Ballot measure readability scores, 2025
Ballotpedia's readability report analyzes what level of education voters would need to understand the ballot titles and summaries of statewide ballot measures using Flesch Reading Ease (FRE) and Flesch-Kincaid Grade Level (FKGL). A readability score is an estimation of the reading difficulty of a text. Measurements used in calculating readability scores include the number of syllables, words, and sentences in a text. Other factors, such as the complexity of an idea in a text, are not reflected in readability scores.
For 2025, 30 statewide ballot measures were certified for the ballot in nine states: California, Colorado, Louisiana, Maine, New York, Ohio, Texas, Washington, and Wisconsin.
- On November 4, voters in six states will decide on 24 statewide ballot measures.
- Earlier in 2025, voters in three states—Louisiana, Ohio, and Wisconsin—decided on six ballot measures. Two were approved, and four were defeated.
Readability index details
Ballotpedia uses two formulas, the Flesch Reading Ease (FRE) and Flesch-Kincaid Grade Level (FKGL), to compute scores for the titles and summaries of ballot measures. The FRE formula produces a score between a negative (-) number and 100, with the highest score (100) representing a 5th-grade equivalent reading level and scores at or below zero representing college graduate-equivalent reading level. Therefore, the higher the score, the easier the text is to read. The FKGL formula produces a score equivalent to the estimated number of years of U.S. education required to understand a text. A score of five estimates that a U.S. 5th grade student would be able to read and comprehend a text, while a score of 20 estimates that a person with 20 years of U.S. formal education would be able to read and comprehend a text. Ballotpedia uses Readable.com to calculate the scores.
- Learn more about these formulas in the formulas section below.
Analysis
Measures
In 2025, 30 ballot measures were certified for elections. The following individual ballot measures had the highest or lowest scores or word counts:
- The ballot measure with the highest ballot title grade level, at 39, was Louisiana Amendment 2, which proposed several changes to tax and budget policies. While no formal education level, including post-graduate, corresponds to a 39th-grade level, this indicates that the language structure is advanced or difficult, based on the FKGL.
- The ballot measure with the lowest ballot title grade level, at 8, was New York Proposal 1, a constitutional amendment addressing the Mount Van Hoevenberg Complex and Adirondack Park. Based on FKGL, this is equivalent to an 8th-grade reading level.
- The longest ballot title, with 298 words, was for Ohio Issue 2, which allowed the state to issue up to $2.5 billion in general obligation bonds to assist local governments in funding public infrastructure improvement projects.
- The shortest ballot title, with 13 words, was for Texas Proposition 16, a constitutional amendment to provide that "persons who are not citizens of the United States" cannot vote in Texas.
States
The following table provides the average readability grade levels for ballot titles of each state. The lowest average score was eight in New York, which had one measure on the ballot, and the highest average score was 26 in Maine, which had two measures on the ballot.
| State | Average Title Grade | Average Title Words | Measures |
|---|---|---|---|
| California | 20 | 16 | 1 |
| Colorado | 24 | 128 | 2 |
| Louisiana | 23 | 44 | 4 |
| Maine | 26 | 52 | 2 |
| New York | 8 | 64 | 1 |
| Ohio | 15 | 298 | 1 |
| Texas | 21 | 33 | 17 |
| Washington | 12 | 54 | 1 |
| Wisconsin | 22 | 38 | 1 |
Author
Since ballot measures often consist of multiple pages of legal text, each state assigns a person or group to write a shorter title for voters to read on the ballot. The responsibility for drafting this language varies by state. In some states, different offices or individuals draft the language for different measures. Additionally, some states require collaboration in the drafting process. For instance, the secretary of state may write the ballot language, but it might need to be approved by the attorney general.
The type of author with the lowest average title grade was state board, with a score of 12, while secretaries of state had the highest average title grade, with a score of 26.
| Author | Average Title Grade | Average Title Words | Measures |
|---|---|---|---|
| Attorney general | 16 | 35 | 2 |
| Secretary of state | 26 | 52 | 2 |
| State board | 12 | 181 | 2 |
| State legislature | 21 | 43 | 24 |
Type
Measures can be citizen-initiated (initiatives) or referred (referrals) by the legislature, a commission, or an automatic constitutional process. Measures can also be constitutional amendments or statutes, including statutes for non-binding measures and general obligation bonds.
| Type | Average Title Grade | Average Title Words | Measures |
|---|---|---|---|
| Initiative | 26 | 52 | 2 |
| Referral | 20 | 52 | 28 |
| Constitutional | 20 | 47 | 26 |
| Statutory | 25 | 90 | 4 |
Historical comparison
Since 2017, Ballotpedia has tracked the readability scores of statewide ballot measures. The following table shows the average ballot title grade, word counts, summary grade, and the number of measures for each year. The year with the lowest ballot title grade was 2019, with 15 years of education, and the year with the highest was 2025, with 21 years of education.
| Year | Average Title Grade | Average Title Words | Average Summary Grade | Average Summary Count | Measures |
|---|---|---|---|---|---|
| 2025 | 21 | 52 | 20 | 70 | 30 |
| 2024 | 16 | 68 | 16 | 125 | 159 |
| 2023 | 19 | 73 | 22 | 246 | 41 |
| 2022 | 19 | 66 | 18 | 130 | 140 |
| 2021 | 18 | 53 | 14 | 289 | 39 |
| 2020 | 17 | 60 | 14 | 187 | 129 |
| 2019 | 15 | 41 | 15 | 104 | 32 |
| 2018 | 20 | 66 | 16 | 210 | 167 |
| 2017 | 20 | 58 | 18 | 240 | 27 |
Graph
The following bar charts illustrate the average grade level for ballot titles and summaries from 2017 to 2025.
2025 readability scores
Below is a list of measures certified for ballots in 2025. The readability scores will be published closer to the election.
Educational attainment in the U.S.
In 2022, the U.S. Census Bureau reported on the highest level of education of the population age 25 and older in the United States. Of that population, 14% completed advanced education such as a master’s degree, professional degree, or doctorate. The chart below shows the breakdown by highest level of education.[1]
Formulas
The Flesch Reading Ease and Flesch-Kincaid Grade Level formulas use the same variables and are inversely correlated, meaning that as one increases the other decreases.
Flesch Reading Ease
In the 1940s, Rudolf Flesch developed the Flesch Reading Ease (FRE) test. The U.S. Department of Defense uses the FRE to help craft its documents and manuals.[2] The FRE computes a score based on the number of syllables, the number of words, and the number of sentences in a text. The FRE formula is as follows:[3]
The FRE formula was designed to produce a score between 0 and 100, with the highest score (100) representing a 5th-grade equivalent reading level and the lowest score (0) representing college graduate-equivalent reading level. However, a score can be negative, representing increased difficulty. Therefore, the higher the score, the easier the text is to read. Rudolf Flesch created the following guide to interpreting FRE scores:[3]
| Score | School level |
|---|---|
| 90 to 100 | 5th grade |
| 80 to 90 | 6th grade |
| 70 to 80 | 7th grade |
| 60 to 70 | 8th and 9th grade |
| 50 to 60 | 10th to 12th grade |
| 30 to 50 | College |
| 0 to 30 | College graduate |
Flesch-Kincaid Grade Level
In 1975, J. Peter Kincaid recalculated FRE to give a score in the form of a U.S. school grade level for use by the U.S. Navy. This new formula became known as the Flesch-Kincaid Grade Level (FKGL) test. Like FRE, the FKGL computes a score based on the number of syllables, the number of words, and the number of sentences in a text. The FKGL formula is as follows:[4]
The FKGL produces a score equivalent to the estimated number of years of education required to understand a text. A score of 9 estimates that a U.S. 9th grade student would be able to read and comprehend a text, while a score of 18 estimates that a person with 18 years of U.S. formal education would be able to read and comprehend a text.[5]
Limitations
As the FRE and FKGL, along with other readability tests, do not measure the difficulty or complexity of the ideas expressed in ballot measure titles and summaries, they may underestimate or overestimate the ability of voters to comprehend a text. Political scientist Shauna Reilly, who utilizes readability indices in her research, noted their limitations, stating:[2]
| “ | There are limitations to the value of these measurements. No mathematical formula can tell us how complex the ideas of the passage are nor whether the content is in a logical order. Further, these mathematical equations exist in a vacuum and cannot explain the context of the passage.[6] | ” |
Prior research
Ballot Question Readability and Roll-off: The Impact of Language Complexity
In 2011, political scientists Shauna Reilly and Sean Richey published an article in Political Research Quarterly on research they conducted to answer the question of whether the difficulty or complexity of ballot measure language correlated with voters skipping voting on a ballot measure. The authors referred to voters casting ballots but skipping a ballot measure as voter roll-off. To measure the difficulty or complexity of ballot measure language, Reilly and Richey calculated Flesch-Kincaid Grade Level scores for 1,211 statewide ballot measures from 1997 to 2007. Reilly and Richey concluded that lower readability scores correlated with higher rates of voter roll-off. In their model accounting for state and year variations, Reilly and Richey only found one variable with a stronger influence on voter roll-off than readability—whether or not a ballot measure was on a primary election ballot compared to a special election ballot.[5]
Reilly and Richey calculated the mean Flesch-Kincaid Grade Level score for each state, except Arkansas, Illinois, and West Virginia, with at least one ballot measure during the 10-year period from 1997 to 2007. The state with the highest mean score was New Mexico, which had a mean FKGL score of 28 years of education. The state with the lowest mean score was Oklahoma, which had a mean FKGL score of nine years of education. The following table is from Reilly and Richey's research and contains the number of ballot measures analyzed in each state, the mean, minimum, and maximum readability score of measures in each state, and the standard deviation of the readability scores for measures in each state:[5][7]
| State | Measures | Mean | Mean U.S. equivalent | Standard deviation[7] | Minimum | Maximum |
|---|---|---|---|---|---|---|
| Oklahoma | 38 | 9 | High school | 1.1 | 7 | 12 |
| Connecticut | 1 | 11 | High school | 0 | 11 | 11 |
| North Carolina | 1 | 11 | High school | 0 | 11 | 11 |
| South Dakota | 36 | 12 | High school | 2.1 | 7 | 17 |
| Alaska | 30 | 13 | Associate's degree | 5.3 | 8 | 30 |
| California | 105 | 13 | Associate's degree | 1.8 | 9 | 18 |
| North Dakota | 13 | 13 | Associate's degree | 2.8 | 9 | 18 |
| Idaho | 16 | 14 | Associate's degree | 2.3 | 12 | 20 |
| Iowa | 5 | 14 | Associate's degree | 4 | 11 | 21 |
| Massachusetts | 18 | 14 | Associate's degree | 2.1 | 10 | 19 |
| Michigan | 18 | 14 | Associate's degree | 3.1 | 9 | 21 |
| Mississippi | 3 | 14 | Associate's degree | 5 | 8 | 18 |
| Oregon | 94 | 14 | Associate's degree | 1.7 | 11 | 18 |
| Rhode Island | 35 | 14 | Associate's degree | 6.1 | 6 | 33 |
| Washington | 57 | 15 | Bachelor's degree | 2.8 | 10 | 22 |
| Montana | 29 | 16 | Bachelor's degree | 7.4 | 11 | 52 |
| New Hampshire | 8 | 16 | Bachelor's degree | 5 | 10 | 27 |
| Utah | 6 | 16 | Bachelor's degree | 5.3 | 10 | 24 |
| Arizona | 70 | 17 | Master's degree | 3.1 | 11 | 26 |
| Florida | 40 | 17 | Master's degree | 5 | 8 | 38 |
| Indiana | 6 | 17 | Master's degree | 3.5 | 13 | 23 |
| Louisiana | 61 | 17 | Master's degree | 6.8 | 8 | 44 |
| Ohio | 19 | 17 | Master's degree | 4.9 | 9 | 30 |
| Tennessee | 6 | 17 | Master's degree | 5.8 | 10 | 25 |
| Vermont | 1 | 17 | Master's degree | 0 | 17 | 17 |
| Alabama | 32 | 18 | Master's degree | 6.4 | 12 | 35 |
| Kansas | 4 | 18 | Master's degree | 1.7 | 16 | 20 |
| Maine | 66 | 18 | Master's degree | 6.6 | 8 | 37 |
| Nebraska | 37 | 18 | Master's degree | 3.4 | 11 | 25 |
| Wyoming | 12 | 18 | Master's degree | 12 | 12 | 25 |
| Missouri | 27 | 19 | Ph.D. | 8.2 | 8 | 44 |
| Nevada | 36 | 19 | Ph.D. | 6.4 | 11 | 42 |
| New York | 8 | 19 | Ph.D. | 8.3 | 8 | 35 |
| Maryland | 11 | 20 | Ph.D. | 4.1 | 13 | 26 |
| Texas | 84 | 20 | Ph.D. | 12 | 12 | 45 |
| Wisconsin | 3 | 20 | Ph.D. | 16.6 | 17 | 23 |
| Georgia | 33 | 22 | Ph.D. | 10.4 | 10 | 57 |
| Hawaii | 10 | 22 | Ph.D. | 10.9 | 10 | 44 |
| Kentucky | 7 | 22 | Ph.D. | 6.1 | 14 | 30 |
| Virginia | 3 | 22 | Ph.D. | 3.2 | 19 | 25 |
| New Jersey | 20 | 23 | Ph.D. | 6.6 | 13 | 34 |
| Pennsylvania | 6 | 24 | Ph.D. | 5.4 | 17 | 33 |
| South Carolina | 19 | 25 | N/A | 10.8 | 16 | 63 |
| Minnesota | 1 | 26 | N/A | 0 | 26 | 26 |
| Colorado | 62 | 27 | N/A | 15.2 | 5 | 95 |
| New Mexico | 14 | 28 | N/A | 9.3 | 12 | 39 |
| Arkansas | N/A | N/A | N/A | N/A | N/A | N/A |
| Illinois | N/A | N/A | N/A | N/A | N/A | N/A |
| West Virginia | N/A | N/A | N/A | N/A | N/A | N/A |
See also
Past readability analyses:
- Ballot measure readability scores, 2018
- Ballot measure readability scores, 2017
- Ballot measure readability scores, 2019
- Ballot measure readability scores, 2020
- Ballot measure readability scores, 2021
- Ballot measure readability scores, 2022
- Ballot measure readability scores, 2023
- Ballot measure readability scores, 2024
External links
Additional reading
Footnotes
- ↑ U.S. Census Bureau, "Census Bureau Releases New Educational Attainment Data," February 16, 2023
- ↑ 2.0 2.1 Reilly, S. (2015). "Language Assistance under the Voting Rights Act: Are Voters Lost in Translation?" Lanham, MD: Lexington Books. (pages 55-56)
- ↑ 3.0 3.1 University of Canterbury, "How to Write Plain English," accessed April 19, 2017
- ↑ U.S. Naval Technical Training Command, "Derivation of new readability formulas (Automated Readability Index, Fog Count, and Flesch Reading Ease Formula) for Navy enlisted personnel," February 1975
- ↑ 5.0 5.1 5.2 Reilly, Shauna, and Sean Richey. "Ballot Question Readability and Roll-off: The Impact of Language Complexity." Political Research Quarterly 64, 1. (2011): 59-67.
- ↑ Note: This text is quoted verbatim from the original source. Any inconsistencies are attributable to the original source.
- ↑ 7.0 7.1 The standard deviation (SD) measures how spread out around the mean the scores of individual measures were. The smaller the standard deviation, the closer the scores of individual measures were to the mean. The larger the standard deviation, the farther apart the scores of individual measures were to the mean.
