Tuesday 5 May 2009

Comparing the growth effects of openness

Openness (equal to a country's imports plus exports, all divided by GDP) is often included in growth regressions. Other determinant variables may describe internal features of the economy, such as the saving rate, education, institutional quality, and political stability. Sometimes the determinant variables may include foreign direct investment and other measures of external exposure, but by no means always.

External exposure brings some disadvantages and some advantages for a country. The advantages include more optimal allocation of global short-term resources (but usually only short-term, not long-term, as a country may get stuck in producing cash crops, for example, rather than industrialising), and international technology transfers. By technology transfer, I mean use in production of information about methods or input combinations, where the information has been taken from abroad. Estimates on the growth importance of transfer vary, from quite important to really important (explaining over half of growth in some countries).

Openness is only one possible way of getting technology from overseas. Foreign direct investment, licensing, joint ventures, student and teacher movement, expatriate return, seminars and the internet are all possible sources of information about it, and may be more effective. Moreover, studies indicate that imports may be more effective for transferring technology than exports, and capital goods imports may be better than consumer goods imports. So having openness as the measure of external exposure will capture the (perhaps dominant form of) external influence on growth only partially and with considerable error.

The other variables in the regressions will pick up the effect of international exposure if they are correlated with the exposure, and correlation will often occur if people and companies in the economy try to maximise their income at all, since people are likely to adapt to encourage transfer if they think it is advantageous. So the effect of, for example, education or institutional form will be overestimated. Misspecification or incomplete specification increases the chance of interpretational error in estimations where internal factors seem to matter much more than external relations. On theoretical grounds, we may question this outcome, as it does not seem believable that so many countries are suddenly getting their internal arrangements far better than almost every other country in history, as measured by their economic growth.

No comments: