Finding information about voter ID laws and their effects was not difficult. There have been various attempts to study these laws and their effects, showing that they depress voting rates somewhat, especially among minority voters who tend to vote for Democrats; yet, the problem has only gotten worse in recent years. The same Supreme Court ruling that I cited in my previous post, that struck down part of the Voting Rights Act in 2013, has allowed states with a history of voting discrimination against minorities, to legislate voter ID laws with impunity, so they have increased in number over the last 4 years. However, these kinds of laws have been present in some states for decades. Wikipedia’s entry on voter ID laws, in addition to its usual comprehensive coverage of a topic, lists the type of voter ID law present in each state as well as showing them in a map (https://en.wikipedia.org/wiki/Voter_ID_laws_in_the_United_States). Overall, 32 of the 50 states have some kind of voter ID law. (However, I happen to live in one of the states with no voter ID law, California.) “Red states” tend to be more likely to have such laws, especially in their stricter forms, as these laws have typically if not always been passed by Republican majority state legislatures. The 18 states with no voter ID law are mostly, but not all, “blue states.” States with strict voter ID laws, rigidly require an ID; this includes 9 states (Georgia, Indiana, Kansas, Mississippi, Tennessee, Virginia, Wisconsin, Arizona and Ohio). States with non strict voter ID requirements, which includes 23 states, accept such techniques as affidavits or verifying a voter’s identity after the vote is cast.

The crucial question, however, is how these laws effect the actual vote. I found the Wikipedia article lacking in this regard, as the anonymous author claimed that such laws appear to have a negligible impact on voting turnout, then proceded to cite numerous studies, the large majority of which do in fact show reduced turnouts usually ranging from a 1% to 4% reduction in the number of votes cast. (Wikipedia is thought by many to be be a tool used by anonymous conservatives, probably paid by think tanks, to promote their point of view.) The best article I found on the topic was one that actually estimates their effect on numbers of votes for Republicans versus Democrats, by none other than Nate Silver of the website FiveThirtyEight (https://fivethirtyeight.blogs.nytimes.com/2012/07/15/measuring-the-effects-of-voter-identification-laws/). However, keep in mind that his analysis was done in 2012, before part of the Voting Rights Act was struck down by the conservative majority on the Supreme Court. Thus, the effects of these laws have if anything, only gotten worse since then.

According to Silver, aggregating the research on voter ID laws, shows that they reduce turnout by about 2% of the registered voter population, on the average. Silver also makes a cogent point that it is rather absurd to have a null hypothesis which presumes no effect of the laws on voter turnout, when these laws can basically only decrease turnout, not increase it. Thus, studies which find “no significant effect” of voter ID laws on voter turnout, have in all likelihood, simply lacked the statistical power to determine the amount of impact that such laws have. Silver goes on to show that studies of voter turnout across states showed that for every 1% increase in the turnout, Obama did about 0.6% better in 2008. Similarly, a comparison of voter turnout in the same state between 2004 and 2008, showed that for every 1% increase in the turnout, Obama did about 0.4% better in the voting results. Overall, this means that for every 1% decrease in voter turnout caused by voter ID laws, the Republican should end up with about 0.5% more of the vote total than would otherwise have happened, according to these estimates. With an average decrease in voter turnout of 2% as a result of these laws, that translates into an estimated 1% increase for the Republican versus the Democrat. Of course, that number will vary according to the actual amount that voter turnout is depressed by voter ID laws in a particular state, but it is enough to give Republicans a very significant edge in close elections.

How many elections (on all levels) are decided by one point or less? Clearly, it is quite a lot of them. As it turns out, Trump won Pennsylvania, Wisconsin and Michigan by much less than 1 percentage point. Pennsylvania has no voter ID law, but Wisconsin has a strict photo ID requirement, and Michigan has a non strict photo ID requirement. Wisconsin is “worth” 10 Electoral College votes, and Michigan, 16, for a total of 26 Electoral College votes between these 2 states. Thus, Nate Silver’s analysis suggests that voter ID laws were likely responsible for Trump winning Wisconsin and Michigan. If so, Trump still would have won the Electoral College, but by only 280 to 258 instead of 306 to 232. However, remember that in addition to voter ID laws, the Democrat has an uphill battle against an inadequate number of polling places in many areas, and efforts to remove voters from registered lists, as well as psychological (eg. anti-voting memes) and demographic (eg. youth, lack of voting experience) factors suppressing the liberal vote. These factors could have been enough to flip Pennsylvania and perhaps several other states to Trump in the 2016 election. The overall effect of voter ID laws, of course, goes far beyond the presidential election. It also effects congressional, state and local elections. I was unable to find information regarding what percentage of elections are decided by one percentage point or less, but there are many examples of very close elections; some have even been decided by a single vote. There are a great many elections in the states which have voter ID laws in even numbered years; some of them are sure to be decided by such small margins, all of which would be skewed toward favoring the Republican candidate, based on the evidence regarding voter ID laws.

A more recent study on the effects of voter ID laws, completed in 2016 before the general election, specifically shows a depressed turnout among minority voters, especially in states with strict photo ID laws (https://www.washingtonpost.com/news/wonk/wp/2016/02/04/new-evidence-that-voter-id-laws-skew-democracy-in-favor-of-white-republicans/?utm_term=.14178785f161). Blacks, Hispanics and other minority voters showed reduced voting turnout rates compared to whites in these states. Overall, it was estimated that the number of votes for the Democrat was decreased by 7.7 percentage points, while that for the Republican was also decreased, but by only 4.6 percentage points, giving the Republican a 3.1% boost. That is clearly a lot more than the 1% that Nate Silver estimated, but this number specifically applies to states with strict photo ID laws, which are the strictest in existence. Smaller boosts for Republican candidates can be expected in states with other kinds of voter ID laws. This study has been critiqued, as some calculation and methodological errors were discovered in it, but even critics admit that its findings are largely valid.

Moving forward, we need to attack these voter ID laws, both on legal grounds, and by having Democrats or any non-Republicans, take back state legislatures and cancel such laws. Legal challenges to voter ID laws are being made in several states, which Wikipedia reports are likely to end up in the Supreme Court. I wish we had a Supreme Court that was more up to the task of understanding the unfairness and disingenuousness of these laws, but even with the current Supreme Court, which is at this time deadlocked on the gerrymandering issue, there is a good chance of them ruling in favor of abolishing voter ID laws. Importantly also, the public needs to be educated about the real purpose and effects of these laws, which will give the voters yet another reason to vote Republicans out of office and pressure legislatures to abolish counterproductive voter ID requirements.