But look further afield and the affinity between open politics and open markets seems clear. As use of the ballot box has spread, especially to poorer parts of the world, tariffs have fallen. In 1981, for example, the world had only about 40 democracies; and the average tariff in developing countries was almost 30%. By 2003 the roll call of democracies had more than doubled and tariffs had fallen by more than half.
Why For one thing, voters are also consumers who do not like paying extra for imported goods. Democratic governments can withstand some consumer disgruntlement, especially if it is too thinly spread to swing many votes. But autocracies need pay it no heed whatsoever.
Voters are also workers. The poor countries, where many of the new democracies have flowered, are typically endowed with abundant labour but scarce capital. If they cut themselves off from trade, manpower will be cheap relative to capital. The plutocrats who profit from this economic isolation are the natural allies of autocratic government. Democracy, by contrast, enfranchises a wider circle of people who stand to gain from selling their labour at something closer to world prices.
For both of these reasons, freer trade often follows freer elections.
One recent estimate found that a transition from airtight autocracy to full-throated democracy (in a hypothetical developing country of average size, income, and so on) yields a fall in tariffs of seven percentage points, from about 22% to 15% or so. Indeed, Daniel Kono, of the University of California at Davis, claims this finding is among the most robust in the field of international political economy.
But it may be less robust than it looks, he argues*. Democrats may shy away from simple tariffs, but they still bash trade by other means. These include safeguards which pop up and down as imports surge and recede, and a bewildering array of sanitary and phytosanitary standards aimed ostensibly at keeping out pests and disease. Russia, for example, imposed onerous inspections on American poultry exporters, because it said their chicken legs, imported in great quantities after the arrival of democracy, might contain salmonella. The European Union banned Mauritania's award-winning camel cheese because the camels were milked by a pastoralist's hand, not by a gleaming machine.
By Mr Kono's reckoning, the transition from despotism to democracy results in lower tariffs but higher barriers of other sorts. Indeed, the share of imports touched by quotas, antidumping duties and the like would rise by seven points, he finds. Moreover, the product coverage of quality, health and safety controls would increase from less than 9% to more than a fifth.
The uses of obscurity
Tariffs may irk price-conscious consumers but at least they raise revenue for the public coffers. Why then do governments resort to other kinds of barriers, such as quotas and voluntary export restraints, which impose costs on consumers without raising any duty Their appeal lies in their obscurity, Mr Kono argues. Politicians indulge in optimal obfuscation. They resort to trade barriers that are difficult for voters to discern and tricky for political opponents to attack.
The burden of a tariff is easy to explain to the electorate: my opponent wants you to pay more for your milk and cars. Antidumping duties are a more slippery target: dumping does not sound like something a responsible politician should favour. And campaigning against health and safety standards can easily backfire: who wants to be in favour of drowning sea turtles in fishing nets so that people can eat cheaper shrimp
You may ask if such standards deserve to be attacked. If voters want to conserve sea turtles, ban shoddy imports and stamp out salmonella, democratic politicians surely have a duty to respond. Perhaps these trade barriers simply reflect genuine consumer concern. Perhaps. Mr Kono looks at several proxies for consumer sensitivities, including the stringency of a country's environmental regulations, the purity of its water and the number of quality-marks its companies receive from the International Organisation for Standardisation. Countries that fiercely enforce safety, greenery and quality at the border are not, he concludes, especially anxious to enforce these things at home.
In other cases, however, governments have promised to fight dumping in order to win support for radical trade reform. Several of Latin America's young democracies, for example, were keen to slash tariffs and peg their exchange rates to fight inflation.
They promised to defend companies against super-cheap imports as a way to sugar this free-trade pill. Mexico, for example, launched 83 antidumping investigations in 1993, more than any other country. But this was partly to shore up support for the North American Free-Trade Agreement.
Trade is probably still freer under democracies than under the alternatives. It is just that this hunch, as Mr Kono shows, is more difficult to prove than previous scholars had thought. Which is more damaging to trade: a tariff on Mexico's tuna or a demand that its fishermen show greater courtesy to dolphins The answer is obscure, optimally so.
*Optimal Obfuscation: Democracy and Trade Policy Transparency. American Political Science Review, August 2006.
The Economist Newspaper Limited 2006