There's an election coming up, and Greg Walsh wants to be sure he's ready. So he shoots a quick e-mail to his county election commission to clear up what documentation he needs to vote. The note goes like this:

I've been hearing a lot about voter ID laws on the news. What do I need to do to vote?
Thank you,
Greg Walsh

An official at the commission shot him a link to his state's voter ID regulations, which clearly state if he needs an ID, under what circumstances an ID is needed, whether it can be expired, whether it needs to be a photo ID, etc.

Luis Rodriguez has similar questions. He sends an e-mail to his local election official. In fact, he sends the exact same e-mail Greg sent:

I've been hearing a lot about voter ID laws on the news. What do I need to do to vote?
Thank you,
Luis Rodriguez

He didn't hear back.

Greg and Luis aren't real people. They're aliases used in an experiment by Harvard political science grad students Julie Faller, Noah Nathan, and Ariel White, which finds that local election officials are less responsive to requests for basic information about voting when they come from people with Latino-sounding names.

Faller, Nathan, and White "contacted every local official or election commission responsible for overseeing elections for each county or municipality at which elections are administered in 48 states."* One quarter of the e-mails used a Latino-sounding name, like "Luis Rodriguez," and asked the voter ID question you see above. Another quarter used a non-Latino-sounding name, like "Greg Walsh". The other half asked a control question ("Do you have to vote in the primary election to be allowed to vote in the general elections?") using both sets of names to see if asking about voter ID in particular had any effects.

After all the responses were back, they had a sample including 6,825 sent e-mails to officials in 46 states.** At least 4,557 officials replied. But the interesting story is in who they did and didn't reply to. "Responses to Latino names," the researchers write, "are three-and-a-half to four percentage points less likely than to non-Latino white names." The bias against Latino e-mailers was about three points greater in voter ID questions, though the difference between that and the primary election question wasn't statistically significant.

The finding holds up when you drop certain regions, when you drop small towns, and when you control for whether officials are elected or appointed. What's more, they find that there are actually statistically significant differences in the quality of response from officials, depending on what kind of name is used. Responses to Latino voters were likelier to be non-informative, less likely to be "absolutely accurate" (that is, giving complete and accurate information about the relevant topic), and even less likely to take a friendly tone:

It's worth noting that an "absolutely accurate" response could be something as simple as e-mailing a link to a state's voting regulations. "We ended up going with what we thought was the most conservative approach, which was giving them credit when they sent you to an institution that should be giving out the data," White tells me. But many didn't do even that.

Interesting, the researchers found no evidence that states with stricter voter ID laws are likelier to discriminate against Latino e-mailers. But they note that the results still bear on that debate. They write, "Our results indicate that changes to existing voting regulations are likely to differentially increase information costs for Latino voters because public officials are less responsive to their inquiries than to non-Latinos."

Similarly designed studies have also found evidence of bias on the part of local political officials. Yale's Daniel Butler and David Broockman (now at Berkeley) found that state legislators are less likely to respond to requests for information from people with black-sounding names. The researchers in this study said they focused on Latinos in part due to public debate over whether or not lax voter ID laws were letting undocumented Latino immigrants vote. "When we started thinking about voter ID, we thought this might be the most relevant reference group," White says.

For what it's worth, voter ID advocates dismiss the findings. Catherine Engelbrecht, president of True the Vote, calls the paper "a conclusion in desperate search of a viable methodology." "In reality, the Pew Hispanic Center found that 71 percent of Latinos support voter ID, which is consistent with national polling just short of 80 percent," she said. "Pew also found that 97 percent of Latinos surveyed were confident they had the identification required to meet local standards. This safeguard has long been a part of the Latino electoral culture, because every single Latin American country uses photo voter identification."

Update: Keesha Gaskins, senior counsel at the Brennan Center for Justice and a critic of voter ID laws, comments: "Based on the results of this study, we see another example of institutional barriers, beyond the control of individual eligible voters, that may impact their ability to vote. Understanding that voter ID laws affect the ability for voters to obtain a ballot, where there are institutional barriers that are exacerbated by race, it provides additional reasons why these laws are bad policy."

* Maine didn't list e-mail addresses for officials, and in Alaska the officials' boundaries didn't correspond to Census categories that the researchers used to see districts' demographics, such as racial makeup, average income, etc. So neither state's officials were e-mailed.

** Minnesota officials conversed with each other about the e-mails they were getting, and Virginia officials forwarded e-mails to each other, resulting in some officials getting multiple categories of an e-mail. The researchers dropped the states from the sample because of these issues.