What Does an ICOR Tell about Efficiency?

The concept of ICOR (The Incremental Capital-Output Ratio) is intuitively appealing, but its mechanical application may lead to unwarranted conclusions. Implicit in it is the notion of productivity. The intuition is as follows: a country with a high ICOR derives relatively little additional output from its investments, which suggests that capital is not very productive. Conversely, small investments can go a long way in a country with a low ICOR. However, this intuition ignores the way in which capital is combined with other factors of production, such as labor.

To illustrate the point, instead of two countries consider two sectors, such as SOEs and private enterprises. And start by the levels of capital and output, instead of their increments. In 2006, Vietnamese SOEs accounted for 51.9 percent of all capital in the Enterprise Survey of GSO, but they only generated 35.8 percent of the total turnover. It would be tempting to conclude that investments are considerably more productive in the private sector, and society as a whole would gain from reallocating capital away from SOEs. The same Enterprise Survey shows that Vietnamese SOEs employed only 28.3 percent of total labor. However, it does not follow that labor is considerably more productive in SOEs than in the private sector, and society as a whole would gain from reallocating labor towards SOEs. This time the conclusion is certainly wrong, even if the reasoning is basically the same.

How to reconcile the apparent logic of the reasoning in relation to capital with the patent absurdity of the conclusion in relation to labor? The difference between the concepts of average productivity and marginal productivity is useful in this respect. The reasoning in the previous paragraphs is based on average productivity (of capital and labor respectively) in the two sectors of the economy. But what matters for efficiency is marginal productivity. To know whether reallocation of resources should go in one direction or the other, it is indeed more useful to think about the effect of reallocating one worker, or one unit of capital, from one sector to the other.

A critically important strength of a market economy is that it takes care of this reallocation on its own. If a worker is more productive in one sector, employers in that sector should be able to pay a higher salary, and poach him or her from the lower productivity sector. The same happens with capital: a sector with more profitable investment projects should be better able to attract credit from banks. The problem is then whether market mechanisms are operating efficiently or not.

If SOEs have easier access to credit than private enterprises do, then indeed they may have too much capital compared to what is socially optimal, and the ICOR will be high. Similarly, if their personnel policies are influenced by non-commercial objectives, then their may be over-staffed compared to privates enterprises with a similar amount of capital, and the ICOR will be low. This is why a microeconomic analysis focusing on the distortions in the allocation of resources, all the way from the mechanisms through which funding is raised to the way in which investment projects are implemented, may be more revealing than an ICOR calculation.