Harder to beat than it looks

Harder to beat than it looks

I’m not going to write anything specific about what yesterday’s election results mean for America or the world, but in this post I thought I’d make a few random observations about the electoral process, polling and Democratic strategy, followed by two comments relating this election to past role-playing campaigns of mine.

  1. The US electoral process is a mess: The composition of the senate and the electoral college process are a joke that protects power for small rural states at the expense of the large and populous urban centres. That California (population 40 million) has the same number of representatives in the senate as Louisiana (population 4.5 million) is ridiculous, and ensures that Louisiana’s residents have nearly 10 times as much voting power as Californians; to a lesser extent this is replicated in the electoral college, where they have 8 compared to 55 electors. This is why we see the strange phenomenon of Republicans winning the presidency and the balance of power in the senate even though they do not win the popular vote, and cannot win significant numbers of senators in the cities. Furthermore, not only is voting not compulsory, but it is held on a Tuesday and many states don’t allow early voting or postal voting, or only have early voting in business hours. Every US election comes down to turnout, and this is a huge problem for a functioning democracy. Reform of all these aspects of the US system is desperately needed.
  2. Polls cannot predict anything while turnout is volatile: Polls consistently showed Clinton leading in the popular vote, which is what happened at the end, but they didn’t come anywhere near predicting the final result. I think this is because a) the polls don’t necessarily reflect the population of the state they’re taken in, so don’t reflect how it will vote, and b) even if they accurately estimate individual voting intentions, they need to weight this by turnout patterns in order to accurately estimate the final vote, and in the absence of accurate knowledge about who will vote, knowing how they would like to vote is irrelevant. Even guessing based on demographics won’t work, since we don’t know whether for example the white people who turn out to vote will be the Democrat voters or the Republican voters. In electoral systems with high turnout (e.g. Australia) this is not an issue, since the effect of fluctuations in turnout will be small compared to the total pool of voters, but in countries with low turnout this is a big problem. Especially since much of the result turns on subtle differences in a few states. Donald Trump would not yet have been declared victor if Pennsylvania were not his, and he won there by 1.2%. Even a small difference in turnout would flip that result. To flip that state Clinton needed just a 3% (not a 3 percent point!) increase in turnout. Note also that many states are won by such narrow margins that predicting the result would be impossible even if we had good estimates of turnout – the error margins on the turnout estimate combined with the voting intention estimates would surely swamp the margin of final victory, producing very high probabilities of error. Polling is no better than reading tea leaves in this situation. If the electoral college system were abolished this wouldn’t matter, since the law of averages combined with big margins in larger states would make polling more effective. But basically the only way to predict the result of a US election is to accurately estimate turnout using huge samples in a few swing states, and then give predictions like ‘there is a 55% chance Pennsylvania will fall to the Republican candidate’.
  3. Get out the vote strategies are not so wonderful: Trump’s ground game was famously bad, and his efforts to get turnout very poor, while Clinton was supposed to be running a well-oiled GOTV machine; yet 7 million fewer people turned out than in the last election, and Clinton got 4 million fewer votes than Obama. This tells me that Trump got better turnout than Clinton without having a ground game. This suggests that having a charismatic candidate is more important than turnout; and that a great GOTV effort is insufficient if the candidate is not well liked. From the party’s perspective this should probably be the only consideration. In four years’ time the Democrats should be asking themselves, is it better to have an inexperienced and popular candidate like Michelle Obama, or an experienced and unpopular candidate like Tim Kaine? And if they’re tempted to say “GOTV should make up for Kaine’s flaws,” they should look at yesterday’s disaster for some helpful pointers. Which brings us to …
  4. The demographic strategy is not working for the Democrats: One often reads that the Democrats are on a winning streak because the proportion of white voters is declining and the proportion of African American/Latino voters is growing, but there are two reasons why this isn’t working for them, at least in the medium term. The first is that the decline of white voters is due to ageing, and older people are more likely to be Republican, more likely to be able to vote, and more likely to be energized to vote; and the second is that fluctuations in turnout will wipe out even large demographic gains. When turnout can fluctuate by 10% between two elections (48.6% this election vs 55% in 2012, according to Wikipedia), demographic gains will be swamped by the patterns of turnout – especially if turnout is not consistent across all demographic groups. It’s also not clear to me that the growing trend in Latino/African American populations is a sustainable windfall for the Democrats, since the reason these populations are growing is that they are younger, and younger people are more likely to vote Democrat; but will this be true as these populations age? Democratic policies appeal more to young people, and the population of young people, although growing, may not be growing fast enough to offset the ageing of slightly older people into more Republican groups. If they are going to be competitive, Democrats need to appeal to older white people in rural areas, and that is very hard for them to do when those areas are completely shut off from Democratic modes of communication through Hate Radio, Fox news, and the growing echo chamber of the Republican right.
  5. Trump did not win poor people: The first exit polls I read suggested that Clinton did much better than Trump amongst people on below median income and below 50,000 US$ income. This group is disproportionately young and African American/Latino, probably also more likely to be women, and it shows that the Democrats are facing an ageing problem – this is the baby boomer dividend for the Republicans. In my experience Boomers are very vulnerable to climate change denialism, deficit terrorism, and arguments about deserving vs. undeserving poor, and this makes them easily convinced to vote for Republicans. This was an election fought along wealth lines, with a heavy leavening of racism and sexism to drive up turnout, and it’s not the case that Trump won by appealing to the poor and those “left behind” by the “neoliberal order”. He won by getting out older, wealthier people to vote against the change that their own children are pining for. This is exactly the same story as Brexit, where the people most likely to be affected by leaving the EU – young people and poor people – overwhelmingly voted to remain, while older and wealthier people voted to leave.

I can’t see an easy way back from this for the Democrats, not because they can’t win elections – Obama showed they can, and resoundingly – but because the Republicans will use their time in power to further drive down the ability of poor and young people to vote, and further attack the social organizations – like unions – that support activism in support of these groups.

I had a bet with two Aussie friends that the Republicans would be out of power for a generation, and I think my position was based on a misunderstanding of the importance of points 2 – 4, and now as a result I have to post an expensive 1.8L bottle of nihonshu (sake) to Australia. Which just goes to show the importance of understanding demographics, and also that this election has been a great tragedy for me, and Americans should apologize to me for my loss.

To bring this post back to RPG-related issues:

  1. A few years ago I played in a World of Darkness campaign that was set in a dystopic near-future, in which an inscrutable and ineffable evil force was working to reduce all the universe to its whims, using America as its primary point of access to the mortal world. Of course it was manipulating US politics through the Republican party, and so America had become a proto-fascist hellzone ruled by President … McCain. We thought this was hilariously cynical at the time, but now I think we were showing a remarkable lack of imagination. Shame on us.
  2. My character in the Cyberpunk campaign I recently played in was fond of saying that Asia was where the future was, and comparing shattered, collapsed America and failing Europe to the vibrant and optimistic megalopolises and future civilizations of Asia. This election shows the truth of her view: with Trump likely to sink all forms of action on climate change, China will become the global leader on response to warming; if Trump can repeal Obamacare the US will again be well behind many Asian nations in progress towards universal health coverage; and in comparison to the lunatic electoral decisions of the UK and US, the one-party administrative states of Asian states like Japan and Singapore are looking decidedly responsible and stable. I’ve said before that China is going to present a genuine alternative model to capitalist democracy if it can weather its economic and environmental problems without instability, and certainly the Chinese press have been presenting this US election as an example of why democracy is an ugly thing. As my Cyberpunk character was fond of saying (if her vocabulary extended to it), it’s time moribund European and anglosphere states started looking more seriously to Asia for ideas on politics and governance, because frankly, from my perspective, they seem to be flat out of ideas.

These are my first and probably last thoughts on the US election. I’ll be tracking Trump’s impact on Obamacare and writing about it as it happens, but the rest of this is too depressing for me to want to take on. Just the sight of a qualified woman being beaten to an important job by an incompetent, unqualified man with a history of workplace sexual harrassment allegations leaves me so cold I couldn’t watch her concession speech, and I certainly want to minimize my exposure to Trump before he forces himself onto us from the oval office. So I think I’ll be avoiding further posts about US politics for the foreseeable future … Godspeed America, I think you’re in for a rough and probably tragic ride.

Advertisements