Raphael Warnock spoke of growing up along the cobblestone streets of Savannah. Of attending Morehouse College in Atlanta and preaching at Ebenezer Baptist Church in the same city.
Of living among the oak trees and Spanish moss.
“In a word,” he said, “I am Georgia.”
And the freshman senator, delivering his first speech in Congress last March, was disturbed by what he was seeing in his home state.
After Georgia voters chose Joe Biden for president, and then Warnock and fellow Democrat Jon Ossoff for the Senate in a runoff election two months later, Republicans decided not to change their message, he said, but to change the rules.
Advertisement
State lawmakers were moving to curb early voting, to make it harder to vote by absentee ballot, and even to forbid the delivery of food and water to people waiting in hours-long lines to cast ballots.

The pretext for these measures and similar legislation in other GOP-controlled states was “the big lie of voter fraud” in the 2020 election, he said. But the real motive, he contended, was suppressing the votes of people of color. “We are witnessing right now a massive and unabashed assault on voting rights unlike anything we have seen since the Jim Crow era,” he said. “This is Jim Crow in new clothes.”
Soon after Warnock’s speech, Georgia lawmakers would pass the measures he warned about — and another provision giving them greater control over local elections. By the end of the year, a total of 19 states would pass 34 laws restricting voting.
The senator’s antidote was two pieces of federal legislation. The Freedom to Vote Act would require states to provide at least two weeks of early voting and permit vote-by-mail for every voter, among other things. And the John R. Lewis Voting Rights Advancement Act would effectively restore parts of the Voting Rights Act struck down by the Supreme Court in 2013, requiring that states with histories of discrimination get approval from the federal government before enacting voting changes.
Advertisement
But now that two moderate Democratic senators have refused to set aside the filibuster rule and allow for a simple majority vote on these measures — even after President Biden urged them to do so — reality is setting in.
The GOP’s voter restrictions are here to stay. More could be coming. And Democrats are voicing concern not just about the fate of the right that undergirds all others but about their own electoral prospects.
“We feel if they can do these voting rights laws and other voting rights laws, we will never have a majority,” said Senate Democratic leader Chuck Schumer, referring to the GOP-approved state laws, in a postmortem interview with The Washington Post. “That’s the bottom line.”
Or is it?
With the new laws set to face their first big test in the midterm elections this fall, a close examination of the restrictions — and the research on voting behavior — suggests a more complicated picture.
A natural experiment in Texas
At the heart of the big lie was Donald Trump’s relentless campaign to sow doubts about voting by mail.
The pandemic-inspired expansion of absentee voting was a “big scam,” he said, with “Democrat governors” scheming to blanket “Democrat areas” with ballots in a bid to rig the election.
That was bunk, of course.
But you don’t have to be a conspiracist to wonder if widespread absentee voting — fairly administered — had an effect on the 2020 election. Turnout did reach record highs, after all. And voters delivered full control of the federal government to Democrats.
Advertisement
Did vote-by-mail play a role in either? Or both? No, says a recently published paper by a team of Stanford and UCLA political scientists.
The researchers tested the impact of mail-in voting in two ways.
First, they compiled turnout data for states such as Alabama, Pennsylvania, and South Carolina that moved to full no-excuse absentee voting in 2020 — no longer requiring an excuse like disability or election-day travel to fill out a ballot at home. And they compared it with data from states including Indiana, Louisiana, and Mississippi that didn’t make the switch.
If mail-in voting made a real difference, it would show up in this analysis; states that made it much easier to vote absentee should have seen significantly higher turnout.
But the data shows that turnout increased in both sets of states in roughly equal measure. It was up 5.6 percentage points in the no-excuse absentee ballot states, compared with the 2016 election, and 4.8 percentage points in the others.
And the modest 0.8 percentage-point gap separating the groups may have been little more than statistical noise. When the researchers compared turnout changes in the same two sets of states between the 2012 and 2016 elections — before widespread adoption of vote-by-mail — they found a gap of 1.7 percentage points. That’s an effect more than twice as large, and absentee voting had nothing to do with it.
Advertisement
The researchers’ second approach was even more precise.
Here they took advantage of a natural experiment in Texas and Indiana, where voters age 64 and younger need a reason to get an absentee ballot, while those 65 and older can apply for one without an excuse. If absentee voting makes a difference, 65-year-olds with much easier access to absentee ballots should turn out at higher rates than their 64-year-old peers.
But even in the heart of the pandemic, when people were especially likely to vote absentee — in 2020, nearly one in five of the votes cast by 65-year-old Texans was absentee — there was no discernible boost. Combining absentee and in-person voting to measure overall turnout, the researchers found virtually no difference in the voting rates of 64- and 65-year-olds.
And absentee voting had no discernible effect on the partisan composition of the electorate, either. Notwithstanding Trump’s loud warnings — and his opponents’ quiet hopes — mail-in voting provided Democrats with no advantage.
Daniel Thompson, a political science professor at UCLA and coauthor of the paper, says research on “convenience voting” reforms like no-excuse absentee voting and early voting is mixed. Some of it shows the reforms can actually depress turnout, perhaps by diminishing excitement around Election Day. Other research shows gains.
One of his own earlier papers on vote-by-mail found a modest 2 percentage point boost in turnout — though no partisan advantage for Democrats.
Advertisement
On the whole, Thompson says, the research suggests the effects are quite small: “These sorts of reforms that just allow people more options for when and how to vote don’t seem to change participation all that much.”

Same-day registration — which allows people to register on the day they vote — may do a bit more to increase turnout. But it hasn’t been at the center of the voting wars.
Montana’s legislature voted to repeal same-day registration last year. But Republicans in other parts of the country left the practice in place — or didn’t have a same-day registration law on the books to revoke.
Republicans have moved, in larger numbers, to tighten one of the most contentious restrictions — laws requiring voters to show an ID before casting a ballot.
Liberals have long argued that the laws aren’t designed to stamp out fraud, as the GOP claims, but to disenfranchise low-income people, people of color, and members of other traditionally Democratic constituencies who are less likely to have an ID.
And several Republicans have acknowledged that voter suppression is indeed what they’re after.
In 2012, as former Massachusetts governor Mitt Romney was campaigning for president, Mike Turzai, the majority leader in the Pennsylvania House of Representatives, proudly declared “voter ID, which is going to allow Governor Romney to win the state of Pennsylvania, done.”
Four years later, Glenn Grothman, a Republican congressman from Wisconsin, predicted in a TV interview that the state’s voter ID law would make it more difficult for Democrat Hillary Clinton to win the presidency.
And that same year, a three-judge federal appeals panel threw out a North Carolina law that cut back on early voting and imposed strict voter ID requirements, finding that Republican lawmakers had targeted “African Americans with almost surgical precision.”
Thousands of pages of emails and other documents produced under court order showed that GOP officials were, in fact, engaged in a systematic effort to disenfranchise voters of color.
“Of course it’s political,” North Carolina Republican consultant Carter Wrenn said in an interview with The Washington Post. “Why else would you do it?”
However, one of the largest and highest-quality studies of voter ID to date — conducted by Vincent Pons of Harvard Business School and Enrico Cantoni of the University of Bologna, in Italy — suggests the GOP efforts have been for naught.
Poring over voter data spanning a decade, the economists found that ID laws “have no negative effect on registration or turnout, overall or for any group defined by race, gender, age, or party affiliation.”
One possible explanation for this sort of finding is that voter ID laws do, in fact, have a suppressive effect but that countermobilization nullifies it. Targeted voters might hear about the laws, get angry about attempted suppression, and make an extra effort to cast a ballot — or the Democratic Party and its liberal allies could dedicate resources to reaching out to affected voters and making sure they have the ID they need to vote.
The concern here is that voter anger could eventually subside or Democratic organizing could tail off, allowing the fundamentally suppressive nature of ID laws to prevail.
But Pons and Cantoni, while acknowledging that countermobilization is difficult to measure, couldn’t find much evidence that it’s happening. If strict voter ID laws were generating a backlash on the ground, that might show up in increased campaign contributions, greater participation in political meetings, or more people putting up campaign signs. But the researchers found no significant uptick.
And while voters of color do report greater contact from political campaigns in voter ID states — presumably from Democrats looking to offset the effects of the laws — the increase isn’t large enough to have a significant effect on turnout, Pons and Cantoni found. If Democrats are offsetting voter suppression effects, then they are very minor effects in the first place.
This doesn’t mean voter restrictions are unimportant.
Political scientists say a restriction that makes a small difference in turnout could help tilt a very close election.
And they add that there is something novel about putting several limitations on voting in place at the same time, as Georgia and other states have done. It’s possible that voters who see early voting curtailed, face a shorter window for applying for an absentee ballot, and lose a drop box in their neighborhood — all at once — will be less likely to vote.
That’s what activists are worried about.
“It’s like death by a thousand cuts,” says LaTosha Brown, cofounder of Black Voters Matter, which is based in Atlanta. “One more cut might not hurt you, but if you’ve got enough cuts, you can bleed to death.”
And there is some troublesome early evidence of the new laws’ effects.
In Texas, elections officials are rejecting hundreds of absentee ballot applications either because they don’t comply with a new requirement that applicants list a driver’s license number, state ID number, or the last four digits of their Social Security number on the paperwork — or because the county doesn’t have the information in its voter file to match with the numbers the applicants provide.
There is still time to correct any errors. And even if some applicants never get an absentee ballot, they could very well find other ways to vote.
But as the midterm elections approach, activists and Democratic Party officials aren’t taking any chances. They’re planning to spend tens or even hundreds of millions of dollars to register voters and get them to the polls.
That means the GOP-approved restrictions, whether they suppress votes or not, are exacting a real cost — diverting cash that could be spent promoting individual Democratic candidates.
And that’s not the only concern. As the turnout battle gears up, Republicans are mounting a separate and potentially larger threat.
They’re putting machinery in place that could allow them to meddle with the count after votes are cast.

The subversion problem
In the wake of the 2020 election, America had an unnerving encounter with subversion.
President Trump weighed the seizure of voting machines. Hundreds of his supporters stormed the Capitol, brutalizing police officers and waving the treasonous flag of the Confederacy in a citadel of American democracy. And at least five states — Arizona, Michigan, Pennsylvania, Texas, and Wisconsin — launched partisan reviews of the election.
The revolt wasn’t enough to prevent the transition of power.
But it left activists and observers worried about future acts of sabotage — and mulling how to prevent them.
At the federal level, attention has focused on the Electoral Count Act, a 19th-century law that governs how Congress counts the electoral votes of each state in the presidential election.
Trump and his supporters exploited ambiguities in the statute to claim that Vice President Mike Pence, who oversaw the count, had the power to toss out states’ votes — a claim the former president amplified just last weekend when he said that Pence “could have overturned the election.”
A bipartisan group of senators has been working on legislation that would clarify that the vice president plays only a ceremonial role in the count and raise the threshold for challenging a state’s results in Congress; under current law, just one representative and one senator can register an objection and force a vote.
Rick Pildes, a professor of constitutional law at New York University, says it’s critical that “there be buy-in from both parties” for reform of the Electoral Count Act — not only because Republican votes will be required to overcome any filibuster but also because GOP support makes it more likely that a narrowly divided Congress will handle a contested election properly next time.
Even if Congress is able to muster a rare bit of bipartisanship, though, it’s not clear that the legislation will rein in election meddling on the state level.
And the threat of that sort of meddling is rising. Nearly two dozen Republicans who have disputed or raised doubts about the 2020 election results are running for secretary of state — hoping to oversee elections not just in GOP strongholds but in swing states like Michigan, Nevada, and Arizona where they could do real damage.
Mark Finchem, a Trump-endorsed Arizona candidate, attended the rally preceding the storming of the Capitol on Jan. 6. And last month, he joined the former president at an event outside Phoenix and called on the state legislature to decertify the results of Arizona’s 2020 presidential election, which Biden won by almost 11,000 votes.
“Ladies and gentlemen, we know it and they know it — Donald Trump won,” Finchem said.
The worry is that if Trump allies become secretaries of state, they could cast doubt on election results or refuse to certify them — clearing the way, in some states, for governors and legislatures to substitute their own slate of electors for president.
Even if they don’t win these races, there is still room for GOP manipulation of election results.
In Georgia, where Republican secretary of state Brad Raffensperger rebuffed Trump’s false claims about voter fraud, lawmakers have taken the secretary of state off the state election board, asserted control over the panel, and empowered it to remove county elections officials.
And in Arkansas, a new law allows a special legislative committee to investigate county elections officials and make recommendations to a state board, which can then remove those officials and “take over and conduct” local elections.
But dark as some of the scenarios are, there is reason to believe the system will hold.
As the Wall Street Journal editorial board has pointed out, even the widely panned Georgia law includes safeguards.
In order to suspend and eventually fire a county elections official, the state election board has to either find that the official has violated three elections rules over the course of two elections and failed to remedy them or find “by clear and convincing evidence” that the official has demonstrated “nonfeasance, malfeasance, or gross negligence” in the administration of two elections over the course of two years.
Moreover, some of the most brazen legislative proposals at the state level — including measures that would have allowed state legislatures to directly overturn election results — have gone nowhere, suggesting there are limits to what the GOP will do, even in thrall to Trump.
And if politicians do go too far, the courts should be a bulwark against election chicanery just as they were in the aftermath of the 2020 election.
Still, if the immediate threat isn’t as dire as Democrats like Warnock, Schumer, and Biden have suggested, it would be a mistake to dismiss the new election laws as harmless.
It’s not just that the voting restrictions could tip some close elections or that meddling with results is a possibility now. There is something broader.
Republican lawmakers say they’ve passed these laws to restore confidence in American elections. But research on the long-standing voter ID laws shows they have had no such effect, raising doubts about whether the new wave of “anti-fraud” measures will do any better on this score.
Indeed, it seems entirely possible that they will only deepen unfounded suspicions about the integrity of American elections.
“Passing restrictions — it doesn’t make people feel better about our elections,” says Sean Morales-Doyle, acting director of the voting rights and elections program for the Brennan Center for Justice. “It reaffirms that there’s something to fix.”
At stake, here, is faith in the American project. And if we’ve learned anything over the past several years, it’s that our faith is fragile.
David Scharfenberg can be reached at david.scharfenberg@globe.com. Follow him on Twitter @dscharfGlobe.