Why nations?
2007-09-27 · Original WordPress address
I agree with The Economist and Arnold Kling that the nation seems to be an arbitrary unit of measurement when considering economic outcomes.
To the extent economics is about efficiency, then we should talk about increasing efficiency for everyone. Trade is good, on balance, for everyone, foreign and domestic. Immigration is good, on balance, for everyone, natives and immigrants.
Yes, the "on balance" aspect of those statements is sticky. There will be some losers and some winners to trade and immigration, but "on balance" means that the gains of the winners outweighs the gains of the losers ((Rodrik claims many economists get mixed up here --- "A pervasive such false belief, for example, is that trade necessarily benefits more people than it hurts." --- but I don't know any of those economists.)). Economists give themselves an out by just saying the winners can compensate the losers so we don't have to worry about the distributional issues. For example, trade with Japan puts American car manufacturers out of business, but all of the tax payers who benefit from having higher quality and cheaper Toyotas compensate the auto workers by paying for their retraining.
Economics isn't just about efficiency, though, its about distribution too. Distribution is much harder than efficiency to get your head around. With efficiency, you either increase it or you don't. There's only one way to go. With distribution, on the other hand, its not obvious what the goal should be. Should we care about equality of outcomes? of opportunities? Should we care about the poorest members of society more than others? The answers to these questions aren't obvious.
It is obvious that we shouldn't choose answers to those questions arbitrarily. When Borjas argues against immigration, he has to arbitrarily assume natives are more important in the calculations of welfare than the immigrants. When Rodrik argues against trade, he has to arbitrarily assume the current distribution of jobs, the number of American auto workers versus Japanese ones, is more important than total welfare.
Of course distributional issues matter. Of course it matters that some people gain and some other people lose. But pointing this out does not make for a good argument against immigration (or trade). What matters is identifying those winners and loser and quantifying the degree to which the winners win and the losers lose... whether or not the winners are "us" and the losers are "them".
So, yeah, when deciding what is right and wrong, the nation seems to be an arbitrary unit of measurement. But I can come up with two reasons why the nation is a good way to study economic phenomenon:
- Nations have different institutions. If you want to test how institutions effect economic outcomes, nations are a pretty good starting point.
- Nations have statistics bureaus. Basically data is collected at the national level. There's starting to be some good micro/international data sets, but for the most part if you want to look at data across countries, its aggregated at the national level.