Pass the Freedom to Vote Act: How Elections Would Look Different This Year and in the Future

Since 2019, Congress has worked to pass key voting rights legislation that would improve election administration and security nationwide and guarantee easy access to the ballot box for all Americans.¹

This year will mark the third federal and second presidential election cycle administered without the pro-democracy policies in the Freedom to Vote Act (FTVA). The U.S. House of Representatives has delivered for the American people by passing the FTVA—transformative legislation that would strengthen voting rights, safeguard the electoral process, prohibit partisan gerrymandering, and curb dark money in politics—but the legislation has been blocked in the U.S. Senate by a minority of members through the use of the filibuster.² The blocking of the FTVA is just the most recent example in the long history of the filibuster being used to obstruct critical advances in civil and voting rights policies.³

The FTVA is needed now more than ever to strengthen U.S. elections and democracy, particularly at a time when Americans are increasingly losing faith in government and institutions, including Congress.⁴ Large-scale election and institutional reforms are necessary to improve trust in American democracy and make the people feel that their government is representative of their interests and responsive to their needs. While democracy and election experts, as well as progressive legislators across the country, agree that the FTVA is desperately needed to improve and modernize U.S. elections, conceptualizing how truly transformative this legislation would be for American democracy has been a challenge.

Using academic and expert research examining different voting policies and their impacts, this report provides analysis and statistical extrapolation on the ways that this year’s elections would look different if some of the FTVA’s key voting rights, voter registration, and election administration policies had been enacted prior to the 2024 presidential election cycle.

The statistical findings in this report should not be conflated or accumulated; rather, they should be viewed as the impacts individual policies could have on voting patterns and voter turnout. This report aims to demonstrate why this legislation is desperately needed and why it would make it so much easier for Americans to make their voices heard.

The Freedom to Vote Act would ensure that all eligible voters have access to automatic, same-day, and online voter registration. Many of these voter registration methods have been standard practice across the country for decades, but the FTVA would ensure all Americans have access to these methods and that their ability and ease in registering to vote is not determined by their ZIP code.⁶

Automatic voter registration (AVR) helps eligible citizens seamlessly and effortlessly register to vote and update their voter registration information through interactions with their department of motor vehicles (DMV).

The Freedom to Vote Act would ensure that AVR is enacted in states where it’s not already offered. Under the policy, information that eligible Americans provide to their DMV would be used to register them to vote or update their voter registration information, unless they affirmatively opted out. Since the 1990s, federal law has required DMVs to offer voter registration services, but the process often can be confusing and tedious for voters and requires additional work on their part.⁷ AVR essentially flips the process on its head and ensures a better integrated and more efficient system for voters, DMV workers, and election officials; it also helps keep voter rolls clean, accurate, and up to date.

Additionally, experts have supported the implementation of AVR in order to reduce racial and economic gaps in voter registration rates and subsequently in voter turnout.⁸ The policy in the FTVA would thereby not only be critical for generally enhancing voter participation and election administration but also for ensuring a more representative democracy by helping close historic gaps in voter registration and voter turnout.

Methodology for AVR analysis

Analysis for this report set out to project three primary findings for how the 2024 general election would look different if the 25 states without AVR had implemented the policy ahead of the 2024 general election:¹²

• How many more eligible citizens likely would have been newly registered to vote for the election?
• How many already registered voters likely would have updated their voter registration information with AVR?
• Of the newly registered voters, how many would likely cast a ballot in the general election?
• How many more eligible Black, Hispanic, and Asian citizens would have likely been newly registered to vote and, subsequently, likely cast a ballot in the general election?

In conducting this assessment, the author used existing studies and statistical data combined with original analysis, primarily utilizing voting and voter registration data collected by the U.S. Census Bureau. (Note: The author did not count North Dakota among states without AVR as the state does not conduct voter registration and therefore does not have a need to register voters through methods such as AVR.)

For the first finding, the author relied on analysis from the Brennan Center for Justice and FiveThirtyEight of the short-term impact that AVR had in seven states (Alaska, California, Colorado, Georgia, Oregon, Rhode Island, Vermont) and Washington, D.C., during the 2018 election cycle.¹³ These seven states and Washington, D.C., which implemented AVR in the two years prior to the 2018 election, provide a good case study of the short-term effects that the AVR policy in the FTVA could have had on states if it had been enacted a couple years prior to the 2024 general election. While many states have approved AVR within the past few years, they are still in the process of implementing the policy, so the author decided to rely on the same set of states analyzed by the Brennan Center and FiveThirtyEight as a case study where the short-term impacts of AVR were reliably reported. (Note: This analysis excluded findings from Alaska and Oregon, which were both studied by the Brennan Center and FiveThirtyEight, because Alaska utilizes the state’s permanent dividend fund for administering AVR and Oregon utilizes a back-end AVR model. The FTVA, in comparison, would enact a front-end AVR system through the DMV.¹⁴ States utilizing different forms of AVR are grandfathered into the FTVA, meaning the systems in these states would not be required to change their models. However, to more accurately project the effects of AVR as it would be enacted in the group of 25 states, the author determined that relying on data from front-end AVR systems implemented at DMVs would be the most accurate and useful.)

While the Brennan Center reported the real measured growth of registered voters in the seven states and Washington, D.C., following the implementation of AVR, the author derived the average percentage of registered citizens reported for those states for the 2020 election—which took place two to four years after implementation—in order to produce a reliable data point for projecting the short-term effects of AVR. The rationale for using the average registration rate of citizens relied on two factors:

• Some of the states in the group without AVR already had relatively high registration rates as of the 2022 election—despite the group as a whole having a measurably lower registration rate.
• Each state will generally reach a maximum or equilibrium citizen registration rate by implementing a front-end AVR solely through the DMV.

Following this methodology and these assumptions, the author calculated that by the 2020 general election, the remaining five states (California, Colorado, Georgia, Rhode Island, and Vermont) and Washington, D.C., had, on average, 74.2 percent of their citizen population registered to vote. In comparison, during the 2022 election cycle (the most recent election cycle data are available for), states without AVR had, on average, 68.4 percent of their citizens registered to vote—5.8 percentage points less than the average registration rate of states two years after the implementation of AVR.

To project how many more voters would be registered to vote if each of the 25 states without AVR had 74.2 percent of their citizen population registered to vote for the 2024 election, the author used voter registration data from the U.S. Census Bureau for the November 2022 general election.¹⁵ As of the time of publication, these data were the most recent voter registration data published by the U.S. Census Bureau. For states that had less than 74.2 percent of their citizen population registered to vote, the author calculated how many citizens would additionally be registered to vote if those states had 74.2 percent of their citizen population registered to vote. From this, the author subtracted the existing 72.9 million citizens registered to vote across the 25 states and determined that an additional 7.9 million additional voters would be registered to vote if the group of 25 states had, on average, 74.2 percent of their citizen population registered to vote.

This analysis assumes that the states that do not currently have AVR would likely follow similar trends in increases in citizen registration rates as shown by states that have enacted AVR. The projection of this report is a conservative estimate not only because it uses the average registered voter rate at 74.2 percent for states with AVR as the basis for projections, but also because this average is below or close to the voter registration rate of a handful of states without AVR, such as Kansas, with 76 percent, and Missouri, with 75.9 percent, of the states’ citizen populations registered to vote. In these instances, the author projected that voter rolls would not grow in those states with the implementation of AVR. However, in reality, this likely would not be the case. Following the implementation of AVR, even states with high citizen registration rates saw meaningful increases in voter registration rates. The author’s decision to not project growth for those states was made to provide a more conservative estimate for the impacts of AVR. It is also important to note that AVR is an incredibly effective tool for helping eligible citizens register and for keeping voter rolls accurate and up to date, which is another reason why AVR should be implemented even in states with high rates of registered voters.

Next, the author of this report set out to project how many of the existing 72.9 million registered voters across the group of 25 states would likely update their voter registration information through AVR ahead of the 2024 general election. For this, the author relied on analyses conducted by FiveThirtyEight in 2019 of states that enacted AVR through the DMV.¹⁶ FiveThirtyEight’s analysis looked at the same set of seven states and Washington, D.C., that the Brennan Center studied and relied on data reported for the 2018 general election. For the same reason as noted above, analysis for CAP’s report excluded Alaska and Oregon.

In order to project the number of existing voters who would have likely updated their registration information, analysis for this report relied on:

• Reporting from FiveThirtyEight on the number of already registered voters who used AVR to update their registration information in the two years following the implementation of AVR in their state.
• U.S. Census Bureau voter registration data for the 2018 and 2022 general elections. Using the data from FiveThirtyEight and the 2018 Census Bureau data for the five states and Washington, D.C., the author calculated the percentage of total registered voters who updated their voter registration.

For example, FiveThirtyEight reported that 1.6 million voters in California updated their voter registration information ahead of the 2018 election, and the Census Bureau reported 15.7 million registered voters in California for the 2018 general election.¹⁷ From this information, the author calculated that 10.2 percent of California voters updated their voter registration information in the lead-up to the federal election, immediately following the state’s implementation of AVR. This voter registration update rate was calculated for all five states (California, Colorado, Georgia, Rhode Island, and Vermont) and Washington, D.C., and then the average voter registration rate was calculated across the group. As highlighted in this report, the author found that for the relevant datasets, 2.8 percent to 48.3 percent of already-registered voters updated their registration information, and the average update rate for the dataset was 17.1 percent. Then the author projected that if, on average, 17.1 percent of already-registered voters across the group of 25 states updated their registration information ahead of the 2024 election, that would amount to 12.5 million voters. The projection of the number of likely registration updates was calculated based on data from the 2022 general election, the most recent voter registration data reported by the Census Bureau at the time of this report.

The author then set out to determine how many voters newly registered through AVR likely would have cast a ballot in the 2024 election. While studies show that voter turnout rates for voters registered through AVR tend to be lower than the voter turnout rates for voters who register themselves, turnout rates for voters newly registered through AVR are still significant.¹⁸ The same 2019 FiveThirtyEight reporting used for prior projections in this report found that between 42 percent and 54 percent of voters who registered through AVR cast a ballot in the 2018 election, compared with between 46 percent and 76 percent of voters who registered to vote via other methods.¹⁹ With the exclusion of Oregon; no reported data for Alaska, California, and Georgia; and unreliable data reported by Vermont, analysis for this report relied on data collected for Colorado, Rhode Island, and Washington, D.C. The author found that the average voter turnout rate across this group was 48.3 percent for voters registered through AVR. From there, the author projected that of the 8 million newly registered voters (as projected above) who would be registered across the group of 25 states without AVR, if, on average, 48.3 percent of those new voters cast a ballot in the 2024 election, that would equate to 3.8 million additional voters casting a ballot.

Lastly, the analysis looked to find the specific impact that AVR would have on voter registration and turnout for Black Americans, Hispanic Americans, and Asian Americans (three demographic groups designated by the Census Bureau). Based on the measured change in registration rates recorded in states following the implementation of AVR, the policy had the most significant impact on the registration rate of Black citizens out of the three demographic groups, as well as the most significant impact on closing the registration gap between Black citizens and white citizens.

For the same set of five states (California, Colorado, Georgia, Rhode Island, and Vermont) and Washington, D.C., used for the previous part of the analysis, the author found that from 2018 to 2020, the average voter registration rate for Black citizens (Black alone or in combination) increased from 59.5 percent to 69.4 percent, a 9.9 percentage-point increase. In that same period, the average voter registration rate for Hispanic citizens increased by 5.5 percentage points, from an average registration rate of 57.1 percent to 62.5 percent. The average voter registration rate for Asian citizens (Asian alone or in combination) increased by 4.2 percentage points, from an average registration rate of 62.1 percent to 66.4 percent. In comparison, the registration rate for white citizens increased by 6.3 percentage points, from an average registration rate of 69.8 percent in 2018 to 76.1 percent in 2020. Of the three demographic groups analyzed, the growth in registration rate of Black citizens alone outpaced the growth of the registration rate of white citizens.

The average voter registration gap between white citizens and Black citizens decreased by more than 3 percentage points, from an average 10.3 percent registration gap in 2018 to an average 7 percent registration gap from 2018 to 2020. For Hispanic and Asian citizens, the voter registration gap slightly increased despite significant increases in the overall registration rates of the two demographic groups.

The author then applied these short-term observed trends and findings to calculate the likely number of newly registered Black, Hispanic, and Asian citizens in the 25 states with AVR and find how many of those newly registered voters would likely cast a ballot in the federal election following the implementation of AVR.

According to the most recent census voter registration data available for the 2022 general election, the 25 states without AVR had an average Black citizen voter registration rate of 60.6 percent, an average Hispanic citizen voter registration rate of 48.4 percent, and an average Asian citizen registration rate of 56 percent. In comparison, the overall average voter registration rate of these states was 68.4 percent.

First, the author set out to calculate how many more Black Americans would be registered to vote if the implementation of AVR in the 25 states without AVR followed established precedent. Of the 25 states, South Dakota had the lowest voter registration rate at 30.1 percent while New Hampshire had the highest at 81.2 percent. Given the very wide variance in voter registration rates for Black Americans, a slightly different approach was used to estimate the number of likely Black registrants. Idaho, Montana, and Wyoming were excluded from projections because they had unreliable population sizes of fewer than 10,000 Black citizens or because the census was not able to report population numbers. For the projection, it was estimated that a state’s voter registration rate would increase by either the average observed increase of 9.9 percentage points or up to the observed average Black citizen registration rate of 69.4 percent following the short-term implementation of AVR—whichever was less in order to not overestimate growth. Additionally, if a state already had an average Black citizen registration rate higher than the observed 69.4 percent average, no increase was taken into account for the projection, even though one would most likely be observed following the implementation of AVR, as has been the case for other states with already relatively high registration rates. For example, in Washington, D.C., in 2018—the year that AVR was implemented—72.1 percent of Black citizens were registered to vote, but just two years following the implementation of AVR, 83.3 percent of Black citizens were registered to vote. South Dakota’s Black citizen registration rate was calculated to increase the observed average increase of 9.9 percentage points, from 30.1 percent to 39.1 percent; Tennessee’s Black citizen registration rate was calculated to increase from 65.4 percent to the observed 69.4 percent average Black citizen registration rate (an increase of only 4 percentage points); and no increase was calculated for New Hampshire with a more than 69.4 percent Black citizen registration rate. Using this methodology, the author calculated that an additional 1.2 million Black citizens would be registered to vote across the group of states without AVR in the two years following the implementation of AVR.

The same methodology was used to calculate likely increases in the voter registration rate for Hispanic citizens and Asian citizens. For Hispanic citizens, an average 5.5 percentage-point increase or an increase up to the average registration rate of 62.6 percent was calculated, whichever was less; for Asian citizens, an average 4.2 percentage-point increase or up to the average registration rate of 66.4 percent was calculated, whichever was less. The same methodology for excluding statistically unreliable population sizes and not projecting increases for states with existing higher than average registration rates for each demographic was carried over. Using this methodology, the author calculated that an additional 642,000 Hispanic citizens and an additional 66,000 Asian citizens would be newly registered to vote across the 25 states without AVR in the two years following its implementation.

Then, using the previously explained analysis using data from FiveThirtyEight that, on average, 48.3 percent of voters newly registered through AVR cast a ballot in the following election, the author calculated how many more newly registered Black, Hispanic, and Asian citizens would likely cast a ballot. FiveThirtyEight’s analysis was not broken down by race, but the author decided to use the overall finding for each specific demographic. Using this methodology, the author calculated that an additional 600,000 Black citizens, an additional 310,000 Hispanic citizens, and an additional 32,000 Asian citizens would be registered to vote.

Same-day voter registration (SDR) ensures that eligible Americans can register to vote at the polls and that registered voters can update their voter registration information—changes in address, contact information, and party affiliation—at the polls. The FTVA would ensure that eligible voters in all states have access to SDR during the early voting period as well as on Election Day. SDR also helps mitigate issues with inaccurate voter roll purges that often disproportionately affect voters of color.²⁰ The SDR policy in the FTVA would ensure that if a voter is inaccurately removed from the voter rolls ahead of Election Day, they are not only able to re-register at the polls but also able to cast a nonprovisional ballot.

SDR greatly benefits communities that tend to move more frequently and therefore need to update their voter registration information more often or register to vote in a new county or state. Studies and U.S. Census Bureau data have shown that Hispanic Americans and young Americans are among the demographic populations that move the most frequently.²¹ For this reason, they were a focus of the analysis for the potential impacts of SDR.

Methodology for SDR analysis

The author set out to determine the most significant effects of implementing SDR policies across the group of 27 states that have not yet implemented the policy.²⁶ (Note: Because North Dakota does not conduct voter registration, the state was excluded from the group of states without SDR. North Carolina only conducts voter registration during the early voting period, and since the FTVA would require states to conduct voter registration both during the early voting period and on Election Day, it was excluded from analysis and not considered as a state with or without voter registration.)

Given the existing research and focus on the impacts of SDR on Hispanic/Latino Americans, Black Americans, and young Americans, analysis for this report also focused on showing the likely impacts of SDR on these groups.²⁷ The author set out to determine how many more Hispanic, Black, and young Americans would likely cast a ballot if they had access to SDR.

One academic study from the University of Massachusetts and Dēmos found that within the same U.S. geographic region, voter turnout rates were up to 17 percentage points higher for Black and Latinx voters living in states with SDR compared with those living in states without SDR across the 2012, 2014, and 2016 election cycles.²⁸ Based on these initial findings, the author conducted original analysis looking at data from the U.S. Census Bureau for the 2018, 2020, and 2022 general elections and compared voter turnout rates for Black voters and Hispanic voters in states that had SDR with voter turnout rates in states without SDR. It was not only helpful to look at the three most recent election cycles because many states have implemented SDR within the past decade, but also because even in 2016, not many states had SDR. The author therefore determined that using data prior to 2018 would not be statistically reliable.

The author set out to determine the average voter turnout rate for Hispanic voters and Black voters for both groups of states. For example, for the 2022 election cycle, the voter turnout rate for Hispanic voters in the group of 27 states that did not offer SDR was added up and divided by 27 in order to find the average Hispanic voter turnout rate for states without SDR. Similarly, this statistic was calculated for Hispanic voters in the 22 states that did offer SDR during the 2022 election cycle. The analysis used the average voter turnout rate rather than the turnout rate of the entire group of states (the number of voters casting a ballot across the entire group divided by the citizen population of the entire group) to prevent populous states from having an outsize influence on the analysis. The average voter turnout methodology was used rather than a general voter turnout approach in order to better isolate the effects of SDR on a state’s voter turnout rate.

Once the average voter turnout for Hispanic and Black voters was calculated for the group of states with SDR and the group of states without SDR for each of the three election cycles, the difference between the two groups was calculated. For example, in the 2022 general election, Hispanic voters in states with SDR had an average voter turnout rate that was 6.2 percentage points higher than Hispanic voters in that same election in states without SDR. Similarly, the voter turnout for Black voters in that same election was 4.1 percentage points higher in states with SDR than in states without SDR. To obtain a more accurate estimate of the likely effects of SDR, the difference in average voter turnouts was averaged over all three election cycles. Across all three election cycles, for Hispanic voters, the average voter turnout was 6.6 percentage points higher in states with SDR compared with those without SDR; for Black voters, it was 1.1 percentage points higher.

This analysis was helpful in showing a likely relationship between the use of SDR and higher voter turnout rates for Hispanic and Black voters and establishing a baseline to more confidently analyze the likely increase in voter turnout for these groups of voters by implementing SDR. However, simply using these percentage point differences in turnout would have resulted in likely unrealistic levels of turnout for Black and Hispanic voters in a presidential election year, given that the analysis found significantly greater differences in turnout rates for midterm election cycles. This also indicates the importance of SDR and making voting accessible during the generally lower-turnout midterm election cycles.

Instead, analysis for this report sought to calculate how many more voters would cast a ballot in 2024 if the average voter turnout rate for the group of states without SDR equaled that of the average voter turnout of states with SDR during the previous presidential election cycle. During the 2020 election cycle, Black voters in states without SDR had a 0.8 percentage point higher turnout than Black voters in states with SDR. One of the primary factors for this is likely that significant portions of the country’s Black citizens live in states without SDR. While over the three election cycles, the author found that states with SDR had higher turnout for Black voters on average, the author decided to only use analysis to project potential differences related to SDR for Hispanic voters. Based on the average voter turnout rate difference of 4.5 percentage points during the 2020 election, the author calculated that based on the most recent reported number of Hispanic citizens, if the average Hispanic voter turnout rate of the 27 states without SDR were 4.5 percentage points higher, that would equate to approximately 1.1 million additional Hispanic voters casting a ballot.

Data from states that reported population numbers of fewer than 10,000 for a demographic group were excluded to derive a more accurate analysis. For example, during the 2022 election cycle, Vermont reported a Hispanic citizen population of fewer than 10,000 people. As a result, Vermont was excluded from the calculator for average voter turnout for Hispanic voters in the 2022 general election. Similarly, states where the Census Bureau could not reliably report citizen population numbers were excluded. For example, the Census Bureau was not able to report reliable data for the number of Black voters in Idaho for the 2022 election, and therefore, the state was excluded from the calculator for average voter turnout rate for Black voters in 2022. Lastly, in some instances, the Census Bureau was able to report data for both citizen population data and votes cast but did not report voter turnout data (the number of votes cast divided by citizen population) because the dataset would have been too small for the Census Bureau to deem it statistically reliable. However, where those two datasets were available, analysis for this report calculated and used voter turnout data for those states for subsequent calculations. For example, for Delaware in the 2018 general election, the Census Bureau reported that there were 52,000 Hispanic citizens in the state and 20,000 Hispanic voters cast a ballot. However, because the voter turnout data would have represented fewer than 75,000 voters, the Census Bureau did not include a voter turnout rate. However, the author calculated that rate at 38.5 percent and used Delaware to calculate the average voter turnout rate.

Another academic study on voter turnout from The University of Chicago found that youth voters—those ages 18 to 24—were the age group most affected by SDR. The study found that if states implemented SDR, youth voter turnout in presidential elections would likely increase between 3.1 and 7.3 percentage points.²⁹ The author then set out to find how many more youth voters would likely cast a ballot if the SDR policy in the FTVA were enacted ahead of the 2024 general election. To do this, the author sought to determine how many more youth votes would be cast if youth voter turnout increased between 3.1 and 7.3 percentage points in the 27 states without SDR. In this case, because detailed academic research over multiple election cycles had been conducted for the impacts of SDR on youth voter turnout, the author simply looked at Census Bureau voter turnout data for voters ages 18 to 24 for the 2020 presidential election and relied on the University of Chicago study for projections rather than conducting original analysis. Based on voter turnout in the 2020 general election for the 27 states without SDR, the author calculated how many more youth voters would cast a vote if turnout increased between 3.1 and 7.3 percentage points for that group. Using this methodology, the analysis concludes that an increase of 3.1 percentage points would mean an additional 460,000 youth voters would cast a ballot, and an increase of 7.3 percentage points would mean an additional 1.1 million youth voters would cast a ballot. Thereby, as many as 1.1 million youth voters would likely cast a ballot if SDR were implemented in all states.

Online voter registration (OVR) is one of the easiest ways for eligible citizens to register to vote and update their voter registration information. Online resources and technologies are thoroughly and seamlessly integrated into the everyday lives of Americans, including a plethora of federal services. Despite this, millions of Americans still cannot access OVR for federal elections. In contrast, federal tax filing has been available online for 38 years.³⁰ The FTVA would ensure that all states enable eligible voters to fill out and submit a voter registration form online. Many states that still do not offer OVR are among the most rural states in the country, where access to OVR would meaningfully improve access to the ballot box.

Methodology for OVR analysis

Analysis for this report used data from the Census Bureau for the 2022 general election to calculate the registered voter population and citizen population for the seven states where OVR will still not be available for the 2024 general election.³⁴ The analysis of the cost of processing paper voter registration applications relied on previous reports, surveys of election officials, and data from the U.S. Election Assistance Commission’s (EAC) 2022 EAVS.³⁵ (Note: North Dakota was excluded from analysis as the state does not conduct voter registration and therefore does not have the need for OVR.)

A 2015 American Civil Liberties Union report found that California saved $2.85 per voter application by moving to an OVR system, while Arizona saved $0.80 per application.³⁶ A more recent 2017 survey of election officials in 34 states conducted by Vote.org found that the labor cost of just processing paper registration forms amounted to $3.54—not including supplies, printing, and mailing costs.³⁷ The same survey found that the average election office spends $113,445 on staff time processing paper voter applications. Because the Vote.org survey was conducted in 34 states and was more recent, the author chose to use $3.54 to measure the cost of processing a paper voter registration application. Even this estimate can be considered conservative as it only accounts for staff time processing, not for the other costs associated with the total costs of paper registration forms—the cost of paper and costs associated with printing and mailing voter registration applications.

Using the $3.54 average labor and materials cost associated with paper voter registration applications, the author sought to determine the amount of election funding savings on processing paper registration forms that the seven states without OVR would have realized since the 2018 general election. While some paper registration forms would still be submitted by voters, as shown with OVR in other states, the number of paper registrations submitted would be significantly reduced. In the most recent Election Administration and Voting Survey (EAVS) of the 2022 general election, the EAC gathered and reported data from states on the number of voter registration applications that were received by different methods in the two years leading up to the 2022 general election.³⁸ (Note: The author would have analyzed data from the 2020 presidential election or previous elections to better calculate the average or specific number of paper registrations processed by states, but data at this level of detail for all states were not available in previous versions of the EAVS). Analysis for this report calculated that in total, 2.3 million voter registration applications were received by the seven states by mail, email, or fax and in person at an election office. For the 2022 election cycle alone, the $3.54 per application would have amounted to $8.3 million spent on processing paper voter registration applications across the group of states. To extrapolate how much states would have saved since the inception of the FTVA, the author set out to calculate the total sum for applications received leading up to the 2020 election cycle, the 2022 election cycle, and the 2024 election cycle. Using data from the 2022 election cycle again should be considered a more conservative estimate as states receive more voter registration applications leading up to a presidential election cycle than they do for a midterm election cycle and data from the 2022 midterm election was used to project for this time range. Based on the 2.3 million voter registrations received for the 2022 midterm cycle alone, if the entire group of states received the same number of applications for each of the three election cycles, that would amount to nearly 7 million voter registration applications. At the cost of $3.54 per application, the seven states would have spent $24.8 million solely for processing paper voter registration forms.

Voters and election officials deserve significant resources for elections. However, OVR as a system improves voters’ access to the ballot box and saves election officials considerable time and resources that they could spend on other areas of election administration and improvement. By failing to have transitioned to the system years ago, states have continued to waste election resources in the form of staff time and funding, while denying voters one of the easiest ways to register to vote and to update their voter registration information.

The Freedom to Vote Act would ensure that all Americans can cast their ballot in the way that works best for them. The legislation would expand voting options by guaranteeing access to two weeks of early in-person voting and mail-in voting for all Americans. The FTVA would further increase representation by restoring the right to vote for returning citizens who have paid their debt to society after serving a felony sentence—a right that is still denied to approximately 3.3 million Americans.³⁹

In the 2024 election cycle, 10 states will bar many citizens who have served their sentence for felony convictions from voting.⁴⁰ Another 15 will not allow returning citizens to vote if they are on parole or probation.⁴¹ In addition, many states prohibit returning citizens from regaining their right to vote if they haven’t fully paid outstanding fines and fees associated with their sentence.⁴² In practice, these prohibitions serve as a pay-to-vote system and a modern-day poll tax that disenfranchises Americans too poor to pay to regain their constitutional right to vote. Because of overincarceration and the racial inequalities inherent to America’s legal system, significant portions of populations in many states cannot vote, with a vastly disproportionate amount of those citizens being Black or Hispanic/Latino.⁴³ American democracy cannot be truly representative so long as many citizens remain disenfranchised.

Methodology for re-enfranchisement analysis

In order to estimate the number of citizens who would be re-enfranchised by the policy in the FTVA, the author relied on detailed data collected by The Sentencing Project in 2022.⁴⁷ In The Sentencing Project’s overall estimations of disenfranchised individuals, those currently serving time in prison were counted among those disenfranchised. The policy in the FTVA, however, would not re-enfranchise individuals serving time in prison, it would re-enfranchise individuals upon their release and return to society, individuals serving parole or probation, as well as any individual who is post-sentence but continues to be denied their right to vote. Based on this policy and The Sentencing Project’s data, the author calculated the number of people who are denied the right to vote based on “parole,” “felony probation,” and “post sentence” in the states that still restrict voting on this basis. Based on the latest data by The Sentencing Project, there are 314,000 people denied the right to vote based on “parole,” 967,000 people denied the right to vote due to a “felony probation,” and more than 2 million people still denied the right to vote “post sentence.”

In addition to early in-person voting, voting by mail offers millions of Americans—especially those living in rural or remote areas—a critical way to cast their ballot ahead of Election Day. During the COVID-19 pandemic, many states were unprepared to widely offer voting by mail, and election officials had to scramble to ensure Americans could vote safely and securely.⁴⁸ While most of those states provided temporary exemptions to voting-by-mail restrictions, they’ve mainly been rolled back, and tens of millions of Americans will not have the option to cast their ballot by mail for the 2024 election cycle.⁴⁹ The FTVA would ensure that all Americans can choose the method of voting that works best for them—including voting by mail.

Additionally, millions more Americans will lack easy access to secure ballot drop boxes to return their mail-in ballots. Not only are ballot drop boxes one of the easiest ways for voters to return ballots, but they are also a more efficient way to return ballots to the hands of election officials and can save states significant election funding as they do not require return postage. Despite these facts—as well as the safe and secure use of thousands of ballot drop boxes in numerous states over the past decade and more—several states have restricted the use of ballot drop boxes.⁵⁰ The FTVA would ensure that all voters have access to secure ballot drop boxes 24 hours per day, 7 days per week during the early voting period and on Election Day.

Methodology for voting-by-mail and ballot drop box analysis

Original analysis for this report set out to project the following:

• How many more voters would likely cast a ballot if they had the option to vote by mail?
• How many more voters of the existing voting population would likely cast their ballot by mail rather than in person?
• How many more drop boxes would be available to voters based on the requirements of the FTVA?
• How many more voters would likely return their ballot to a drop box if they had access to ballot drop boxes?

To determine how many more voters would likely cast a ballot if they had the option to vote by mail, the author determined the difference in average voter turnout between states allowing registered voters to request a mail-in ballot and states requiring voters to provide an excuse permitted under state law. For this analysis, states that automatically mail out ballots to voters were excluded, as they tend to have higher voter turnout rates and because they would not provide an accurate representation of the policy in the FTVA, which would still require voters to request a mail-in ballot. Additionally, this analysis used data from only the 2022 midterm general election to calculate average turnout rates. While data from multiple election cycles were used for other analysis in this report, two factors affected the use of vote by mail in previous election cycles. First, in 2020, nearly all states allowed voters to request a mail-in ballot due to the COVID-19 pandemic and only a handful of states continued to require an excuse, so comparing these sets of states for turnout would not have been reliable.⁵² Because of the pandemic, vote-by-mail numbers also soared and, in general, the 2020 election will likely continue to be an anomaly in vote-by-mail trends, even though it fundamentally altered many Americans’ awareness of the ease of voting by mail, which continues to have an impact on voting trends today. Additionally, before the 2020 election cycle, voting by mail was much less widely used by Americans and would not provide an accurate representation of its future widespread use.⁵³

When looking at the average voter turnout rate of no-excuse states compared with excuse states using U.S. Census Bureau data from the 2022 general election, the analysis found that states that did not require an excuse to vote by mail had a 54.5 percent average voter turnout rate compared with an average voter turnout rate of 47.9 percent for states that require an excuse—a difference of 6.7 percentage points. Then, using data from the 2020 election to project turnout for a presidential election, the author found that a 6.7 percentage point increase in voter turnout among states that require an excuse to vote by mail would equate to more than 5 million additional votes being cast.

To determine how many voters who participated in recent elections would likely choose to cast their ballot by mail instead of in person, the author used data from the EAC’s 2022 EAVS on the number of mail-in ballots cast in each state and calculated the percentage of total votes that were cast by mail in each state, also known as the vote-by-mail rate. For example, during the 2022 general election, Connecticut reported to the EAC that 150,000 mail-in ballots were cast in the state, and the Census Bureau reported 1.3 million total ballots were cast. The author then calculated that Connecticut had a 12 percent vote-by-mail rate. For states that required an excuse to vote by mail, the average vote-by-mail rate was 5.4 percent, and for no-excuse states, it was 19 percent for the 2022 general election—a difference of 13.6 percentage points.

In order to project how many more ballots would be cast by mail in these states during the 2024 election, if the overall voter turnout rate did not increase, the author projected how many ballots would likely be cast for the 2024 election if the vote-by-mail rate of each state stayed consistent with that of the 2022 general election. For example, Connecticut had a 12 percent vote-by-mail rate in the 2022 election and voter turnout of nearly 1.7 million votes cast in the 2020 presidential election. Using these data for the most recent vote-by-mail rate and turnout in the most recent presidential election, the author projected that approximately 202,000 mail-in ballots would be cast in Connecticut for the 2024 general election. Similarly, South Carolina, with a 3.2 percent vote by mail rate in the 2022 general election, would expect to see around 78,000 mail-in ballots cast for the 2024 election at that rate. The author then calculated how many additional mail-in ballots would likely be cast if each state had the average mail-in voting rate of 19 percent. In Connecticut, a 19 percent vote-by-mail rate would mean 319,000 mail-in ballots cast in the 2024 election, accounting for similar voter turnout as the 2020 presidential election. Therefore, the author projected an additional 118,000 mail-in ballots cast if all Connecticut voters could request a mail-in ballot. For each state, the additional mail-in ballots were calculated based on the difference between the 19 percent average rate and the expected rate from 2022. Based on this methodology, the author calculated that across the 14 states without no-excuse voting by mail, an additional 4.9 million ballots would likely be cast by mail.

There is no compelling reason to believe that for the group of excuse states, many of which have conservative state governments, the vote-by-mail rate would stay low following the implementation of no-excuse voting by mail.⁵⁴ In other states with similar political makeups but that allow for no-excuse vote by mail—such as Florida, Iowa, Idaho, Nebraska, North Dakota, Ohio, and South Dakota—there is a relatively high vote-by-mail rate. Whereas the average vote-by-mail rate for the entire group of no-excuse states was 19 percent, for the 2022 general election, Florida had a 36.4 percent mail-in voting rate, while Iowa’s was 30.4 percent, Idaho’s was 18.8 percent, Nebraska’s was 36.7 percent, North Dakota’s was 24.1 percent, Ohio’s was 21.9 percent, and South Dakota’s was 27.8 percent.

To analyze the effects of the drop box policies in the FTVA, this analysis predominantly relied on the 2022 EAVS, which collected data on the number of drop boxes states deployed throughout the election cycle. The FTVA would require that each state have one ballot drop box for every 45,000 registered voters. Using Census Bureau data from the 2022 election cycle, the author calculated the number of drop boxes that would be required in each state under the FTVA. For example, there were 1.4 million registered voters in Arkansas for the 2022 general election, so with the FTVA’s requirement of one drop box for every 45,000 voters, the state would need to provide at least 30 drop boxes. Arkansas reported using only six drop boxes during the 2022 election cycle; therefore, to meet the FTVA requirement, the state would need to provide at least an additional 24 drop boxes. Using this same methodology for all states, including states that do not require an excuse but do not have a sufficient number of drop boxes under the FTVA, at least an additional 946 drop boxes would need to be deployed across the United States. Of these, 822 would be in the 12 states that have no drop boxes, 24 would be in states that require an excuse to vote by mail but do provide some drop boxes, and 101 would be in no-excuse states that do not have the minimum number of drop boxes required under the FTVA. It should be noted that the FTVA requires either a drop box for every 45,000 registered voters or a drop box for every 15,000 ballots cast by mail—whichever number of drop boxes is greater. The author calculated the required number of drop boxes under both methodologies and found that currently, for all states, the requirement for a drop box for every 45,000 registered voters would produce the higher number.

Lastly, the analysis tried to project how many more existing voters would be returning their ballot to a drop box, even if voter turnout did not increase but the average vote-by-mail rate of excuse states reached the 19 percent average vote-by-mail rate of no-excuse states, as calculated above. Based on 2022 EAVS data, the author estimated the average drop box rate of return for no-excuse states with more than 10 drop boxes to be 27.9 percent, meaning that on average, 27.9 percent of mail-in ballots were returned via drop box.

The author calculated that with 5.8 million total mail-in ballots projected to be cast in the 12 states without any drop boxes, at the 27.9 average return rate, 1.6 million would likely be returned to a drop box.

Additionally, for four states that reported drop box use but that do not offer no-excuse voting by mail, the author calculated the difference between: 1) the number of mail-in ballots that would likely be returned to a drop box based on the average 19 percent vote-by-mail rate and 27.9 percent drop box return rate; and 2) the number of ballots that would be projected to be returned by drop box at the 19 percent vote-by-mail rate but return rate a state reported for the 2022 general election. For example, in Connecticut, a 19 percent average vote-by-mail rate would equate to 319,000 ballots cast and at a 27.9 percent average drop box return rate, 89,000 of those would be returned to a drop box. However, as some voters can cast a ballot by mail with an excuse in Connecticut and have some drop boxes available, in order to project the number of additional ballots returned under the FTVA, the author needed to consider how many ballots would be returned to a drop box for the 2024 election without the policies in the FTVA. In 2022, Connecticut reported a 20.8 percent drop box return rate; at that rate, 66,000 ballots would be expected to be returned in the state if all voters had the option to cast a ballot by mail. Therefore, the author projected that an additional 23,000 ballots would be returned in Connecticut. Based on this methodology, the author projected that an additional 80,000 ballots would be returned to drop boxes in the four states that do not allow all voters to cast a ballot by mail but have some available drop boxes.

In total then, 1.7 million additional drop boxes would be returned across states that do not allow for no-excuse voting by mail and states without drop boxes.

Every election cycle, early in-person voting (EIPV) ensures that millions of Americans can cast their ballot prior to Election Day. However, millions of Americans either have no access to EIPV opportunities or have access to limited days that do not provide the necessary flexibility for them to access the ballot box easily. During the 2022 election, the most common reason that eligible Americans gave for not casting a ballot was because they were too busy on Election Day and had conflicting work and school schedules.⁵⁵ The FTVA would set a national floor and ensure that all Americans have access to at least two weeks of EIPV, including on weekend days.

Methodology for EIPV analysis

The original analysis for this report looked at the difference in voter turnout rates between states that did and did not offer at least two weeks of EIPV during the 2022 midterm election and 2020 presidential election. The author sought to estimate how many more voters would cast a ballot if they had access to two weeks of EIPV. States that automatically mail out ballots to all voters were excluded from the entire EIPV dataset, as those states persistently have higher voter turnout rates and the vast majority of voters in those states cast ballots by mail, not in person. While the FTVA would require two weeks of EIPV where voting is available in the same form as on Election Day—not merely in-person absentee voting where voters can fill out an absentee or mail-in ballot at a designated location—for the purposes of this analysis, a distinction between those two forms of early voting was not made, as long as it was available for a minimum of two weeks. EIPV voting where voting is widely available in the same form as Election Day would likely increase early voting rates and turnout in many states that currently offer in-person absentee voting. However, increases for those states were not projected, and to simplify and provide a more conservative projection, those states were treated as offering two weeks of EIPV. Additionally, the EIPV analysis did not look at data from the 2018 election cycle because the 2020 election significantly transformed public consciousness around early voting methods that carried on to the 2022 election cycle and beyond.

As with other analyses in this report, the author sought to determine the average voter turnout rate for each group of states. The author chose this approach instead of finding the voter turnout rate of the entire group of states (the number of voters who cast a ballot in the entire group divided by the total number of eligible voters in the group) so that populous states did not have an outsize influence in the dataset. The analysis found the average voter turnout rate by adding up the voter turnout rate of states with two weeks of EIPV and, separately, the voter turnout rate of states without two weeks of EIPV and divided those sums by the number of states in each group. The approach of finding the average voter turnout rate therefore treated each state’s voter turnout rate as an equal data point.

Using data from the U.S. Census Bureau, the analysis found that states that offered at least two weeks of EIPV during the 2022 midterm general election had a voter turnout rate 3.6 percentage points higher than those that did not. Similarly, during the 2020 general election, states that offered at least two weeks of EIPV had an average voter turnout rate that was 2.3 percentage points higher than those that did not. To find a more accurate statistic, the difference from these two elections was averaged out, resulting in findings that states with at least two weeks of EIPV had voter turnout rates that were, on average, 3 percentage points higher over the past two election cycles.

Using this data point, the author set out to project how many more voters would likely cast a ballot if the nine states not offering two weeks of EIPV for the 2024 general election did so. To carry out this projection, the author looked at the overall voter turnout for those nine states during the 2020 general election to ensure turnout was representative of higher turnouts in presidential election cycles.⁵⁸ The author then calculated how many more voters would likely vote early in person if voter turnout across those states increased by 3 percentage points. Based on voter turnout in the 2020 election, a 3 percentage-point increase in voter turnout would result in an additional 1.4 million votes cast across the nine states.

The author then set out to find how many voters within the existing voting population (those who have cast a ballot in previous elections) would likely choose to cast their ballot early in person. Based on data from the EAC for the 2022 election cycle, the author calculated that on average, the EIPV rate (how many of the total votes cast in the election were cast early in person) for states that offered some, but not 14 days, of early voting was 19.7 percent, compared with an EIPV rate of 22.4 percent for states with at least 14 days of EIPV.⁵⁹ Based on this, accounting for no changes in voter turnout, the author calculated that at a 22.4 percent EIPV rate, across the nine states an additional 1.7 million voters would choose to cast their ballot early. This could be seen as a conservative estimate since West Virginia and Florida, with 10 and seven days of early voting, respectively, had EIPV rates of 23.9 percent and 30 percent higher, respectively, than the calculated average for the group of states with 14 days of EIPV; therefore, additional votes for these states were not considered, but additional days of early voting would almost certainly result in additional early ballots cast.

Analysis in this report leaves little doubt that the Freedom to Vote Act would transform elections and increase access to the ballot box for millions of Americans. In many ways, when looking at the policies analyzed in this report and the number of states that have had these policies in place for multiple federal election cycles, it’s clear that the FTVA does not seek to innovate or upheave elections. Instead, the legislation would help ensure that all Americans have baseline access to the ballot box, regardless of their state or ZIP code.

Many recent elections have been won with small margins, and these elections have had profound consequences on charting Americans’ futures. This analysis shows that election results could be more representative and democratic if all Americans could cast their votes with equal access and ease.

In states where elections have been decided by even tens of thousands of votes, a single policy in the FTVA could have swayed the outcome. This is important not because of party or ideology, but because it is vital for the health and longevity of American democracy that elections reflect the will of the people. The only way that can truly happen is by establishing meaningful baseline voting election standards as the FTVA would do.

Americans’ trust and faith in government institutions and democracy are waning. Transformative change is needed to expand, not restrict, access to the ballot box. The analysis in this report clearly shows how the 2024 election cycle would have likely seen increased voter registration and turnout had the voting policies in the FTVA been enacted. It’s past time that Congress pass the Freedom to Vote Act to demonstrate a genuine commitment to defend American democracy. Congress cannot allow another federal election cycle to pass without ensuring that all Americans’ constitutional right to vote is upheld.

The author would like to thank Sydney Bryant, Alice Lillydahl, Beatrice Aronson, and Carl Chancellor for their contributions to this report.