“How Much of a Difference Did New Voting Restrictions Make in Yesterday’s Close Races?”

Wendy Weiser writes for the Brennan Center.  Once again, I’m skeptical.  Consider the analysis of North Carolina:

In the North Carolina Senate race, state house speaker Thom Tillis beat Senator Kay Hagen by a margin of 1.7 percent, or about 48,000 votes.

At the same time, North Carolina’s voters were, for the first time, voting under one of the harshest new election laws in the country — a law that Tillis helped to craft. Among other changes, the law slashed seven early voting days, eliminated same-day registration, and prohibited voting outside a voter’s home precinct — all forms of voting especially popular among African Americans. While it is too early to assess the impact of the law this year, theElection Protection hotline and other voter protection volunteers reported what appeared to be widespread problems both with voter registrations and with voters being told they were in the wrong precinct yesterday.

Some numbers from recent elections suggest that the magnitude of the problem may not be far from the margin of victory: In the last midterms in 2010, 200,000 voters cast ballots during the early voting days now cut, according to a recent court decision. In 2012, 700,000voted during those days, including more than a quarter of all African-Americans who voted that year. In 2012, 100,000 North Carolinians, almost a one-third of whom were African-American, voted using same-day registration, which was not available this year. And 7,500voters cast their ballots outside of their home precincts that year.

The relevant question is: how many people who WANTED to vote this year DID NOT DO SO (and reasonably could not have done so) BECAUSE of the changes in the voting rules?  In other words, to know whether these restrictions were outcome determinative we would have to control for:

1. a potential decline in turnout for reasons unrelated to these laws (e.g., less enthusiasm in a midterm election year without an African-American candidate on the ballot); and

2. the extent to which voters who voted early on days which were cut (including through same day voter registration) wanted to vote but reasonably could not have voted under the alternative early voting days, absentee balloting, or voting on election day. There is some evidence that early voting turnout may have increased in North Carolina despite the fewer number of days because the hours of voting were extended, making it easier for some working voters to vote before or after work.

I’m not saying Wendy’s conclusion is wrong—only that it is unproven and would take a much more nuanced analysis than this.


Comments are closed.