Is it better to be a small nation?

Of the ten richest countries in the world in terms of GDP per head, only two have more than 5m people: the United States...and Switzerland, with 7m. A further two have populations over 1m: Norway, with 4m and Singapore, with 3m. The remaining half-dozen have fewer than 1m people.

The Size of Nations, a new book by Alberto Alesina and Enrico Spolaore, addresses why some small countries have done so well. Here is a related working paper by Alesina, here are some related working papers by Spolaore. As some of the larger empires of the past break up, questions of national size increase in importance. More than half the world's countries have fewer than six million people, roughly the population of the state of Massachusetts.

The Economist offers the following summary:

The book argues that the best size for countries is the result of a trade-off between the benefits of scale and the costs of heterogeneity; and that openness to trade alters this trade-off. The gains from being big are considerable. Large countries can afford proportionately smaller government (although they often don't). Essential running costs can be spread over many taxpayers. Embassies, armies and road networks are all likely to cost less per head in populous countries. Defence in particular is cheaper for giants. “It is only safe to be small in a peaceful world,” say the authors (who, unusually for economists, offer two stimulating chapters on conflict, war and the size of nations).

Large countries are able not only to spend more efficiently; they can also raise taxes in more cost-effective ways. Income taxes are more efficient than customs duties, but require a bigger initial bureaucracy. Large countries have bigger internal markets, allowing more specialisation and returns to scale. And they can redistribute resources geographically, providing insurance when one part of the country is hit by disaster or recession and shifting income from rich regions to poor ones.

So why don't all countries merge into one large superstate? Well, smallness has its benefits too:

...large countries are also likely to have a diverse population whose varying preferences and demands a government may find hard to meet: America, Brazil and India are cases in point. A study of local government in the United States suggests that Americans are willing to put up with the higher running costs of small municipalities and school districts in exchange for living in communities with little variation in income, race or ethnicity. This could imply that people also prefer to live in more homogeneous countries. With the main exception of America, successful big countries (such as Japan) have relatively homogeneous populations.

The authors argue that a worldwide regime of free trade will make the optimal size of nations smaller. If you can trade with other nations, there is no need to be large to ensure an open internal marketplace. So rising globalization should make secession easier to endure, which indeed seems to be the case.

My take: I am less convinced of the benefits of smallness. Think of small countries as having greater scope for experimentation, and thus a higher variance of outcomes. They also pop in and out of existence at a higher rate. Brazil will always be Brazil, but the fortunes of Croatia have varied over the years. If we look at the small countries that continue to exist, there is positive selection bias. We should expect them to do better than average, as the failures disappear, unlike with the less politically fluid larger countries. The observed superior performance of small countries does not mean that ex ante you should prefer to live in San Marino. In a small country, you face some very real chance that your system will fail, and that you will cease to exist, possibly under unfavorable terms. Especially if you are risk-averse, there is much to be said for the security of living in a larger nation.