Voter turnout and military coups
In my last post, I wondered if there was any evidence that low voter turnout was a problem. Patrick Crozier has been thinking similar thoughts, and, like Patrick, I am convinced that it matters not a jot. What matter are the policies that the parties offer, and since they are trying to win an election by getting as many votes as possible, in a two party system, we may expect, to a first approximation, both parties to offer very similar policies that satisfy the preferences of the median voter . If the parties are offering bad policies, this may be attributed, in part, to the rational ignorance and rational irrationality of the median voter . Thus, if we wish to argue that high voter turnout is a good thing, we would have to show that the median voter of the set of all people entitled to vote is better informed and more rational than the median voter of the subset of people who exercise their vote. There is no a priori reason to suppose that this is so. Indeed, if we could encourage the view that if you take little interest in politics and economics, and spend little time trying to understand how policies would work, then it is perfectly acceptable and responsible not to vote, then we may even see the parties offering better policies.
However, there is another argument that can be made. It is possible that high turnouts may help to prevent coups. A revolutionary leader has to enjoy widespread support, or else he may be faced with a counter-revolution. In a democracy, there is a common perception that the winner of an election is somehow morally entitled to rule. So, is it not possible that electoral turnout may signal to would-be coup-plotters how much resistance they may expect in the event of their enacting their coup, and so influence their judgement about whether it is worth risking it? With a high turnout, most people accept that the winner is legitimate and so any coup would be met with stiff resistance. With a low turnout, most people regard the election as a sham, and so a coup leader would be, in their eyes, no less a legitimate leader than any winner of an election, and if they can't be bothered to vote, they are even less likely to risk their lives to help put down a revolution. This is what I had in mind when I asked my question a couple of weeks ago.
I still haven't found any academic research into the consequences of low voter turnout (although, to be honest, I haven't spent much time looking), but I did find the website of the International Institute for Democracy and Electoral Assistance, where they have published turnout data for general elections around the world since 1945. If we are willing to select our anecdotes and commit the post hoc ergo propter hoc fallacy, then we could point out that in Pakistan, the general election in 1997 that preceded General Musharaf's coup of 1999 had a turnout of 35% of registered voters (or 32% of the voting-age population (VAP)); in the Phillipines, Ferdinand Marcos's suspension of parliament in 1972 and assumption of absolute power in 1973 was preceded by a general election in 1969 in which only 13% of the VAP voted (due to a low registration rate - there was a turnout of 89% of registered voters); Ne Win led a military coup in Burma in 1962, the election of 1956 had a turnout of 35% of the VAP (52% of registered voters) (according to the BBC's timeline for Burma, there was an election in 1960, but IDEA's data appear to be incomplete); in Thailand the general election of 1946 had a turnout of 27% of the VAP (32% of registered voters) and was followed by Phibun Songkhram's coup of 1947; Chad had an election in 1997 in which 26% of the VAP voted (50% of registered voters) and in 1998, Youssouf Togoimi launched an armed rebellion against the government; Kenneth Kaunda turned Zambia into a one party state in 1972, the election of 1968 had a turnout of 32% of registered voters (50% of the VAP - out of a VAP of 1.9 million, only 3 million people were registered to vote); Idi Amin became ruler of Uganda in a coup in 1971, the election of 1962 had a turnout of 32% of the VAP (67% of registered voters).
Of course, an analysis like that should not be taken seriously. With a little more effort, I'm sure we could find examples of stable democracies that often have low turnouts at elections and a proper analysis would have to account for all sorts of other factors that might contribute to a situation that would result in a coup. I'd also bet that there is too little data to get a statistically significant result. Nevertheless, it does suggest that I'll have to be careful when I use my snappy response to any scolding I get for not voting.
Here's a joke about the downfall of democracy:
Johnnie Beattie, the Scottish comedian, is reputed to have attended a fancy dress party dressed in a soldier's tunic, with a single angel's wing on his back and udders dangling between his legs. When asked what he was dressed as, he said, "I'm a right-wing military coo." (It has to be read in a Scottish accent to work.)
 Any differences in the parties' policies may reflect uncertainty as to what the preferences of the median voter are, deviations of voters' preferences from a strictly one dimensional spectrum, deviations from a two party system, collective irrationality of internal party politics etc.
 This doesn't mean that the median voter is personally responsible for the parties' bad policies: if he recognises his rational ignorance and rational irrationality and strives to correct them, he will merely join the sect of irrationally informed and irrationally rational cranks, and someone else, who nearly shared his original preferences, will assume the position of median voter. Only if large numbers of voters change their preferences will the parties' policies significantly follow suit.