Wednesday, November 16, 2016

Wisconsin and many reasons

In social science we recognize that there is more than one reason that explains social phenomena. It is the one good impulse behind the complicated models we use in academic writing to try to prove a particular hypothesis, but leaves most people out of the conversation. This is a frequent complaint we hear that our journals are too mathematically oriented and don't address real world concerns. But like with many bad ideas, there is more than a grain of truth in the premise. There is no direct cause and effect that makes one thing the true answer. It is many things.

So why did Hillary Clinton lose Wisconsin to Donald Trump (by a skinny 27,000 votes)? There are most likely many reasons. First and foremost, the decline in voter turnout. As of today the total vote cast in Wisconsin is down 120,000 votes from 2012. Based on Dave Wasserman's vote count total (which seems very comprehensive) and state archives, that decline is almost all explained by the reduction in the Clinton vote from where Barack Obama stood in 2012. Obama 1.62 Million in 2012 and Clinton 1.382 Million while the Romney to Trump vote was down only 5000.

In Milwaukee county alone the turnout was down 52,000 votes from 2012 and there is a great explanation of most all of Clinton's loss of 27,000 statewide. However, in Dane County (home to Madison) the vote total was about the same as in 2012 and Clinton actually performed a small amount better than Obama (2000 votes) and overall turnout was up 6000 plus. Clinton also did a small amount better than Obama in suburban Waukesha County (where she and other democrats do poorly but still) while Trump was down 20,000 votes from Romney's total four years earlier. In fact, across the three counties Clinton's margin increased over Obama's by 20,000 (245,811 from 225,425).

So the rest of the state must have been a disaster. And it was. Compare the 2012 map with the 2016 map and you can easily see the wave of red that took over other counties in the state. So the question is still why? The mix of answers that we have seen across the flipped states probably apply ... voter suppression (a form of racism), the impulse for change, poor campaign messaging and tactics (for the region), the email nonsense, sexism and general lack of enthusiasm for the candidates. In order to address these issues the Democratic Party should take the time in 2017 to go talk to voters in the flipped counties and get a sense for what is on their minds. This is critical to mount a comeback at the state and local level in 2018.

Saturday, November 12, 2016

What happened in Michigan?

Michigan, like some hanging chad from an earlier election, has yet to be called by the AP and other news organizations. In the midst of the Electoral College upset on Tuesday, I went looking deeper into the numbers there to see if there were simple answers as to why Trump carried Pennsylvania, Wisconsin and perhaps Michigan (given that the re-canvas may change the results but it won't change the overall outcome of the election).

Critical to Democratic chances in Michigan is a strong showing in Wayne County that includes Detroit. There the obvious problem was a decline in voter turnout since 2012 where total votes cast went from 820,654 down to 779,971. Moreover, the Obama total of 595,846 slid to 517,447. Some have focused on the growth in Green Party vote in the state as a problem but I find that to be a thoroughly weak minded argument. The fact is that in Wayne County the Clinton vote was down 78,000 votes from the Obama total and Stein only received 5,000 more votes there. Moreover, the Libertarian Party (who wasn't on the ballot in 2012) received 18,753 votes way up from the scattering of write-ins they received in 2012. But the total write-ins also went up this year to 5.545. As did the blanks (meaning ballots that had no preference) which increase three-fold to 5,687. Finally the Trump vote increased over Romney by 15,000. All told the Clinton vote in Wayne was 66% of the true total (include blanks) down from 72 % in 2012 of the true total.

So most of Clinton's loss was a decline in turnout, increase in Trump support over Romney and a softness in the Clinton vote in all possible categories ... blanks, write-ins, alternative candidates. So why did this happen. Well, a good authority would be Wayne County Member of Congress Debbie Dingell of the long-time political dynasty in the area. Her Washington Post op-ed tells a tale of ignored working class voters that doomed the campaign. The punchline for the Michigan analysis is that overall turnout in Michigan was up in 2016 from 2012. There were over 92,000 more voters statewide casting ballots. The Clinton total was down over 300,000 from Obama. This wasn't an issue of third parties. Over 87,000 people took a ballot and didn't vote for President in Michigan. It was an issue of not running a campaign that got people's attention and made them feel that you were on their side. This is the inescapable conclusion. More on what that meant in coming days.