Dirty elections

To the rigger the spoils
Oct 22nd 2009
From The Economist print edition

New research on how many countries rig elections and why


HAMID KARZAI’S acceptance this week, through gritted teeth, of a run-off in Afghanistan’s presidential poll is a reminder both of how common rigged elections have become, and of how rarely incumbents fail to get away with it. Think, in the past two years, of Iran, Russia, Zimbabwe and Kenya—and those are just places that made the news. With two exceptions (Middle Eastern autocracies and East Asian communists), almost all countries now hold nationwide polls. And, according to Freedom House, an American NGO, an increasing number are unfree.

Faced with vast quantities of empirical evidence, economists like to run cross-country comparisons and regressions. So far, political scientists have done less of this sort of thing—so the more reason to welcome an attempt by Paul Collier and Anke Hoeffler, two Oxford academics,* to organise the political material into a usable database of 786 elections (clean and dirty) from 155 countries in 1975-2004 and to draw conclusions, striking and otherwise.

In the obvious category: rigged elections work for the rigger. Incumbents running in clean elections average six and a bit years in office; in rigged votes, 16 years. “Well, duh,” says Duncan Green, head of research at Oxfam, a British charity. Fair enough, it is obvious—but ten extra years may be more than expected.

Strikingly, the authors contend that “dirty elections are bad for economic growth by skewing politicians’ incentives.” This is because, they find, good economic performance makes a huge difference to an incumbent’s chance of re-election whether the vote is free or rigged, adding about three years’ to his or her tenure. Although economic success wins rewards in both systems, in clean ones, it adds 40% to a president’s time, whereas in dirty ones, the rewards of growth are swamped by those of rigging, which more than doubles the time in power. So rigging makes the economy less important to a president’s future—a rejoinder to the Chinese claim that in developing countries “managed democracy” is better for growth than an electoral free-for-all.

Obviously, not all developing countries rig the polls. Big nations seem less likely to rig than small ones—perhaps because they have more competing interest groups, making it harder to fake credibility by staging a poll win. Large government revenues from raw-material taxes makes rigging more likely by increasing incentives to get your hands on all that money. A few things make rigging less likely: term limits, the independence of the courts, parliament or press. And aid makes almost no difference. Even if outsiders are keeping the entire country afloat, their influence is patchy. As Mr Karzai earlier showed.



*“Democracy’s Achilles Heel”. Centre for the Study of African Economies. www.csae.ox.ac.uk.