Changes in state election laws have little impact on results, new study finds

In recent years, U.S. politics has been consumed by partisan fights over states’ election policies.

But a new study by two political scientists is causing a stir by finding that state legislators’ changes to election laws — both those that tighten election rules in the name of integrity, and those that loosen rules to expand access — have almost no impact on which side wins.

“Contemporary election reforms that are purported to increase or decrease turnout tend to have negligible effects on election outcomes,” write the authors, Justin Grimmer and Eitan Hersh, political scientists at Stanford University and Tufts University, respectively, in “How Election Rules Affect Who Wins,” which was published online as a working paper June 29.

These laws, the authors write, “have small effects on outcomes because they tend to target small shares of the electorate, have a small effect on turnout, and/or affect voters who are relatively balanced in their partisanship.”

That doesn’t mean these laws don’t matter. Many advocates, as well as the authors themselves, say there are plenty of reasons beyond partisanship to care about voting policy — not least the effect some can have on non-white voters.

“If we can take the temperature down on some of these issues and separate the partisan consequences from some of the other consequences, the public discussion would actually be a lot better,” Hersh said in a phone interview. “Right now, it seems like one of the reasons this stuff is toxic is because every minor thing, from having mail voting to having voter ID, is treated as some democracy-ending reform. And I think that’s quite dangerous.”

Indeed, Grimmer and Hersh’s conclusion, which is largely supported by other recent research, is at odds with the behavior of much of the political and advocacy worlds.

In recent years, the parties and outside groups have poured countless dollars and hours into the battles over voting, seeking to gain an electoral edge, stop their opponents from getting one, or fight voter suppression. Now, some are asking: What does the emerging consensus that these laws have minimal effects on election outcomes mean for that ongoing work?

Elections bill in House 

The study appears just as a heated debate is flaring again in Congress over the partisan and racial impact of recent voting laws.

On July 10, at a U.S. House Administration Committee field hearing in Atlanta, Republican lawmakers unveiled the American Confidence in Elections Act, new legislation that would tighten voting rules in numerous ways.

To make the case for the measure, the GOPers repeatedly criticized Democrats for predicting that Georgia’s 2021 election law, which imposed stricter rules on several types of voting, would suppress votes, especially among minorities. (“The left lied,” declared a GOP video on the issue that was shown at the hearing.)

Republicans noted that the state’s turnout in fact went up last year — though Democrats countered that Black turnout had gone down relative to white turnout.

Grimmer, a senior fellow at the Hoover Institution, served as an expert witness for Georgia in its defense of the law after the state was sued by the U.S. Department of Justice and voting-rights groups.

Meanwhile, some in the trenches of the voting wars reject the new study’s conclusions out of hand.

“Republicans are targeting the rules of voting because they know they matter,” said the Democratic super-lawyer Marc Elias — who filed the first lawsuit against the Georgia measure — in a statement to States Newsroom. “Studies of cherry-picked practices from years and decades ago may be interesting to some political scientists but they don’t solve the problem of armed vigilantes at drop boxes or states changing laws to make voter registration more difficult.”

As the conventional wisdom has it, laws that restrict access tend to help Republicans, since those most likely to be blocked or deterred by stricter rules — often racial minorities, students, renters and low-income Americans — lean Democratic. And laws that make voting easier, the idea goes, tend to boost Democrats, since the people likely to be helped by them similarly lean Democratic.

Indeed, Republican-led states have lined up to pass restrictive new voting laws, while fighting Democratic efforts to pass expansive laws. Democrats have done the reverse — including raising hundreds of millions of dollars to file court challenges to the GOP’s measures. And at election time, both sides have mobilized vast armies of volunteers to hunt for fraud, or protect voting rights, at the polls.

Politicians have been quick to blame election rules for defeats. Hillary Clinton has said, with little evidence, that between 27,000 and 200,000 Wisconsin voters “were turned away from the polls” in the 2016 presidential election because of the state’s ID requirement. Former President Donald Trump has gone much further, frequently blaming his 2020 loss on loose voting rules that, he falsely claims, enable fraud.

Advocates and much of the media have likewise prioritized the issue, seeing a chance to hold powerful actors accountable, protect or expand access to the political process, or spotlight a set of urgent challenges to U.S. democracy.

But Grimmer and Hersh describe this Sturm und Drang — at least the part that’s focused on partisan outcomes — as a tempest in a teapot.

“The caustic rhetoric that suggests the partisan stakes for election administration reform are very high is detached from empirical reality,” Grimmer and Hersh write. “Even very close elections are decided by margins larger than the magnitude of election reforms we examine in this paper.”

Little evidence of impact on results

Though that finding may surprise political operatives, advocates, and journalists, academic experts say it’s very much in sync with existing research on the issue — making the Grimmer-Hersh study much harder to dismiss as an outlier.

Scholars have struggled to find evidence that changes like early voting and election-day registration have significantly boosted turnout. (One possible exception is mail voting, where at least one recent study did find significant effects, while others didn’t.)

Nor have most studies found that even very controversial restrictive measures do much to lower voting rates. A 2019 paper released by the National Bureau of Economic Research found that strict voter ID laws “are unlikely to have a meaningful impact on turnout or election outcomes.” And a paper published this year by two Notre Dame political scientists found that ID laws “motivate and mobilize supporters of both parties, ultimately mitigating their anticipated effects on election results.”

The Grimmer-Hersh study tries to clarify why the partisan effects are so negligible. Unlike most earlier studies, it doesn’t look only at one type of law — voter ID laws, for example — but rather on the entire category of election laws that might affect turnout, including both those that make voting harder and those that make it easier.

The authors give an example of a hypothetical law that imposes additional requirements for voting, targeting Democratic-leaning groups.

The requirements target 4% of the electorate, and cause a 3 percentage-point decline in turnout among this group — figures that the authors say are consistent with the effects of real laws. The result would be a 0.12 percentage point drop in overall turnout.

That’s already small, but because that group is likely to be around 60% Democratic, not 100%, the swing toward Republicans would be even smaller, just 0.011 percentage points. Only the very closest elections in history would be affected by a swing that tiny.

Even laws that contain several prongs that affect voting in different ways are still likely to affect results only in the very tightest elections, the authors write. North Carolina’s omnibus elections bill currently moving through the legislature there is an example, though it isn’t mentioned in the study.

In one section that may raise the hackles of voting-rights advocates, the authors note that there has been no significant turnout decline in the mostly Southern states that were affected by the Supreme Court’s 2013 ruling in Shelby County v. Holder, which removed the requirement for those states to have their election changes pre-approved by the federal government, to ensure they don’t hurt minority voting. In fact, they say, voting rates among non-whites have increased since the ruling.

Advocates and journalists — including this one! — have poured resources into documenting the slew of restrictive new rules, from voter ID laws to reductions in polling sites, that were imposed in the wake of Shelby, at times painting the onslaught as an urgent crisis of democracy.

Even far-reaching structural reforms that go beyond targeted measures like voter ID may not do much to affect election outcomes, the paper suggests.

Many predicted that the 1993 National Voter Registration Act, which has added millions to the rolls by requiring motor vehicles departments and other state agencies to offer registration, would help Democrats, the authors note. (“Who wins under this bill?” asked Rep. Spencer Bachus, an Alabama Republican, during the debate over the measure. The law, he answered, “will result in the registration of millions of welfare recipients, illegal aliens, and taxpayer-funded entitlement recipients. They’ll win.”)

In fact, the authors write, the law had essentially no partisan impact.

‘Beyond the voting wars’

Still, some critics note that many high-profile elections these days, including presidential elections, are decided by razor-thin margins. In both of the last two presidential elections, the winner won three pivotal states by 1.2 percentage points or less — in 2020, it was 0.7 percentage points or less. (And that’s leaving aside Florida 2000, a unicorn event that was so close that almost everything made a difference.)

“[U]sing the Hersh-Grimmer framework, some of these laws would’ve had a plausible chance of swinging the 2016 presidential election because the election was so close,” said Jacob Grumbach, a professor of political science at the University of Washington, via email. “ I could understand people might think that’s a big deal.”

Hersh acknowledged it’s possible that a multi-pronged law, or a set of laws acting together, could cause a swing that big. But he argued that because of the high level of uncertainty involved in the analysis, there’s no reliable way to predict what the partisan effects of a given law will be.

“Yeah, collectively these small policies could aggregate,” Hersh said. “But no one knows how they aggregate. Even the ones that liberals call suppression.”

“There’s no way lawmakers can sit around and be like, ‘OK, we’re going to do these six things and this is going to help Democrats or Republicans,’ and actually know what they’re talking about,” Hersh added.

The authors also acknowledge more than once that there are plenty of valid reasons to worry about election policies that have nothing to do with results —  “such as whether they make voting convenient, more secure, more cost effective, and whether they are motivated by discriminatory intent.” All of those effects would be important to pay attention to, even if they didn’t have a partisan impact.

In fact, both Grimmer and Hersh stressed in interviews that one goal of the paper was to encourage a focus on these other issues by making questions of partisanship recede.

​”States should be seeking out policies that they think will improve the functioning of elections,” said Grimmer. “And they can be comforted knowing that when they make those changes, it’s not going to end up with wild swings in partisan balance.”

Others see a different take-away from the paper.

David Nickerson, a political science professor at Temple University in Philadelphia, was involved in a project that looked at the impact of stadium voting, which took place in the 2020 election at over 48 professional sports stadiums as a convenience measure during the pandemic. Nickerson said project organizers hoped that when they showed stadium voting has no partisan effect, Republican officials would drop their opposition to it.

But that didn’t happen, Nickerson said. The experience suggests to him that differences over how elections should be run — and in particular over how easy or hard voting should be — are as much about ideological principle as they are about political advantage.

“You’ll even hear Republicans openly say they think voting should be harder, and you should only vote if you really want to and care and are committed — which I can’t imagine a Democratic official saying,” Nickerson said. “It’s a different worldview.”

Michael Morse, a professor at the University of Pennsylvania Carey Law School, said the study should lead advocates to focus relatively less on laws that affect the individual voting experience, like most voter ID laws, and more on structural issues, like getting more people on the voter rolls, or stopping gerrymandering.

“We have a limited amount of resources for reform,” Morse said. “The agenda for reform should be informed by this type of empirical political science.”

Indeed, the reality that election laws barely affect results is a good thing for building bipartisan coalitions for voting rights, Morse added.

“I would like the public discussion of these issues to be less partisan,” he said. “It’s the only way forward beyond the voting wars.”