A Democracy Crisis in the Making

In June 2023, Protect Democracy, along with our partners at the States United Democracy Center and Law Forward released the latest edition of our report, A Democracy Crisis in the Making.

In April 2021, we published the first edition of A Democracy Crisis in the Making: How State Legislatures Are Politicizing, Criminalizing, and Interfering with Election Administration. That Report identified a burgeoning trend in state legislatures: bills that would increase the risk of election subversion — that is, that the declared outcome of an election does not reflect the true choice of the voters. Through the 2021 and 2022 state legislative sessions, we tracked nearly 400 legislative proposals that would make election subversion more likely. Fifty-six of them ultimately became law in 26 states.

In the two years since our first Report, the danger of election subversion has drawn wider attention. During the 2022 midterm elections, the future of nonpartisan election administration was a campaign issue. Exit polling showed that “democracy” was a top concern for voters. That election came and went without major crises, and the accurate results were ultimately certified on time and in accordance with the law. Moreover, voters in certain states decisively rejected election deniers, whose beliefs were rooted in the baseless conspiracy theory that the 2020 presidential election was stolen from Donald Trump. Some states are moving towards proactively safeguarding against election subversion by advancing legislation to enhance security and privacy protections for election workers or refocusing investigative efforts on improving voter access.

These developments are cause for optimism. Yet we cannot sound the all-clear.

Election deniers did win key statewide and local election administration posts, as well as seats in state legislatures and Congress. As we demonstrate in this edition of the Report, the danger of a democracy crisis has not passed. In only the first few months of the 2023 state legislative year, from January 1 through May 3, we have identified 185 bills in 38 states that would increase the risk of election subversion — a pace roughly on par with that of the previous two years. This count includes bills in each of the five categories of election-subversion legislation that we outlined in previous editions of this Report. The bills we have identified in early 2023 illustrate that the election-subversion threat is evolving as legislators develop new ways to interfere with election administration and double down and expand on previous trends. Many of these bills are designed to inject confusion and delays into the election process, which increases the likelihood of attempted subversion and can give rise to disinformation, further eroding public trust and confidence in election results.

The bills included in this year’s Report show that the subversion threat is very much still alive:

  1. Usurping control over election results. These bills would give legislators or other state officials direct control over election outcomes. In one extreme example, a bill in Texas would give the secretary of state power to order a new election in the state’s most populous county under certain circumstances. As of May 3, we have found three bills introduced that fall into this category.
  2. Requiring partisan or unprofessional election “audits” or reviews. Extending a pattern from 2021 and 2022, legislators are considering amorphous post-election review schemes outside the professional standards of traditional post-election audits. These unprofessional reviews could promote subversion and needlessly call election outcomes into doubt. Some bills under consideration seek to review previous elections; others would introduce “forensic audits”—a term with no established meaning among election experts—for elections in the future. Other “audit” bills provide pathways for representatives of political parties to request or initiate audits, which can become a tool to delay or cast doubt on the validity of election results. As of May 3, we have found 25 bills introduced that fall into this category.
  3. Seizing power over election responsibilities. These bills would shift election administration responsibilities away from professional, nonpartisan officials and toward partisan actors in the legislature. In a growing and worrisome trend, some bills would block election administrators from complying with federal law or federal funding requirements without the approval of the legislature—perhaps forcing those administrators to choose between complying with state and federal law. As of May 3, we have found 31 bills introduced that fall into this category.
  4. Creating unworkable burdens in election administration. These bills would interfere with the basic procedures of election administration, increasing the risk of chaos and delay and enabling specious claims of irregularity that could be a pretext for subversion. Accelerating trends in this category include bills that would block alternative sources of funding for already strapped election offices, bills that would mandate burdensome and error-prone hand counts of ballots, bills that would enable efforts to swamp election administrators with records requests, and bills that would hamper the ability of election administrators to ensure safe in-person voting. As of May 3, we have found 104 bills introduced that fall into this category.
  5. Imposing disproportionate criminal or other penalties. These bills would create or expand penalties for election officials in the ordinary execution of their jobs, including criminalizing inadvertent mistakes. Many of these bills would heighten the atmosphere of unwarranted suspicion that already surrounds election officials; some would create entirely new investigative bodies or law enforcement agencies and ratchet up fear among election officials. As of May 3, we have found 73 bills introduced that fall into this category.

At the same time, election-subversion threats have continued outside state legislatures. At least eight states have withdrawn from the Electronic Registration Information Center (ERIC), an organization that helps state election officials maintain accurate voter rolls and prevent illegal voting. In Alabama, Missouri, and West Virginia, the ERIC withdrawal was initiated by a secretary of state who is an election denier. Until recently, this organization was supported by state election officials from across the country and the political spectrum, but it has been targeted by a relentless disinformation campaign. In addition, states face a “brain drain” as nonpartisan election administrators, hampered by chronic underfunding and hounded by threats and harassment, leave their jobs.

Finally, in the 2022 election cycle, local election officials in at least five states refused to certify accurate election results.

Each of those states’ results were ultimately certified, sometimes after a court order, but these rogue refusals highlight a discouraging trend and a danger to what should be routine practice — the lawful affirmation of the will of the people.



We released the first version of this Report in the aftermath of the 2020 election. At that time, we were just a few months removed from a violent attempt to overturn the results of the election and were just beginning to recognize the ongoing and substantial role that election conspiracy theories and election denialism would play in our political system. In short, it was becoming increasingly evident that the attempt to subvert the 2020 election was not an anomaly in the history of our democracy but rather part of a sustained movement to introduce confusion and disruption into the election process and undermine confidence in our elections.

It soon became clear that state legislatures would be important in determining the health of our democracy — capable of mitigating or exacerbating the subversion threat. In our initial Report, we described a disturbing trend of state legislatures attempting to change election law in ways that would undermine our election system and make it far easier for partisan actors to interfere with or manipulate the results — or even to overturn the will of the voters outright. Over the last two years, as we have released new versions of this Report and interim updates, that trend has evolved, but it has not abated. During the 2021 and 2022 state legislative sessions, we tracked nearly 400 legislative proposals that we assessed would, if passed into law, increase the risk of election subversion directly or indirectly.

Reflecting on the 2022 midterms and the state of American democracy, there certainly have been some positive developments that give reasons for optimism. In the 2022 midterm elections, voters in critical states overwhelmingly rejected election denier candidates who were running for statewide offices with some responsibility for administering or certifying elections. And, with a few notable exceptions, those losing candidates quickly conceded defeat. Furthermore, in several states that were hotbeds of legislative activity that would have increased the risk of election subversion, the composition of the legislature has changed such that their new majorities reject election denialism. As we said in our December 2022 Report update: “the decreased likelihood of subversion legislation being passed and the decreased likelihood of state executive branch officials engaging in subversion even if the legislature encourages it…reduce the risk of election subversion in the 2024 election.”

Many of the bills that we have tracked in previous versions of the Report have not become law. And of those that have — a total of 56 new laws enacted in 26 states in 2021 and 2022 — none led to a full-on crisis in the midterms, at least from a subversion perspective. (Whether recently enacted legislation had an impact on the ability of voters to cast their votes is outside the scope of this Report.)

Despite these positive developments, we believe that tracking legislative proposals that would increase the risk of election subversion is imperative for at least five reasons.

First, the election denial movement is not going away. Disinformation about our elections continues to flood the system, and a significant percentage of Americans subscribe to election conspiracy theories. Some of the most notable purveyors of disinformation remain prominent in power and in our politics — including in statehouses, Congress, and state party leadership. And notably, former President Trump — who is running for president again in 2024 — continues to reject the results of the 2020 presidential election. Disinformation and conspiracy theories are providing fuel to state legislatures as they continue to propose (and in some cases pass) the type of legislation that led us to begin tracking this activity more than two years ago. As detailed in this Report, legislatures in 38 states have proposed 185 bills that, if enacted, would increase the risk of election subversion in 2024 and beyond. So the trend that we have documented has not yet run its course, and it is far too early for those of us who care about the health of our democracy to declare victory.

Second, although most election deniers were unsuccessful in their bids to take over election administration in some key states, they did win statewide office in a handful of less closely watched states. And many election deniers were elected to state legislatures, to Congress, and to local positions with some control over election administration (such as county and local clerk positions). Even without new legislation, we can expect that some of these election deniers will abuse their authority in ways that we have already seen and that increase the risk of subversion, such as by engaging in irregular hand counts or unprofessional “audits,” empowering aggressive poll watchers, or refusing to certify election results (especially at the county level). This problem could be exacerbated by some of the legislation tracked in this Report.

Third, although laws enacted in 2021 and 2022 did not lead to a crisis in the 2022 election, that does not mean that they had no effect at all. To the contrary, there is good reason to believe that some of the challenges in administering the 2022 election in certain jurisdictions were attributable at least in part to new burdens placed on election administrators without a corresponding increase in staffing or financial resources. And laws increasing the risk of election subversion remain on the books for 2024 and will undoubtedly be joined by others described in this Report. With each election, there is a risk that these laws will be used by anti-democratic actors to manipulate the results or overturn the will of the voters.

Fourth, even if many of the proposals described in this and previous reports never become law, it is important to understand how subversion legislation propagates (and repropagates) disinformation cycles. To this day, a significant percentage of Americans believe false information about the 2020 and 2022 elections. This disinformation cycle sustains the election denial movement, which leads to more threats and propels the resignations of election officials. Legislation continues to give legitimacy to baseless claims of election fraud and those promoting and profiting off the lie.

Fifth, the legislative proposals tracked in this Report also reveal how trends in the election denial movement are translating into action. For example, before the 2022 election, we recognized an uptick in legislation that would require all ballots to be counted by hand, driven by debunked conspiracy theories that election equipment used to count votes was rigged against Donald Trump in 2020. Although none of these bills became law, hand-counting of ballots did indeed become an issue during the midterms. We expect to see some of the legislative trends described in this Report become an issue as the 2024 election approaches. One trend worth keeping a close eye on is bills that would require a new election if certain conditions are met.

The election-subversion trend has drawn significant attention in academic and media circles in the last two years. A number of organizations, in addition to ours, also track proposed legislation that would increase the risk of subversion, among them the Voting Rights Lab and the Brennan Center for Justice.

This scrutiny has made a difference, most importantly by better informing the electorate about the trend, so that voters incorporated it into their decision making in the 2022 midterm elections.

In addition, the recent focus on election subversion has led to a renewed interest in supporting and improving the nation’s election administration personnel and infrastructure. A recent report by the National Task Force on Election Crises lays out a number of positive steps that state and federal legislators can take both to forestall subversion efforts and to improve elections. Its recommendations include ensuring consistently adequate funding for election equipment and personnel, updating election statutes to deal with emergency situations in more orderly and depoliticized ways, and making vital changes to procedures to conform with the revised federal Electoral Count Act governing the upcoming presidential election.

Defining Election Subversion

Given the wide interest in election subversion from the media, researchers, and lawyers, it is not surprising that one can find varying definitions of “election subversion” (also sometimes referred to as “election sabotage”). This Report has always focused on legislative initiatives that increase the risk that the purported outcome of an election does not reflect the choice of the voters. As we have explained in previous Reports, although a few of the bills that we have tracked would explicitly allow state legislatures or other actors to overturn the will of voters — what we sometimes refer to as direct subversion — the vast majority do not.

Bills that indirectly make subversion more likely are far more prevalent. A more probable scenario is a relatively close election, followed by efforts to create confusion and doubt about the results. Partisan actors could then claim that the true will of the voters cannot be determined, and engineer the outcome of their choice. That being said, as discussed later in this Report, this legislative session has seen an increase in bills pushed by election deniers that would nullify election results if certain conditions are met. These bills are closely related to some of the direct subversion bills that we’ve seen in the past, in that they would allow the will of the voters to be disregarded.

For the last two years, this Report has focused on five particular ways that the risk of subversion is exacerbated by legislative proposals.27 Election outcomes that respect the democratic will of the voters are threatened by laws that require or enable: (1) partisan legislative usurpation of control over election results; (2) partisan or unprofessional election “audits” or reviews; (3) partisan legislative seizure of power over election responsibilities; (4) the creation of unworkable burdens in election administration; and (5) the imposition of disproportionate criminal or other penalties around elections.

As of May 3, 2023, we have identified 185 bills of concern in 38 states. A total of 15 bills have been enacted or adopted and three have been vetoed (with several states still in the midst of their legislative sessions).

This is roughly on par with the number of measures we identified in 2021 and 2022 at approximately the same point in state legislative calendars.

Part I. How State Legislatures Are Increasing the Risk of Election Subversion

Part I. How State Legislatures Are Increasing the Risk of Election Subversion

A. Usurping control over election results

In the previous legislative cycle (2021-2022), a handful of states considered and ultimately rejected bills that would have given legislators more direct control over election outcomes, allowing lawmakers to reject the will of the voters. This year, legislators have introduced three bills that could provide state and local officials with similar opportunities to usurp election results.

For example, the Texas legislature (with Republican majorities) has proposed multiple bills that would give statewide officials greater control over larger counties — functionally creating separate rules for Harris County, which includes Houston, where a concentration of the state’s Democratic voters reside. H.B. 5082 allows the secretary of state to order new elections in counties with a population over 1 million (which applies to only six of the 254 counties in Texas, including Harris County) when there is “good cause” to believe that 2 percent of polling places ran out of usable ballots during voting hours and did not receive replacements within an hour. Texas S.B. 1993 was originally a companion bill to H.B. 5082 but was revised in committee to provide the secretary of state with this authority only in counties with a population of 2.7 million or more (which only applies to Harris County). Both of these bills create routes for the secretary of state, who is appointed by the governor and is affiliated with one party, to order new elections in certain counties (including some where voters predominately support another party). This structure can give rise to partisan manipulation and subversion of election results.

Arizona is considering its own bill that would also empower a potentially partisan statewide official to take over local election administration on flimsy grounds. S.B. 1695 prohibits local election officials, such as the board of supervisors, county recorder, or county election officer, from canvassing the election if an election official has violated any election law (even minor violations of state law or a provision of the Election Procedures Manual) and if voters submit affidavits alleging disenfranchisement. If a court receives the requisite number of affidavits, it must order the canvass delayed and appoint a special master to review the affidavits and examine the facts. If the court declares a “failed election,” the court can order a new election. One grounds for alleging disenfranchisement under S.B. 1695 is if a voter waits more than 90 minutes outside a voting location. Although requiring election officials to follow election laws is essential (and preventing voters from waiting too long to cast ballots is a valuable goal), this law is structured in a way that creates a pretext for a special master to take over from regular election administrators and potentially overturn the will of the voters

B. Requiring partisan or unprofessional election “audits” or reviews

After the 2020 presidential election, many state legislatures launched expansive (and expensive) efforts to reexamine the 2020 election results, with the aim of propagating the false narrative that the election was stolen from former President Trump. In the previous legislative cycle (20212022), many states considered legislation to institute new, amorphous or partisan reviews or “audit” schemes for future elections, which could promote election subversion. This style of “audit” stands in sharp contrast to the professional audits that many states employ to detect any irregularities or discrepancies in the results. These states conducted their standard audits after the 2022 election, which resoundingly confirmed the integrity and accuracy of the election results.

As of May 3, 2023, state legislatures have considered 25 bills that would shift some election administration responsibilities to the control of the legislature. In the 2023 legislative session, “audit” bills have taken a variety of forms. There are still backward-looking bills seeking to audit previous elections:

  • Arizona: S.B. 1471 requires Maricopa County and permits Pima County and Pinal County to retabulate ballots from the 2022 election.
  • New Hampshire: H.B. 599 authorizes a “full forensic audit” of the November 2022 election.
  • Virginia: S.B. 1316 and S.B. 605 authorize a 2020 election “audit.”

Other bills expand which entities can request an election “audit” and introduce “forensic audits” into state statutes. A “forensic audit” of election results is a vague concept that does not have a recognized meaning among election experts and official auditors. The risk of these proposals is that they will lead to “audits” that do not adhere to best practices and instead are fuel for disinformation and distrust of election results.

  • Alaska: H.B. 132 and S.B. 1 allow the Legislative Council, which is composed of 14 state legislators, to “contract with and appoint technical subject matter experts to conduct full forensic audits of election data, algorithms, software, and equipment, including precinct tabulators, storage devices, voting machines, and vote tally systems.”
  • Arizona: H.B. 2078 allows candidates, chairs of county political parties, and chairs of ballot measure political committees to request a post-election investigation that can culminate in an audit by the secretary of state. The legislation does not set forth in statute a standard or basis that these individuals must have before initiating this request.
  • Indiana: S.B. 262 requires the secretary of state to enter into contracts with an independent “forensic” imaging company to complete a full “forensic audit” of certain elections.
  • Minnesota: H.B. 1952 and S.B. 905 allow political parties to request a “forensic audit” after the statewide canvass, choose whether to request a specific or general audit, and be consulted in the selection of the independent organization conducting the forensic audit.
  • New York: S.B. 2335 and A.B. 876 shift the standard for requesting a voting machine audit from the full Board of Elections, whose members are equally bipartisan, to allow just one board member to make the request. Accordingly, such an audit can be pursued with the unilateral support of one political party.
  • South Carolina: H.B. 4259 grants county political party chairs and governing bodies of each county the authority to investigate the election with a “forensic audit,” a full hand recount, and a post- election audit of paper ballots for 22 months after the election.
  • Texas: H.B. 589, similarly to the Arizona bills described above, allows certain individuals (candidates, county political chairs, presiding or alternative presiding judges, or ballot measure political committee members) to request a post- election investigation that can culminate in an audit by the secretary of state, without setting forth in statute the initial standard or basis these individuals must have for initiating the request. The secretary of state can also issue civil penalties against county clerks under this law.
  • Virginia: S.B. 1316 and S.B 605 allow for a “forensic audit” at the request of “any elected official representing individuals in that county or city or of any election official who worked in the election in question or upon a petition signed by at least 1,000 residents of the county or city.” A jury of randomly selected residents is then presented with the audit results and has the power to declare the “validity or invalidity” of the election based on the evidence, which will void “invalid” election results.

C. Seizing power over election responsibilities

In 2023, 14 state legislatures have considered 31 bills that would shift some election administration responsibilities to the control of the legislature as of May 3. These efforts undermine professional and nonpartisan election administration and potentially empower partisan actors to interfere with the process or create chaos and uncertainty. We have detected several persistent trends from 2020 through today in how legislatures attempt to alter the balance of power from professional executive branch officials to the more partisan legislative branch.


In 2021, one of the earliest signals that the subversion tide was rising came from Georgia, where the state enacted several bills that, among other things, enabled partisan takeovers of county election boards through statewide legislation as well as bills restructuring specific counties’ election boards. This year, the legislature passed additional bills that restructure six county election boards: Ben Hill (H.B. 736); Cherokee (H.B. 642, H.B. 644); Columbia (H.B. 730); Schley (H.B. 710); Screven (S.B. 277); and Ware (H.B. 422). Although the restructuring of county boards is not inherently inappropriate and these bills appear neutral on the face of their statutory text, the track record from last year’s bills — combined with the immediate impact of this year’s bill targeting Ware County — demonstrates that these bills have provided pathways for altering the racial composition and skewing the bipartisan distribution of the members of the boards.

This year, other state legislatures are following Georgia’s example. In Texas, a number of proposals target specific counties and Harris County (which includes Houston) in particular. One, S.B. 1933, allows the secretary of state to impose administrative oversight of a county office if a complaint is filed with them and if the secretary has good cause to believe that there is a recurring pattern of problems. In Pennsylvania, S.B. 603 would put a board with a substantial number of legislatively appointed members in charge of creating an elections manual for the counties, which can limit the discretion of election officials. And it would then require that the manual be submitted to the legislature.


One related cluster of proposals would give legislatures authority over how election officials implement relevant federal election statutes or would assert that legislatures have power to approve or reject rules and regulations issued by state executive branch officials. In Kansas, S.B. 260 requires that the Legislative Coordinating Council review all proposed rules or regulations regarding vote canvassing and poll closing procedures. The legislature can then revoke or reject the rules or regulations at any time. In Arizona, disputes over the secretary of state’s manual on election procedures have been ongoing since 2020, and boiled over into litigation in 2022.37 As a result, the 2022 election was conducted without the benefit of a revised authorized version of the manual with up-to-date, detailed instructions on how counties were to conduct elections.38 In 2023, the Arizona legislature is considering a proposal that prolongs the unrest regarding the manual. Arizona S.B. 1213 gives the legislature’s Joint Audit Committee the power to review, approve, or reject the manual. One newer trend in this vein has emerged this year. Four states are considering virtually identical proposals that would bar election administrators from implementing federal election laws or directives or accepting federal funding without legislative acquiescence. Of course, state election officials must comply with federal laws regardless of the state legislature’s actions. These proposals would place state election officials in the untenable position of violating state law in order to comply with federal law.


Other proposals would have the legislature itself conduct some election tasks. For example, Alaska S.B. 1 and H.B. 132 both would have the Legislative Council appoint people to conduct full “forensic audits” of virtually all election equipment, including election data, algorithms, and software. And in Montana, H.B. 905 would have created an “election security team” with half of its members appointed by legislative leaders. The team would have been tasked with conducting a full hand count of elections within one year of each election, thereby providing partisan legislators the means to co-administer a crucial election task. Similarly, legislatures’ ability to approve of or participate in election-related litigation has been a frequent subject of proposed bills. A number of proposals this year would inject legislatures into lawsuits, allowing them to stop professional election administrators from overseeing and setting litigation priorities and settling cases.


Finally, multiple states have proposed bills that would constrain the ability of election officials to respond to or adapt to emergencies. In the 2021-2022 legislative session, these proposals were relatively common as legislatures reacted to the many ways election officials dealt with the COVID-19 emergency. Thus far in this legislative session, several legislatures have considered proposals that would tie the hands of election administrators to respond flexibly to exigent circumstances. These, if adopted, could magnify the risk of election chaos if an emergency were to occur. In some cases, the constraints are broad. For example, in Indiana, H.B. 1505 prohibits the governor from making any changes to elections in the midst of a declared emergency. Others allow some flexibility while limiting officials’ emergency response. For example, in Kansas, under one proposal (S.B. 262), the secretary of state would be barred from making any determination or issuing any rules or regulations in the 30 days before an election unless the legislature has approved an emergency.

D. Creating unworkable burdens in election administration

A trend that gained popularity in 2022 and continues through 2023 is legislation that makes elections more difficult to administer by imposing new unworkable burdens on local election administrators. While the legislative proposals vary, the through line is that these proposals would increase the risk of chaos and delay, which opens the window for specious claims of fraud or irregularities that could serve as a pretext for election subversion.

As we’ve noted in our past reports, although it is appropriate for state legislatures to set the rules for an election, it is not appropriate for legislatures to intrude into the minutiae of election administration in a way that makes elections unworkable and thus renders the results far more difficult to finalize. These proposals are typically fueled by disinformation about election administration and voting procedures, rather than by expertise in professional, nonpartisan election administration.

Thirty-one state legislatures have proposed or passed 104 bills this legislative session in this category. This category includes bills that would limit the ability of election officials to address funding shortfalls by accepting nonprofit or other funding to support election administration; constrain election officials’ ability to respond to partisan poll watchers who harass or intimidate election officials or voters; and require all ballots to be hand counted, which would dramatically increase the error rate and create extensive delays. We have also seen the emergence of a new subcategory this year: bills that would expand the ability of election deniers to overburden local election offices with an excessive volume of records requests and demands for public inspection of ballots.


Since the 2020 election, 24 states have enacted legislation that targets third-party sources of funding for election administration — including banning it completely. This trend started after the Center for Tech and Civic Life, a nonprofit that works to modernize U.S. elections, granted opt-in philanthropic funding for local election administration to local election offices so they could safely administer elections during the COVID-19 pandemic. This funding went to jurisdictions representing voters across the political spectrum, with 2,500 election departments receiving funding across 47 states.

Supporters of the ban have argued, contrary to the evidence, that the private money donated to local election offices improperly influenced electoral outcomes. These efforts have been fueled by false narratives about the 2020 and 2022 elections and come as election officials raise the alarm about inadequate election funding.

Although it would be better if private funding for election administration were not necessary, U.S. election infrastructure is chronically underfunded, often forcing election officials to stretch their limited resources to run safe, smooth, and secure elections. Outright bans on alternative sources of funding, without accompanying increases in state or federal investment, lead to increased risk of crises because elections are harder to administer without proper funding to purchase and maintain adequate voting equipment, hire and train personnel, and provide a safe and secure environment for those personnel and voters.

For example, this year Georgia passed into law S.B. 222, which makes it a felony for local election workers or government officials to solicit or accept private funds to cover the cost of running elections. While Georgia lawmakers had already passed a ban on county election boards’ directly receiving outside funding in 2021, lawmakers took this additional legislative step after DeKalb County received a $2 million grant in early 2023 that the county subsequently allocated to its elections office. Arkansas, Idaho, and Montana recently enacted similar bans.


Starting in 2022, we began to see statehouses consider proposals requiring hand-counting of ballots in all elections, based on false claims that voting machines were used to manipulate the outcome of the 2020 election. This legislative trend has increased in 2023. Although targeted hand counts can be an important tool for post-election recounts and audits, mandatory hand counts of larger jurisdictions, or even entire states, would be nearly impossible to administer, dramatically less accurate than machine tallies, and ripe for creating delay and sowing chaos. Indeed, experts have found hand counts to be both extremely slow and error-prone. In addition to introducing errors, the extreme delay itself creates a risk of election subversion, especially if it means the results will not be certified by the relevant statutory deadlines.

In several states, bills such as Arizona H.B. 2307, South Carolina H.B. 3162, and Virginia S.B. 884 have been introduced that would ban the use of machines to tabulate votes and instead mandate hand counts. Arizona H.B. 2100 would require all ballots to be counted by hand with returns posted within 24 hours after the polls close, which experts assess would be a physical and technical impossibility.

Moreover, restrictions and outright bans on voting machines could be costly for taxpayers. In Shasta County, California (approximate population 182,000), a push to replace voting machines with a hand-count system could cost an additional $4 million over two years. In Texas, the 2021 election overhaul bill S.B. 1 mandated the state purchase voting technology that does not yet exist, which could arguably force some counties to hand-count ballots.


In general, observers and poll watchers, whether nonpartisan or affiliated with a political party, are common and appropriate at in-person voting locations. They can ensure that election procedures are followed, report problems to voting rights hotlines, and improve confidence in elections. As standard practice, their conduct and access is circumscribed to ensure that they do not interfere with or intimidate voters or otherwise hinder election administrators as they do their jobs to ensure a safe and secure voting environment. But state legislators have introduced bills that would effectively create more opportunities to harass poll workers and voters alike. In Pennsylvania, S.B. 516 would expand who may serve as poll watchers by allowing anyone registered to vote in Pennsylvania to serve anywhere in the state. In the 2020 election, election deniers spread disinformation about cities like Philadelphia that are either majority Black or have large Black populations, questioning the legitimacy of their votes, which caused some tense moments at the central count facilities. If this bill were to become law, one can imagine a situation where out-of-town poll watchers would target cities like Philadelphia, observing elections with preconceived notions of fraud that could lead to disruptive behavior

The discretion to make on-the-ground decisions responsive to real-time, specific issues at voting locations has traditionally been vested with local election officials — including by requiring observers to stop certain behaviors, restricting their movement, or ejecting misbehaving observers. But legislation in West Virginia, Oklahoma, and South Dakota would disempower election officials from removing disruptive poll watchers and would increase the penalties for doing so.

E. Imposing disproportionate criminal or other penalties

Seventy-three bills introduced or enacted so far in 2023 in 26 state legislatures would impose new criminal and civil penalties on election officials for routine activities or minor errors that normally would not result in serious punishment. These proposals increase the risk of subversion by making it harder for election officials to properly administer elections and provide a safe environment for staff and voters, given their fears of legal repercussions. This is one factor driving unprecedented numbers of experienced election officials to leave their posts since the 2020 election, compounded by increased harassment and threats (see Section IV, “Subversion beyond the statehouses”).

We’ve identified two trends within this category of legislation. The first trend is proposals that create new penalties or expand existing penalties for election officials performing their jobs and for the minor, inadvertent mistakes that can occur in that process. The second trend is proposals to create new investigative bodies or mechanisms that over- criminalize election administration. Although criminal and civil penalties are sometimes warranted for improper behavior, the proposals we have identified dramatically escalate the criminalization of election administration with no apparent need or benefit. When new legislation creates a fear of criminal or civil penalties, the expense of legal fees for defense, exorbitant fines, and so on, seasoned election officials are more likely to join hundreds of their peers in leaving the field altogether — which will truly imperil the administration of our elections.


Some bills introduced and enacted in 2023 expand or increase civil and criminal penalties, which are particularly concerning when they create significant criminal or civil liability for vaguely defined conduct. For example, Indiana H.B. 1336 creates a chilling effect by allowing absentee board members to be assessed a penalty, and potentially removed, if they fail to perform any of their prescribed duties, regardless of the scope or magnitude of the failure.

Other bills create criminal or civil penalties for attempting to comply with federal laws. For example, Missouri S.B. 235 requires local election officials to give 30 days’ notice to the General Assembly before implementing new federal election guidance and further requires the General Assembly to approve the new guidance. A local election official faces a $5,000 fine for implementing federal guidance that has not been approved by the legislature. This would put election officials in an impossible scenario under which they would potentially violate federal law if they fail to implement federal guidance, or violate state law and incur a $5,000 fine.


This second trend includes proposals that increase law enforcement’s role in elections, escalate surveillance of election officials by establishing procedures for burdensome post-election investigations, and create investigative units specifically to pursue supposed election violations. Without evidence that additional enforcement or penalties are necessary, these provisions feed conspiracy theories that American elections are not securely administered. They also create significant costs, in both time and money, for election officials and others charged to defend themselves from significant liability.

Some bills expand the role of law enforcement in elections and create a labyrinth of post-election investigations. For example, Arizona H.B. 2078 would create a Byzantine process through which candidates, party chairs, and others could request an explanation and documentation of any activity that “appears to violate the statutes” and require response from the election official within 20 days of the initial request. Similarly, in Texas, H.B. 5234 would establish a procedure under which elections could not be certified before a 30-day period, explicitly including time for any individual voter to file a suit to resolve an undefined “error” and remove officials who refuse to correct said “error.”

We have already seen the resulting chaos from a surge in public records requests following elections, which frequently seek large volumes of records or require significant staff resources to search, compile, and produce them. These requests overwhelm election officials (who are often volunteers), fuel baseless claims of voter fraud, and even threaten election security as potentially sensitive information about voting systems is made public. Making it easy for bad-faith actors to request months’ worth of documentation may make it nearly impossible for these officials to perform their jobs, and may even drive them to leave their positions. Officials may have to choose between defending themselves from spurious requests, incurring significant fines and penalties for not responding, and continuing to administer safe and secure elections.

Several bills that are part of this trend include provisions authorizing completely new law enforcement units to investigate election-related crimes. For example, Arkansas H.B. 1513 establishes an “Election Integrity Unit” within the office of the attorney general. This unit is based on Florida’s Election Crimes Task Force, which has already been subject to enormous controversy for failing to produce any evidence of the scale of election crimes that it promised to uncover. In Florida, S.B. 4 will expand the jurisdiction of statewide prosecutors to make it easier to bring cases at the state level with a more specific list of election-related crimes, despite the fact that 0.000677 percent of voters in the state are even suspected of committing election fraud.

Given the extreme rarity of election law violations, these increases in law enforcement’s scope and resources are unnecessary. And if the new units choose to focus their efforts on election officials under new laws criminalizing various aspects of election administration, the threat of prosecution and unfounded litigation will likely lead even more conscientious officials to leave their jobs.

Part II. Subversion Beyond the Statehouses

Part II. Subversion Beyond the Statehouses

While this Report’s primary focus is on legislative trends that increase the risk of election subversion, it is also worth noting some non-legislative trends, as they are fueled by the same conspiracy theories and actors. We describe several non-exhaustive examples to highlight the threat posed by these actions and how, when coupled with similar legislative efforts, they can increase the subversion risk and contribute to deteriorating confidence in the electoral system.

A. Disinformation leads to ERIC withdrawals

A particularly concerning trend in 2023 has been states withdrawing from the Electronic Registration Information Center (ERIC). Governed by its member states, ERIC is a “nonprofit, nonpartisan membership organization created by and comprised of state election officials” to help election officials across the country maintain accurate and complete voting rolls and help prevent and detect illegal voting. Until recently, ERIC was widely supported by state election officials from a geographically and politically diverse cross-sector of states. However, after the program was targeted by disinformation campaigns falsely alleging that the organization is funded by George Soros and is an effort to register more Democratic voters (among other false claims), eight states withdrew from ERIC. Additional states, such as Alaska and Texas, are also considering withdrawing from ERIC, while California is considering joining.

The decision of each of the states to withdraw from ERIC has led to widespread, bipartisan criticism from election officials in other ERIC-member states. Georgia’s Republican secretary of state, Brad Raffensperger, condemned these states, saying that in “[r]eacting to disinformation they’ve hurt their own state & others while undermining voter confidence.” Republican Utah Lieutenant Governor Deidre Henderson, who serves as the state’s chief election official, also voiced support for ERIC, stating it had “served Utah and its member states well.” Democratic Maine Secretary of State Shenna Bellows “described the work ERIC does as ‘technical and boring’ but an important part of the ‘backbone’ of American elections.” She also expressed concern about the departures being a “product of disinformation” that has “made ERIC a lightning rod in some circles.” Indeed, several of the officials from withdrawing states recently cited ERIC as being vital to secure their elections.

While 22 states and the District of Columbia remain in the program, this trend is worrisome because, as each departing state pulls its voter data from ERIC, the remaining member states have fewer records to cross-check against their own voter rolls. This means that both in the 2024 election and beyond, election officials could be missing an additional tool in their election administration and enforcement toolbox. Although some departing states have pledged to use or create a similar tool, no such alternative exists, and a previous attempt to create one failed badly. Moreover, these departures demonstrate how quickly disinformation can fuel abandonment of well-respected and established election administration procedures.

B. Brain drain of election officials

Professional election administration by election officials at all levels of government has been critical to ensuring the integrity and accuracy of America’s elections. Election officials across the country have endured significant challenges in recent election cycles, including but not limited to shortage of election workers, cyber security concerns, insufficient financial resources to strengthen and modernize election infrastructure, harassment and threats, and misinformation about their work and role in administering elections.

As the nation heads into the 2024 election cycle, in which races will range from the local election clerk to the President, the country faces the prospect that a significant number of election officials will have recently left their posts or will leave during this election cycle. An April 2023 survey by the Brennan Center found that 11 percent of current election officials are “very or somewhat likely to leave” their posts before the November 2024 election. As a result, this pivotal election cycle will be administered by many new or interim election officials, who may face additional hurdles while having less experience in election administration.

  • The chief election official in Texas’ third most populous county, Heider Garcia, announced in April 2023 that he would leave his post after five years of service. He has been subjected to harassment and death threats since the 2020 election. This resignation came on the heels of the county’s new election judge — who has trafficked in 2020 election misinformation — debuting an election integrity task force.
  • Two months before the 2022 midterm election, all three election officials in the Texas county of Gillespie (population approximately 27,000) departed, with one citing the death threats she received in the 2020 election cycle as a contributing factor. As of April 2023, the county has yet to hire a permanent replacement.
  • After the election administrator in Harris County, Texas’ most populous county, departed in March 2022, her appointed successor took over in late August 2022 and 10 weeks later administered the November 2022 election. After receiving partisan criticism for his work, his office is now the target of state legislation (S.B. 1750) that would eliminate it.
  • In Hood County in North Texas, an election administrator with 14 years of experience, Michele Carew, resigned because of hounding by supporters of former President Trump and moved into the private sector. Carew was brought into this role just two and half months before the 2020 presidential election and received “assurances from the Texas secretary of state that she was complying with Texas election rules.” Despite Trump winning the county with 81 percent of the vote, “the local GOP executive committee issued warnings to Republican officials who defended Carew.”

The significant number of departing election officials raises the question of who will serve as their replacements. The county election judge in Tarrant County, Texas, has not ruled out considering an election denier for the newly vacated position. In Cochise County, Arizona, an election administrator who had served five years left after refusing to conduct the illegal hand count of the November 2022 election, which was demanded by the board of supervisors.m The board hired a replacement who has promoted false voter-fraud and voting-machine claims on his social media accounts.

Less experienced and under-resourced election officials may not have the depth of experience to quickly catch mistakes. For example, in Pinal County, Arizona, the new county election director took over an office in March 2022 that employed only one staffer when there should have been five full-time employees. During the election four months later, a ballot-printing problem resulted in municipal races’ being omitted from 60,000 primary ballots across seven communities. This was a result of a staff error that went undetected for too long. The combination of experienced election officials departing their posts in significant numbers, the challenges of hiring qualified, nonpartisan replacements, and the lack of sustained financial investment in elections, as well as complex or vague new laws regulating the duties of election officials, may pave the way for election subversion in future election cycles.

In Cochise County, Arizona, the election administrator who had served five years left after refusing to conduct the illegal hand count of the November 2022 election demanded by the board of supervisors. The board hired a replacement who has promoted false voter fraud and voting machine claims on his social media accounts.

C. Local officials’ refusal to certify election results

The long cycle of an election concludes when the official tally of votes is certified as the state’s conclusive election results. ProPublica’s review of the 2022 election revealed that local election officials in five states — Arizona, Nevada, New Mexico, North Carolina, and Pennsylvania — refused to certify the election results. For example, in North Carolina, two local election officials in Surry County questioned the legitimacy of the state’s election laws, court decisions on election procedures, and the outcome of the election. In Arizona, county supervisors from Cochise County refused to certify the 2022 election results, despite repeated warnings that such a course of action would be illegal. Members of the Board of Elections in Luzerne County, Pennsylvania, also initially refused to certify the election results. Notably, in many instances, state authorities did not pursue official consequences against these local officials who refused to execute their duties, although others were held accountable.

Although the statewide results in each of these states were ultimately certified, sometimes by judicial order, these refusals demonstrate a concerning trend of local election officials abdicating their ministerial duties to certify the result. Even when these refusals are ultimately resolved without creating a major crisis, they cause unnecessary delays in finalizing election results, with serious repercussions. Delays in certifying official results can create an opportunity for misinformation in the post-election period. And it takes significant financial, legal, and judicial resources to compel local officials to conduct their official duties, even where there are clear laws requiring them to certify results.

For the presidential election specifically, last year’s passage of the federal Electoral Count Reform Act updated the statutory framework for submitting the state’s presidential vote outcome to Congress.119 Accordingly, states now face strict timelines for issuing and transmitting the Certificates of Ascertainment. Delays and barriers to certification of a state’s election results could affect whether its Electoral College votes are counted during the January 6, 2025, joint session of Congress. Curbing these certification delays and refusals will be key to ensuring that the will of the voters is respected and accepted.

Part III. How States Are Proactively Safeguarding Against Election Subversion

Part III. How States Are Proactively Safeguarding Against Election Subversion

In a heartening trend, more and more states are proactively safeguarding against the risk of election subversion. Legislators are introducing bills that strengthen democracy, some of which have already passed or are likely to pass. And non-legislative actors are using their positions to similarly push back against the continued efforts of would-be election subverters.

A. Protections for election workers

Many states are proposing and enacting additional protections for election workers. As we discussed above and in our August 2022 Report, election deniers and other anti-democratic activists have been driving election officials across many states to resign, bombarding them with harassment and even death threats. This “brain drain” risks leaving openings for inexperienced, incompetent, or bad-faith actors to take over election administration. Thankfully, the 2022 midterms largely passed without major election-subversion crises. Nevertheless, lawmakers across the nation have taken note of the threats and challenges faced by election workers and are acting to curb them.

Between April 2022 and May 2023, lawmakers in at least 27 states have introduced measures to protect elections and election workers.

  • In California, Colorado, Maine, New Hampshire,123 Oregon, and Washington, new laws aim to protect election workers, hopefully encouraging more people to serve in these vital roles. Some of the bills protect election workers’ privacy, and others ramp up the punishments available for people who commit crimes against election workers.

This has become a bipartisan issue:

  • Republicans in Indiana, Missouri, Montana, and Oklahoma, have proposed felony penalties for people who interfere with election officials’ work.
  • In Arizona, a Republican legislator introduced S.B. 1061, which protects the confidentially of public officials’ and election officials’ home addresses. This bill received bipartisan support and has been signed into law.
  • With bipartisan support, New Mexico enacted S.B. 43, which added the secretary of state, county clerks, and their employees to the state’s election code prohibiting threats aimed at interfering with the impartial administration of elections.
  • Minnesota recently enacted structural changes to election administration and voting rights aimed at safeguarding elections from potential subversion threats. This legislation, signed by the governor in May 2023, enacted stiffer penalties for intimidating voters and election officials and knowingly spread false information intended to prevent someone from voting.The legislature has also passed H.B. 1830, which would raise penalties for intimidating and interfering with the duties of election officials.
  • In Nevada, the secretary of state introduced S.B. 406, which would make it a felony to harass, intimidate, or threaten election workers and would protect the personal identifying information of election workers from doxxing.134 The legislation has received unanimous support in both chambers.

In conjunction with other policies, such as protecting privacy and providing adequate funding, increased criminal penalties can help recognize the severity of threats election officials face. However, additional avenues for accountability can still be necessary. Creating new civil causes of action would give election officials another tool to protect themselves in circumstances that do not result in criminal prosecution or if prosecutors refuse to hold bad actors accountable for political purposes. While imposing penalties can serve the admirable goal of protecting election workers, they should be deployed thoughtfully and carefully, so as not to accidentally criminalize benign behavior.

B. Direct responses to January 6, 2021

Some of the pro-democracy legislation making its way through statehouses explicitly responds to the violent insurrection at the U.S. Capitol on January 6, 2021, which sought to overturn the results of the 2020 presidential election. Minnesota’s legislature introduced its sweeping pro-democracy bill this legislative cycle, detailed above, on January 6, 2023. Legislators used this anniversary to highlight the ongoing threats to our democracy, and the need for stronger efforts in its defense. In New York and Pennsylvania, legislators have introduced measures setting up annual remembrances of the insurrection on January 6.

  • The New York bill explicitly acknowledges that President Trump and others “directly incited and encouraged an armed and violent insurrection against the government of the United States, with the express purpose of preventing the peaceful transfer of power and overturning the results of a free and fair election.” The bill then declares January 6 as Democracy Day, in honor of the people who were killed or wounded defending the Capitol” and recognizes the need to “strengthen our democratic institutions.”
  • The Pennsylvania Senate Resolution creates a 1/6 Day that focuses on the survivors of the Capitol attack.

Both bills acknowledge that the January 6 insurrection was a threat to our democracy and the acceptance of free and fair election results, shine a spotlight on the heroes of January 6, and highlight the need to protect democracy.

Arizona’s newly elected attorney general is shifting that office’s focus toward protecting voter access. In our 2022 Reports, we highlighted Arizona as one of the states most vulnerable to election subversion. Legislators introduced the second-highest number of bills that would increase the risk of election subversion in their legislative session (behind only Wisconsin), including bills in all five of our categories. Arizona also exemplified non-legislative actions that increase the risk of election subversion, including insider threats from election deniers and harassment of election officials. Although the Arizona legislature is again considering several bills that would increase the risk of election subversion, other statewide officials are proactively taking actions to reduce the risk of election subversion.

For example, the newly elected attorney general, Kris Mayes, announced that she would shift the focus of the “election integrity unit” that was created in her office in 2019 to protect voter access. That unit previously was created to investigate voter fraud but ultimately found very little and prosecuted only 20 cases from among the state’s millions of voters. At the same time, that unit served as the basis for conspiracy theories and undermining faith in election administration. In addition, Arizona’s newly elected governor, Katie Hobbs, has already vetoed three bills that increased the risk of subversion.”


IV. Conclusion

The 2020 and 2022 elections were conducted fairly, counted accurately, and certified under the law. That is a testament to thousands of election officials who, in each state, acted in good faith and fulfilled their duties. The election system can always be improved, but the system is not broken: statewide canvasses, professional post-election audits, and countless legal decisions have validated these results over and over. The bills described in this Report would only create problems and introduce confusion into our election process.

The decisions being made in statehouses this year will help determine how the 2024 election is conducted. Unfortunately, the bills we describe in Part III of this Report lay the groundwork for any number of scenarios that could allow for election subversion. As we explained, if subversion does happen, it is unlikely to be as blatant as throwing out the election results altogether and installing a handpicked winner — although bills in Texas do allow for calling an entirely new election in a handful of targeted counties. The far likelier scenario is disorder and confusion. To take just a few examples: Fair election results could be needlessly called into doubt by unprofessional and partisan reviews. Mandatory hand counts of ballots could introduce human errors and delay results such that states sail past certification deadlines. Election workers, threatened with prosecution for even minor missteps, could be cowed from responding to poll watchers disrupting the vote. Election administrators could be left with an impossible choice between obeying federal or state law. Any of these scenarios could create a pretext for partisan actors to throw up their hands, declare that the will of the people is for them to discern, and subvert an election. Even if such a scenario does not come to pass, many of these bills are grounded in conspiracy theories and — even if they never become law — further seed the ground with disinformation and erode trust in our elections, not to mention fuel the continued threats to and harassment of election administrators.

As we noted in the December 2022 edition of this Report, the results of the midterm election lowered the risk of election subversion in 2024. In at least some critical states, the defeat of election deniers in statewide races can give voters more confidence that their choices will be respected in the next election. In addition, legislators in an array of states, including members of both parties, are moving to enhance protections for election workers and to counter the inexcusable climate of harassment they have faced because of persistent disinformation. We can take heart from these efforts to bolster protections for our democratic institutions and for the people who sustain them.

But we cannot end our vigilance. Across the country, state legislators are plainly still committed to seizing more power over the administration of elections and, directly or indirectly, more power to influence election results. Bills that would increase the risk of election subversion are showing up in state legislatures this year at roughly the same pace as during the previous two years. The threat is evolving, but it is not receding

Voters should be awake to this persistent threat. And state legislatures should focus their efforts on bolstering the nonpartisan administration of elections, protecting election officials, and respecting the will of the people. Until they do, our democracy is still far too close to a crisis.

Previous Reports

Previous Reports

Protect Democracy, along with our partners at States United Democracy Center and Law Forward, issued our initial report in April 2021, less than four months after the January 6 attack on the U.S. Capitol, itself a violent attempt to subvert the voters’ choice. Since then we’ve continued to track the introduction and passage of state election subversion legislation, updating the report in June and December of 2021, and adding a second volume to the report in May 2022.

December 2022

Across the country, state legislators are proposing and passing bills that would increase the risk of election subversion by giving partisan state legislators greater control over elections and hamstringing experienced state and local election administrators who have traditionally run our elections. These bills would increase the risk of election subversion; that is, of the outcome of an election being contrary to the will of the people.

August 2022

In the 21 months since the 2020 election, we have seen a breakdown in the longstanding consensus that election administration belongs in the hands of professional, dispassionate experts, and that naked political interference in vote counting is anathema to a functioning democracy. Over the course of 2021 and into 2022, state legislatures have embarked on a sweeping campaign to propose, consider, and, in some cases, enact measures that increase the risk of election subversion — that is, the risk that an election’s declared outcome does not reflect the choice of the voters.

May 2022

As of April 8, 2022, legislatures in 33 states are considering at least 229 bills that seek to undermine nonpartisan election administration — 175 introduced in this calendar year alone & 54 that rolled over from last calendar year. A total of 50 bills have been enacted, 32 last year & 18 thus far this year.


December 2021

As 2021 comes to a close, 262 bills have been introduced in 41 states — an increase of more than 100 bills since we began monitoring this trend in April — with 32 becoming law across 17 states.


June 2021

As of early June, there have been at least 216 bills introduced in 41 states that would interfere with election administration.


April 2021

As of April 2021, there were at least 148 bills introduced in 36 states that would interfere with election administration. The Report analyzes more than 50 of the worst bills across the country and finds an alarming pattern in what appears to be a clear reaction to the 2020 election.


Note regarding methodology

Each year, state legislators introduce thousands of bills related to elections. To create this Report, we used a combination of legislation included in the Voting Rights Lab tracker and legislative proposals that we discovered via independent research. The Voting Rights Lab tracks election-related bills under consideration each cycle.

In this report, we included legislation introduced since the start of 2023 through May 3, 2023. The States United Democracy Center, Protect Democracy, and Law Forward worked together to analyze each proposal to determine whether it would — if adopted — materially increase the risk of election subversion, and to filter out those that we concluded did not meet that criterion. A complete list of the bills that fall within the scope of this Report is included as an Appendix.

Related Content