Is Steve M at No More Mister Nice Blog right when he says, This Election Is Going to Be a Nail-Biter? He’s looking at the recent polls that show Donald Trump gaining on Hillary Clinton. Now I don’t think polls mean much right now. But I do keep remembering back to 1980, when Democrats were happy that they were going to be running against the extremist Ronald Reagan. And the data we have indicates that economic considerations account for about half of the outcome in a presidential election. So I thought it might be a good time to discussion some election fundamentals.
Polls Don’t Mean Much Now
Before I get to that, Steve M is being a bit silly. It’s not just that it is way early in this election. It is also that the timing is bad. We just came off the whole “rigged system” narrative that the right wing has been pushing about Clinton not getting indicted. Very few people even know what the issue is, but there is a tendency for people to think that smoke implies fire. We may be innocent until proven guilty in court, but in most people’s minds, we are guilty until proven innocent.
And let’s be prepared for things to get worse. The RNC is next week and that always gives candidates a bump. Just remember that Michael Dukakis had a 15 percentage point lead over the elder Bush after the DNC in 1988. And Dukakis went on to lose California — and didn’t win New York by all that much (4 percentage points — compared to the 28 that Obama beat Romney by in 2012). So we should all calm down.
Election Fundamentals Tell Us More
But to be honest, I haven’t liked what I’ve seen in the election fundamentals either. My simple model just looks at the trend of the unemployment rate (U3) in the first 10 months of the election year. If the trend is negative (unemployment is going down), it is good news for the party in power. If it is positive, it is good news for the party in opposition. And look at what we have for the first six months of this year:
- Jan: 4.9
- Feb: 4.9
- Mar: 5.0
- Apr: 5.0
- May: 4.7
- Jun: 4.9
To the naked eye, it looks flat, because when it comes to trends, we tend to look at the first and last numbers and go with that. But right now, this is a distinctly negative trend -0.017% per month. I know that doesn’t mean much without context. So here’s some context:
- 1988 (Bush-Dukakis): -0.033;
- 1992 (Bush-Clinton): +0.021;
- 1996 (Clinton-Dole): -0.50;
- 2000 (Gore-Bush): -0.007;[1]
- 2004 (Bush-Kerry): -0.033;
- 2008 (McCain-Obama): +0.179;
- 2012 (Obama-Romney): -0.046.
Hillary Clinton Is Winning
So right now the election fundamentals favor Hillary Clinton by just about the same amount as they favored Bill Clinton when he was first elected in 1992. And she’s doing considerably better on the election fundamentals than Mr Gore in 2000. If I had to pick between polls after a bad news cycle for Clinton and the economics at play, I would go for the economics.
Of course, the election fundamentals could change. Although Alan Greenspan was willing to let the unemployment rate get all the way down to 3.9%, it does not seem that the current Fed Chair Janet Yellen is. For years, the Fed has been itching to raise interest rates to slow our already anemic recovery. But I think the worst that will happen is that the unemployment rate trend will be flat. Unless something horrible happens, it is unlikely to go in Donald Trump’s favor.
What that means is that Donald Trump is going to have to make this election about something other than the economy. And that will be hard to do given that a bad economy is the basis of pretty much his entire campaign. The wall? The racism? China? It’s all based on economics. So it’s actually looking pretty good for Hillary Clinton right now. And I didn’t think that before I started writing this article.
I am not doubting your numbers or the primacy of economics in determining elections. What i am doubtful of is whether or not GDP numbers or jobs reports swing a lot of voters nowadays. I just have a hunch that net worth or the nature of one’s employment is more statistically significant than these abstract macro economic figures.
Between student loans, labor precarity and inflated home prices, there is a large contingent of college educated whites who are Democrats for life. At the same time, anyone with a good job and a home and a pension is so well off that they are reflexive GOP voters for life. It seems like if you are in a position where the DOW and GDP numbers are of real and material importance, you are likely to be a Republican already.
Meanwhile, you got evangelicals who are locked in as Republicans and you have besieged minority communities who vote for survival and reliably vote Democrat.
It just seems like every group knew how they would vote, long before this election cycle even began. Obviously, the final election tallies do change. Is it the better off subset of non college whites who actually change sides from one cycle to the next?
I’m not certain what the mechanism is, but it is definitely not that people look at the jobs report and vote based upon it. It seems to be that there are a fair number of voters who drift based upon their feelings. This is why we don’t see a correlation between how well the economy is doing and voting patterns. The unemployment rate could be 3%, and it wouldn’t matter if it had been 2% a year before. People have vague ideas that things are or are not going well.
But it is just a model. And even the most sophisticated models find that this economic effect is part of the total effect. I’m hoping that this election, the other part is going to hurt Donald Trump really badly. But we’ll see.
I suspect it’s a subset of every demographic you mentioned, the portion that states “I hate politics” and considers any joke where the punchline is “politicians are dishonest” or “government screws everything up” to be funny by default.
And these people aren’t wrong, given the information they’re going on. Certainly the conventional news coverage of politics focuses on all personality quirks, all the time; can you even remember the last time a policy decision made the news? (The ACA, maybe?) Even Trump’s bogus “wall” is covered as an example of his kooky persona, not a proposed multi-billion dollar construction project. (Trump’s genius, such as it is, is realizing that nobody will ever pay close attention to the details of what he says. That’s his real “art of the deal” — BS your way through life. Also, investing heavily in golf courses while inequality is vastly metasticizing, that doesn’t hurt.)
I suspect most of these “swing voters” go to the polls out of a sense of civic duty. And, being good social-minded folk, they buy the pablum about how “both sides need to work together.” So they vote “people in charge” when things are kinda going OK and “throw ’em all out” when things kinda aren’t. That’s my best guess, others may have better guesses.
I would say that a single poll never means much. A consistent lead over months is much more significant. Virtually every poll shows Clinton with a modest but substantial lead over Trump. I take that seriously; certainly more so than I would a single poll predicting a blowout against many others showing a tie (hypothetically).
Absolutely. But I don’t even trust the polls overall at this point. After the DNC, things should become clearer and the polls will be more meaningful. But no single poll means much at any time.