But so far, they haven’t really worked as intended. A 2016 study in Texas, home to some of the most restrictive voting laws in the country, surveyed non-voters from that year’s election in the Democratic stronghold of Harris County. Just 1.5 percent of them listed not having identification as the main reason they didn’t vote. A 2014 Virginia study found that of the 2.19 million eligible voters who tried to cast ballots in that year’s election, just 474, or approximately one in 4,600, were stymied by lack of ID. These are not outliers. The UC Davis political scientist Benjamin Highton, after surveying ID laws nationwide, concluded that they had “modest, if any, turnout effects.” Why haven’t voter-ID laws suppressed very many votes? No one really knows. It’s possible that these blatant attempts at voter suppression have created an equal and opposite backlash, driving turnout among the groups intended to be suppressed. It’s also possible that future, more carefully targeted voter-ID laws will be far more effective at manipulating elections. But if this ends up being the case, it will only prove the broader point: Although requiring voter ID is unnecessary, and the politicians who support it tend to do so for deplorable reasons, the act of requiring identification in order to vote is not in and of itself voter suppression... That’s exactly why Manchin’s proposal is, as a matter of both policy and politics, quite elegant. The currently available evidence suggests that, as improbable as it may seem, expanding voter-ID requirements to all 50 states would do little to reduce turnout. But Manchin’s not just trying to require identification nationwide. He would also expand the types of government ID that can be used in elections, for example by allowing voters to cast ballots if they display a utility bill. This version of a voter-ID law would require voters to prove their identity without disenfranchising a large number of voters—in other words, it’s the exact type of voter-ID law most Americans already support.