## Recently in the sister blog and elsewhere

Why it can be rational to vote (see also this by Robert Wiblin, “Why the hour you spend voting is the most socially impactful of all”)

Be skeptical when polls show the presidential race swinging wildly

The polls of the future will be reproducible and open source

Testing the role of convergence in language acquisition, with implications for creole genesis

Also I’m supposed to be blogging on the election at Slate later today, but I’m not quite sure what the link is for that.

P.S. I’m live-blogging now at Slate; it’s here.

1. fraac says:

Isn’t there a corollary to Douglas Hofstadter’s superrationality theory that says if you make a decision it effectively ’causes’ all the people sufficiently like you to make the same decision, because you’ve been influenced by the same factors and people aren’t that different? So it’s rational to vote because you represent hundreds or thousands of votes (or only a few if you’re weird).

• Andrew says:

Fraac:

I wouldn’t say that if you vote for a certain reason that it causes other people to vote that way. Rather, I’d say that if you vote for a certain reason that other people like you will be making the same decision.

2. Jacob Egner says:

From the Robert Wilblin article:

>Over each four year term of a President, the US Federal Government will spend \$14.8 trillion.3

The “3” is to refer to footnote 3, which links to https://en.wikipedia.org/wiki/United_States_federal_budget which says that the USA federal government spend \$3.7 trillion in fiscal 2015. A quick google search shows recent USA GDP to be approximately \$17 trillion, so even saying he confused GDP and US federal government spending does not explain it.

So, I’m tempted to say Robert Wilblin is straight up wrong by a factor of 4, and it is weird that there is a very short chain to his own source contradicting him.

>If you multiply all that spending through a 1 in 10 million chance of changing the outcome, it comes to \$1.5 million. That’s the fraction of the budget you can ‘expect’ to influence by voting in a swing state, in a statistical sense. Of course, much of the Federal Budget is quite stable, and mostly determined by Congress

So he admits that the 1 in 10 million chance of changing a presidential election is inappropriate for the issue of a government budget controlled by congress, and then he admits that most of this budget is going be the same regardless of which of your choices wins, and this is on top of a budget number off by a factor of 4.

So, maybe a more proper crude calculation would be \$3.7T budget * 1 in 100 billion chance of changing party majority * 10% of budget actually different in a good way = expected \$3.70 (yes, roughly four dollars, no SI suffixes) benefit to spending an hour to vote, depending on how close the elections are that you’re voting in, and I feel I’m being generous in the 10% budget goodness swing, especially since you might be wrong. Heck, the outside view says there’s a ~50% chance you’re one of the people who is wrong. All of this cost-benefit analysis is further watered down due to the uncertainty between election victors and what policies/spending will actually happen, and then the uncertainty between policy/spending and actual desirability of outcome.

>For example, if one party wants to spend 0.5% of GDP on foreign aid, and the other wants to spend 0.3%, one vote alone could shift – in expectation – \$14,000 into foreign aid. If that were spent on bed nets – which unfortunately it usually isn’t – it would be enough to save 4 lives from malaria.

Some googling says that the US federal government spend \$35 billion on foreign aid in 2014, with \$5.9 billion of that being military aid. Also \$3.1 billion of economic aid goes to Israel, a developed country. So, (35 – 5.9 – 3.1) / 17000 = 0.15% of US GDP was spent by the US federal government on non-military, non-Israel aid. I keep on reading about how foreign aid is reliably a super low proportion of government spending. So, no, an extra 0.2% of GDP possibly going to non-military non-Israel foreign aid is not hanging in the balance between your choices on the ballot this year. And again, the author is using the inappropriate 1 in 10 million chance of changing a presidential election, when he acknowledges that congress is more important to the issue.

And then he mostly invalidates his point by saying that the foreign aid money spent by the government is spent on very suboptimal things. And again, there’s uncertainty about will your desired election victor really get your desired policy, and will the desired policy really be good? I think you could make some good arguments that some aid does more harm than good, which decreases the estimate of the goodness of your vote in the foreign aid arena.

It’s good when someone acknowledges counterpoints and caveats to their position, but when someone starts out with an argument and admits it doesn’t apply…then it feels like they just didn’t want to give up on their desired conclusion.

I liked the war impact discussion more. I wish he had spent more time on that.

• Jacob Egner says:

One error in my thinking was the budget issue. The author says over a 4 year term (and I misread), and thus his government spending figure does make sense. This error of mine invalidates my first quote-and-response, and it changes my expected-benefit-by-affecting-government-spending calculation from \$3.70 to \$14.30.

Many thanks to `xantes` from reddit for pointing out my error.

3. Jacob Egner says:

Of course, I welcome corrections and sanity checks to my thinking. For instance, if foreign aid spending seems to fluctuate a lot in a proportional sense dependent on what party is in power (as opposed to whether some specific sympathetic disaster happens, causing a spike independent of majority party), then that would be an excellent counterpoint to my argument.

4. Eli Poupko says:

While I’ve discussed this with you previously, Professor Gelman, and I know you disagree, I’d just like to go on record here regarding my take on the issue of “why it can be rational to vote.”

The resolution of the so-called paradox of nonvoting discussed in your The Monkey Cage article relies on the possibility that the perceived benefits of voting may be enormous, given that individuals tend to altruistically consider the collective social benefits of their vote, and not just their own selfish benefits. But this approach continues to assume that the probability of one vote having a causal effect on the outcome of a large election is minuscule.

I wish to draw attention to the argument of Richard Tuck in his 2008 book, Free Riding, who explains how the apparent paradox of turnout is founded on a misunderstanding of the meaning of causation in a collective action situation like voting. According to Tuck, individual votes have a causal effect on an election outcome to the extent that they may be necessary to forming part of an “efficacious set” of votes (I’ve written about how Tuck’s argument applies to the ex-ante probability of having a causal effect on an election outcome here: http://online.liebertpub.com/doi/abs/10.1089/elj.2014.0263. I’m happy to provide a copy of this article to anyone interested.)

The problem I see with the argument for social benefits making voting rational is as follows: It seems to me conceptually inconsistent to assume that voters think their vote has causally efficacy only when it is pivotal, and yet at the same time act altruistically by considering the collective benefits of their vote. To think that your vote only has casual efficacy when it is pivotal is tantamount to saying that you would prefer to free-ride (by abstaining) whenever you don’t expect the outcome to end in an exact tie (or in a one-vote margin). But wanting to free-ride rather than participate in collective action whenever possible is not the behavior of an altruistically rational individual who is socially concerned–rather I’d argue it represents the behavior of the conventional selfishly rational homo economicus.

It is interesting to me how I haven’t seen as much about the rationality of voting in this election as in previous years, even from the hard-core economists and economically-oriented political scientists. It’s almost as if scholars decided that the stakes were just too high this time around to burst the public’s illusion of how every vote matters.