Secondly, it would be nice to know about the effects of trade on the domestic economy.
A third purpose is to evaluate different kinds of policy. Here it is good to remember that most trade theory is based on neoclassical microeconomics, which assumes a world of atomistic individual consumers and firms. The consumers pursue happiness (“maximizing utility”) and the firms maximize profits, with the usual assumptions of perfect information, perfect competition, and so on. In this world choice is good, and restrictions on the choices of consumers or firms always reduce their abilities to optimize. This is essentially why this theory tends to favor freer trade.
One of the most important, and limiting, assumptions in neoclassical trade theory is that firms produce under conditions of perfect competition. Any industry that is controlled by a small number of firms is not perfectly competitive. There is a whole area of economics, initially developed by Joan Robinson in the 1920's, that explores what happens under imperfect competition. We won’t get into it in this course, but if there are significant “economies of scale” (which means that per-unit costs are smaller for bigger firms), then you can get very different policy recommendations out of your model.
Does this mean that the simple neoclassical models are useless? No. Their most important use is as a way to help you think through a set of issues. Neoclassical theory is especially good at pointing out the links between different markets. But you should be suspicious if you hear anyone saying that a theory “shows” that one policy or another is the right one in the real world.
The most famous neoclassical model is also the simplest — the model developed by the English political economist David Ricardo in the early 1800s. It’s simple because Ricardo assumes that there is only one “factor of production” (i.e. type of input) — labor. This model makes the point that trade should, in principle, benefit both parties even if one is more efficient. More sophisticated models were developed in the current century as economists learned more math. The best-known is the Heckscher-Ohlin model, named after a couple of Swedish economists, which is often called Heckscher-Ohlin-Samuelson (HOS) because of the important contributions made by the U.S. economist Paul Samuelson. HOS includes two factors of production (e.g. labor and land), and it shows that particular factors of production may be hurt by trade, though it still agrees with Ricardo that there are overall gains from trade.
There are many other varieties of trade theory, making
different assumptions and getting different results. One kind that
has gotten a lot of attention in recent years assumes increasing returns
to scale, which means that large producers are more efficient than smaller
producers. Ricardo, as noted above, assumed constant scale returns.
Neoclassical theories like HOS assume decreasing returns and get generally
similar results. If you allow increasing returns then bigger is better,
and one nation may end up dominating an industry, but it's hard to say
which nation will do so. In this case, the ability to intimidate
and bluff may be important. So increasing returns undermines the
ability of theory to explain or predict observed trade. Perhaps more
seriously, scale economies may stack the deck against late-developers.