To what extent do voting restrictions affect elections?

Welcome to FiveThirtyEight's weekly political chat. The transcript below has been slightly modified.

sarahf (Sarah Frostenson, Political Writer): From the purge of voter lists in Georgia, which heavily affects black voters, to a North Dakota law that could make voting more difficult for Native Americans, voter identity laws have attracted a lot of media attention until mid-term. . A Pew Research Center study released this week, however, revealed that most Americans (76%) were still supportive of the requirement to present an identity photo issued by the government in order to vote, despite the evidence more and more obvious of a partisan fracture.

The question I am asking you all today is threefold:

1. What do we know about voter identification laws and how they shape elections?
2. What evidence can we indicate to show their effect on the electorate?
3. How to reconcile public opinion with the laws on the identity of voters?

danjhopkins (Dan Hopkins, professor of political science at the University of Pennsylvania and contributor to FiveThirtyEight): I would say that the voter identification laws are of different shapes and sizes. So we often need to break them down by asking: What type of identification is needed, for whom and what happens when voters do not show any piece of identity? I think these distinctions are often lost in debates around these laws – but they can be decisive for their impacts.

julia_azari (Julia Azari, Professor of Political Science at Marquette University and contributor to FiveThirtyEight): I would like to add something and say that some of these laws are identity laws, but some are registration laws. What we have seen happen in Georgia with mail ballots thrown away because their signature does not exactly match their registration dossier is qualitatively different even from a strict law on photo-photo.

danjhopkinsYes, Julia, certainly a key distinction. Part of what has been talked about in Georgia is voter registration, and the decision of the Supreme Court in Ohio earlier this year was also about voter registration, not about voter registration. identity of voters.

For example, if Ohio cancels an elector's registration because an individual did not vote in a few federal elections and does not return a postcard, it is a registration law voters. But if Virginia asks a person who shows up to vote without a proper ID card to vote, it is an elector's law.

julia_azari: So I suppose I would categorize North Dakota and Georgia as examples of voter registration law rather than voter identity law.

danjhopkins: (Except that North Dakota is complicated because it is not technically registered on the electoral lists …)

julia_azari: Yes, so maybe we're talking about voter eligibility.

In Wisconsin, we register the same day, but the voter identification laws are complicated.

In other words, localized elections are therefore a mess.

danjhopkins: But on voter identity laws, several recent studies of different laws seem to converge on a very rough consensus: these laws disproportionately affect voters of color and older voters, but they tend not to have effects of sufficient importance to negatively affect election results. everything but the very the closest groceries. Of course, if someone is denied the right to vote, it is important in and of itself, whether or not a different candidate has been elected.

Recent research in political science has also highlighted the very real possibility that these laws produce short-term reaction effects – in other words, they mobilize some of the same components that are disproportionately affected.

nrakich (Nathaniel Rakich, Election Analyst): Yes, Dan, I find the idea of ​​reaction quite interesting. For example, a group called North Dakota Native Vote and Daily Kos came together to raise half a million dollars as a result of the Supreme Court's decision to leave the North Dakota Identity Act. This money is now used to help Native Americans obtain the necessary documents to vote and to carry them to the polls.

sarahf: I guess I have a hard time believing it. Are voter identification laws that make it more difficult for some people to vote … actually encourage them to vote in greater numbers? Or at least that's the case in the beginning?

danjhopkins: Well, Sarah, I think that's partly because these laws tend to apply to relatively small fractions of the electoral population. When several co-sponsors and I spotted the provisional ballots cast in Virginia in 2014 due to a lack of identity papers, the number was only about 500 in all of them. State. Now granted, tens of thousands of people have traveled to Texas and Michigan without voting in 2016, without the necessary voter identification, which brings me back to my previous point about how the laws in voter identification can vary a lot. But again, the number of people affected rarely adds up to an order of magnitude enough to allow an election to tip.

julia_azari: The fact that these issues have become partisan is terrible for long-term democracy, but perhaps good for short-term mobilization.

It is not surprising that people mobilize when they feel specifically targeted or threatened.

People may be demoralized or immobilized for a variety of reasons without voting, but there is all sorts of evidence that specific threats can actually serve as a catalyst for effective engagement.

danjhopkinsYes, Julia, definitely. The backlash effects of unpopular laws are likely to fade, but if the laws remain in full force, the deterrent effects will probably not be mitigated.

And if we think about it strictly from a partisan perspective, the impacts of the laws can also change over time. As voters age, they are less likely to have their current driver's license. We can see that the partisan effects of these laws change as the electorate ages.

nrakich: And the voting coalitions can change. Hispanic voters can become more Republican, for example. So, voter identity laws could start hurting the GOP.

julia_azari: An issue that I did not see in the literature – and certainly the document has not been published for so long – is whether the mobilization that occurs around the right to vote is a mobilization that otherwise, would not have occurred. Or maybe a mobilization that might otherwise have happened, but around a different problem.

nrakich: Indeed, in addition to motivating them with anger to be more likely to come forward, coverage of the news by a court decision or a new electoral identity law can also help educate people about what they are saying. must bring to the polls. It is not difficult to imagine such a small but significant number (I think of four figures) of people voting for this reason, which counteracts the depressing aspect of the law.

danjhopkins: Not only that, but also the national and local authorities can actively try to educate the public, as did the Virginia officials in 2014. They sent a mail to thousands of voters not having a driver's license, which may partly explain the mute impact of this strict voter identification policy.

nrakichIt is likely, however, that if a republican government actually enacted an electoral identity law in order to reduce turnout (especially among more democratic groups), it would take no action to actively educate voters.

Republicans have sometimes admitted that this was their intention. A Republican from Pennsylvania, for example, argued for voter identity laws in 2012 saying it would "allow the governor [Mitt] Romney to win the state of Pennsylvania. "

danjhopkins: But the Pennsylvania law has been rejected by the courts.

I know something about voting in Pennsylvania as a voter here, and you just have to show a piece of identity the first time you vote in person in an enclosure.

sarahf: I think I heard on the podcast that you were registered to vote in several states, Dan. ?

danjhopkins: Not many, but although most states have been quick to remove me from the list, one has not been decided.

julia_azari: Election expert Rick Hasen wrote an article pointing out that the country is moving in different directions. Some states facilitate voting.

Others do not … do that. And there is a fairly obvious partisan scheme.

danjhopkinsIt should be noted that there is strong research in political science on which legislators and states tend to adopt restrictive laws. It's typically Republicans, and it's typically in competitive states with a large black population.

nrakich: But these laws were quite rare and undisputed until recently. They really started to take off in 2011 … after the Republicans took control of a group of state governments.

danjhopkins: It is also important to note that the Supreme Court decision in the Shelby County case v. Holder in 2013 allowed a number of states and localities to amend their election laws and administration without obtaining approval from the Department of Justice.

Indiana adopted the first of the recent wave of voter identification laws in 2005, which was upheld by the Supreme Court in 2008. But after 2010, the GOP took control of various state hotels, so that he had majorities to promulgate in Wisconsin. new policies.

julia_azari: I think the decision of the Supreme Court is really what makes the difference here.

The court found it constitutional that the Ministry of Justice oversee the voting laws, but that the coverage formula stipulated in the Voting Rights Act was outdated and had to be updated by Congress. But that did not really happen.

I also have a hypothesis: public opinion on voter identification laws will evolve as more and more people are affected by these issues.

nrakich: It's already changing. As I mentioned earlier, voter identity laws were bipartisan and uncontroversial. But now that they have mostly identified with Republicans, there is a partisan split in public opinion (though not as important as issues like abortion).

Political scientists, there is a name for that, right? When do people follow elites for their clues?

danjhopkins: Nathaniel, it's a name associated with that – it's John Zaller. His 1992 book, "The Nature and Origins of Mass Opinion," talks about this.

nrakich: It seems like a good light reading. I have a rather cold week ahead.

danjhopkins: Yes, everyone digs chapter 4, "Compose with instability of answers".

julia_azari: I have considered tattooing an instability of response but I have not done so yet.

danjhopkins: Fortunately it's a conversation or I'll show mine …

sarahf: Hahah, but that's what I did not really understand in all this: the factor of public opinion. There is a fairly broad consensus among Americans that one has to show the identity of an elector, but there is also evidence that Republicans and Democrats disagree about the nature of the problem (or solution) under -jacent.

nrakichTo recount Sarah's chart, it's quite striking (at least for me) that most of the fight against the vote-to-vote policy is about politicians. Voters generally agree: a large majority want electoral identity, but also other policies facilitating the vote, such as registration the same day.

Ninety-one percent of Republicans support voter identification laws, but 63 percent of Democrats do so. Eighty-two percent of Democrats support the re-franking of crimes, as do 55 percent of Republicans.

danjhopkinsYes, this is a case where opinion polls focus on broader issues while political debates focus on very narrow issues. Like, what counts as an identity? Can I vote with a student card from a public university? What about a handgun license?

julia_azariIt's quite sober to realize that even our forms of identity are partisan.

sarahf: Are there any specific initiatives on the November ballot regarding voter identity laws in some states? What are they asking for there?

nrakichYeah. Arkansas and North Carolina will both vote on the inclusion of voter identification in their state's constitution.

On the Liberal side, Maryland will vote for same-day registration and Michigan will vote for a package of changes, including automatic voter registration and absentee voting for absentees.

Oh, and Nevada also votes on automatic voter registration.

danjhopkins: Has anyone seen a vote on these measures? Given that the public is generally supportive of voter identity, I would imagine that it would get substantial support from the public, although perhaps (as Zaller has mentioned earlier) many Democrats will vote against.

nrakichA survey conducted in September in Arkansas revealed that its voter code voting initiative increased from 71% to 21%.

In North Carolina, the voter identity amendment gives an advance of 64% to 27% in early October.

julia_azari: But we need to think about how public opinion deals with paranoia about the vote of non-citizens – and, I argue, is linked to a long history of attacks on the right to vote of blacks US.

danjhopkins: Absolutely. We can not hold this conversation without acknowledging that it is deeply linked to a long history of systematic efforts to deprive black voters and other black people of their rights.

Proponents of these laws often say they are designed to prevent voter fraud in person, but it's great, super rare. One study found that voter identity theft is reported at about the same frequency as extraterrestrial removal.

sarahf: Electoral fraud is more and more often equated with electoral security. Right?

julia_azari: Exactly, which is unfortunate, because the security of the elections poses quite significant problems.

nrakichIronically, one of the most plausible ways to compromise electoral security is for hackers to access voters' files and remove them from voters.

julia_azari: Right. Which is also a big problem with American democracy, even without Russian hackers or whoever it is. Rather than having a problem where too many people are anxious to vote, we have the opposite problem. Millions have the right to vote but do not.

danjhopkins: Earlier, Julia pointed out how decentralized voting is in this country and she is absolutely right: the vote is administered by our neighbors ready to serve as tellers, in very different ways in different places. And that goes both ways. On the one hand, it means that our data is decentralized and the risk that things go wrong in an enclosure is high. But it also means that the level of coordination needed to hack our elections without our knowledge is huge, perhaps unimaginable.

julia_azari: It also means that the implementation of laws depends on the fact that election officials know the laws and are reasonable. Do not denigrate the people who work at the polls – it's a real civic engagement. But it can go wrong.

sarahf: OK, it is unlikely that voter identity laws could change the outcome of an election, but it is clear that these laws affect the fact that people vote both short and long. long term. What should we look for on election night and beyond?

danjhopkins: The most credible studies on the effects of voter ID on voter turnout consist of examining individual data, often merged between different sources. So I do not think we can see anything on election night. But in the following weeks and months, once we get the data from the voter file, we may be in a better position to highlight some of the impacts at the individual level, namely groups that appear to have been dissuaded from voting or. In some cases, we will also be able to examine document traces. And I repeat, the short-term impact of the promulgated laws tends to be felt on a scale that would only reverse incredibly tight elections (think of the Florida presidential election in 2000). But it is extremely problematic that specific people still can not vote accordingly – and that these people are disproportionately people of color and older voters.

julia_azari: At the risk of making my political science card, I will look for stories suggesting that something more systematic could happen during elections: laws that are poorly enforced or used to intimidate voters.