survey questionnaire itself.
When conducting a ballot measure survey, there are various data collection methods to consider. Some methods can be expensive or time-consuming, like mail surveys. To address this, our team has adopted a dual-mode approach that utilizes both MMS text invitations and telephone interviews.
MMS text invitations have proven effective in reaching up to 50% of the registered or resident population, increasing opportunities for respondents to participate and resulting in a more statistically accurate survey. Additionally, telephone interviews provide a way to target senior households, who may be less likely to support tax increases, and offer them an alternative way to participate. Here is a list of different data collection methods:
Online surveys: Surveys that are conducted through the Internet and can be accessed via email or social media.
Telephone surveys: Surveys that are conducted over the phone, either through automated systems or with live operators.
In-person surveys: Surveys that are conducted in person, either through door-to-door visits or at public locations such as malls or events.
Mail surveys: Surveys that are sent through postal mail and completed by the respondent before being returned.
Panel surveys: A group of incentivized online respondents are asked to participate in the survey.
Mixed-mode surveys: Surveys that use a combination of two or more methods, such as online and phone or mail and in-person.
Social media surveys: Surveys that are conducted on social media platforms, such as Twitter or Facebook.
Each data collection method has its own advantages and disadvantages, and the choice of method depends on the research question, target population, and available resources.
To ensure that a survey accurately reflects the demographics of a population, the pollster must have a clear understanding of that population’s demographics. This is particularly important when surveying registered voters, as the pollster must obtain a database of registered voters from sources such as the local county clerk or the Secretary of State Office.
Some polling firms, including Magellan Strategies, maintain a database of registered voters to facilitate survey projects and have ready access to demographic information. For instance, we have compiled a report outlining the voter registration and past-election turnout demographics for Broomfield, Colorado. Armed with this information, our Colorado ballot measure surveys can provide a demographically representative sample for presidential, mid-term, or odd-year election cycles.
Determining the demographics of a survey targeting residents is more complex than doing so for a survey targeting registered voters. The primary reason is that state, local, and federal governments do not regularly conduct population counts. Instead, the U.S. Census Bureau only conducts a census of the entire country once every ten years. However, the Bureau also conducts an annual American Community Survey data, which enables them to estimate population figures on a yearly basis.
These population estimates serve as the basis for producing “best-estimate” local population data by state and local governments. They also allow pollsters to determine the demographics of an adult population. For example, the following table provides the Census Bureau’s 2021 population estimates for the City and County of Broomfield, Colorado.
A ballot measure is unlikely to garner voter support if an organization is perceived to be providing inadequate services or not using taxpayer funds efficiently. As a result, a ballot measure survey should include the following questions:
Do you approve or disapprove of the job Fairfax County does managing street and road maintenance and repair?
Do you approve or disapprove of the job the Martinsville School District does doing educating students?
Is the Town of Jefferson fiscally responsible, and do they spend taxpayer money wisely?
Entities such as governments, school districts, and special districts must be able to account for every dollar in their budgets to instill voter confidence when asking for additional funding. While voters generally dislike tax increases for any purpose, they may be more likely to approve a tax increase if they believe the requesting entity demonstrates fiscal responsibility. Furthermore, voters want detailed information about how new tax revenue will be allocated.
In particular, organizations that are transparent about their finances are more likely to gain voter support. For example, informing voters about road maintenance costs or publishing teacher salaries by school districts is more likely to achieve voter approval. Conversely, organizations that are unable to accurately estimate the costs of their proposed services or projects are more likely to have their ballot measures rejected. Finally, when proposing a property tax increase through a ballot measure, it is crucial to inform homeowners about the potential increase in their tax bill. Such information can improve voter support for the measure by up to four percentage points.
Organizations that are specific and transparent about their finances are more likely to win voter support. For instance, providing information about the costs associated with road maintenance or publishing teacher salaries by school districts can increase the likelihood of voter approval. Conversely, organizations that are unable to accurately estimate the costs of their proposed projects or services may be more likely to have their ballot measures rejected. Finally, when proposing a property tax increase through a ballot measure, it is important to inform homeowners about the potential increase in their tax bill. Such information can improve voter support for the measure by up to four percentage points.
The likelihood of voters supporting a ballot measure is greater when they perceive that the service, issue, or problem being funded is a priority. Examples include road maintenance and repair, hiring additional firefighters and engines to reduce response times, increasing pay for bus drivers, building security upgrades for schools, and adding staff to expand library hours.
However, we have encountered instances where voters rejected a ballot measure because they did not view the need as a top priority. For example, a road maintenance ballot measure in the City of Gunnison in 2022, a school building infrastructure repair ballot measure in Moffat County School District RE-1, and several library districts on the Western Slope of Colorado.
To address this, a ballot measure survey should ask respondents their opinion of the service or need being addressed. For instance, in a road funding ballot measure, ask about respondents’ views on the conditions of the roads in their neighborhood. Similarly, for a housing affordability ballot measure, ask about respondents’ perceptions of the affordability of housing in their community.
Finally, some ballot measure surveys should include questions to ascertain the importance of a service, need, or amenity to a community. We refer to this as the “What is in it for me test.” If voters do not perceive a personal benefit or value in the service, they are less likely to support the ballot measure.
It is a recommended practice to communicate to survey respondents that no definitive decisions have been made to present a ballot measure to voters. This information can provide organizational leadership and elected officials with some flexibility in their decision-making process. Additionally, it is important to let respondents know that their opinions and feedback will have an impact on that decision. This message can help to alleviate public scrutiny and concerns about the organization’s motivations for conducting the survey.
Once an organization’s image rating and job approval have been measured, the next step is to present the ballot measure details, which can be done in two ways.
Firstly, if the ballot language has been finalized, it should be presented to voters as it is. Attempting to paraphrase the language can be confusing for voters, as they will be reading the actual ballot language when they vote.
Secondly, a paraphrased version is necessary if the final ballot language will be written using the survey results. This version should include information that will be in the ballot language, such as the amount of the property or sales tax increase and an estimated amount of funding raised by the tax. A description of the services, amenities, or building projects to be funded by the tax should also be included.
It is important to note that the “uninformed” ballot test question should not contain persuasive language. The paraphrased version should only include the facts that will appear in the ballot measure language, without any added emphasis on the need for the community or amenity to be addressed.
Voters require information to make an informed decision about whether to approve or reject a ballot measure. Firstly, they want to know precisely how the new funding will be allocated. Secondly, they want to understand how much the tax increase will impact their finances. Lastly, they want to know if there are any other options to fund the budget without resorting to a tax increase.
While some of the answers to these questions may be included in the ballot language, more often than not, they are not. As such, organizations should use their campaigns to inform and persuade voters. The survey should include a section dedicated to providing information that will educate respondents about the facts that they will encounter in the campaign.
This information section will allow organizations to tell their story and explain why the ballot measure is being considered. They can also describe how the budget has been stretched thin, the rising costs of providing services, or the impact of inflation on salaries.
In addition, the survey should allow for the ranking of information that builds support for the ballot measure. It can also test different ways of presenting the information using alternative words and phrases. Ultimately, the information section must provide detailed information on precisely how the new funding will be allocated.
Once the organization’s story of why the ballot measure is being considered is conveyed, the survey should include a second ballot measure question, the “informed” ballot measure question. This question should be worded identically to the “uninformed” ballot measure question to ensure an accurate comparison of responses. However, this time respondents are fully informed of the ballot measure’s details, allowing for an accurate reflection of their opinion.
To fully understand voter perspectives on a ballot measure, it’s essential to include open-ended questions in the survey. These questions should ask respondents to provide reasons why they either support or oppose the measure. It’s best to ask these questions after the informed ballot test question, as respondents will have a better understanding of the details of the measure and can provide more relevant responses. These open-ended questions can provide valuable insights into the reasons behind voter opinions and can help organizations tailor their messaging and outreach efforts.
The last part of the survey should include questions that allow voters to identify themselves by their gender, age, and political party affiliation. These details are necessary for the pollster to weigh the survey data accurately and ensure that the results represent the demographic makeup of the population being surveyed. In addition, other common demographic questions such as education level, household income, marital status, race, and ethnicity can also be included.