More Reflections on Democratic Slide

There’s been a great deal of fanfare over what is or is not democratic behavior in the last couple weeks. Certain millionaires attempting to buy votes in the open. Certain other millionaires stifling political speech. And so on.

---

I haven’t given much weight to the polling for this presidential election, because I think the circumstances have eroded our ability to course correct and re-calibrate. Biden bowing out so late in the process means that Harris experienced a very typical ‘announcement bump’ in popularity numbers, but at a very atypical point of time. Then we got the Chicago DNC, and again Harris experienced a very typical ‘convention bump’. Somewhere in that mix we also had the Milwaukee RNC and an assassination attempt. Cumulatively, I don’t think the polling numbers were worth anything until about mid-September.

A significant degree of my trust in election polling comes from the simple fact that we do have intermittent events to test predictions. Primary elections (and related institutions) are the ahem primary example. They are spread over time, held among mutually-exclusive subpopulations with known aggregate characteristics, and the survey recruitment and screening process is extraordinary. Fantastic.

Obviously it is too late for calibrating the predictions to the primaries. We are left with reasonable correlates. Fundraising numbers reflect the way a small subpopulation will vote. And they are ‘putting their money where their mouth is’, so I consider it a reliable signal. Rally attendance numbers reflect the way a large subpopulation will vote, if they turn out, which is to say with some greater level of uncertainty. Betting markets, crowd reactions at debates, etc., etc. This is all to say, there are some correlates that can help to validate polling.

My takeaway from the polling and correlates since mid-September is that things are too close for a serious prediction.

I think that conclusion flows from a quite large margin of error. I do understand that some people interpret the high variance as ‘hedging bets’. I won’t argue against that line of thinking. But I think we also ought to recognize that this isn’t a good case for polling in the first place.

---

Just before the 2016 election, a huge list of political scientists took an unprecedented step of signing and circulating a public letter against Trump’s candidacy. While it wasn’t an official APSA communique, it was signed by the contemporary president and several former presidents. It was understood before he took office that Trump was a threat to democracy.

We dealt with his presidency for 4 long years. Absurdly discriminatory policies were implemented at the highest level, and the Supreme Court took the principled stance of bending over. Research and development were set back years in multiple strategic areas, but especially climate science. Institutional expertise was thrown out the window through reckless decisions like relocating the Bureau of Land Management. We walked on our closest international allies, and ceded regional power to Russia. The nation’s Freedom in the World score fell from 89 to 83.

He has promised to repeat these and worse with a second presidency. The Supreme Court has preemptively assumed a subservient position. And then there’s the overt intrusion by millionaires to whom I alluded at the top.

Things have jumped the shark from awkward public disavowal of any intent to rebel in January 2021, divorced of course from the reality, to boasting about doing it again.

---

Altogether now:

I don’t know how to reconcile with the fact that this country is seriously considering re-electing Trump.

This election is not a runaway for Harris. I don’t really know what to expect tomorrow, but it’s close enough to be disappointed.

I could complain all day about the democratic slide around me, but there’s at least a sizable plurality of active voters who want in on that ride.