The economics of scholarship allocation (3 of 3)

public economics ▪ policy ▪

It's been a while, so let's begin with a recap.

In Part One, I argued that the two principles guiding the allocation of public investments should be used to guide the allocation of publicly-funded scholarships: First, invest in activities with the highest social returns; and second, do not crowd out private investments.

Meanwhile, in Part Two, I described the two rules of thumb used to get around the problem of defining "social returns". These are, first, finding investments with high multiplier effects; and second, bias the allocation towards groups or regions that have the least number of educated people. I then illustrate how, in practice, these rules of thumb had been used to good effect in allocating Inpres elementary schools in Indonesia.

But there is a catch: The second rule-of-thumb, which seems to have worked well in allocating public investments in basic education, may not be good enough in allocating higher education scholarships. Why? Here's a clue: It has something to do with this post.

Before going into why this second rule doesn't work, let's begin by examining why it does.

One can think of education – or “human capital” in economic-speak – as inputs to production processes that generates the aforementioned social returns within a particular “economic unit”. An economic unit is usually defined as firms, but a tightly-knit community where its members’ economic activities are highly interdependent can also be considered as one.

Besides education, physical capital and manpower play equally important roles as inputs to production processes. We can call these three broad types of inputs “factors of production”. Combined together, these factors of production determine an economic unit’s output.

Now, for most goods, one thing that we know about these production factors is that less is more: The scarcest production factor usually is the most valuable. In turn, increasing this production factor usually bring the highest increase in productivity.

This last thing is quite intuitive. Just think of that time when your high priority assignment was stuck in the legal department because it’s overstretched. Didn’t you wish that the firm had added more people in legal so that the office can “work better”? At the same time, it wouldn’t make sense to add an extra department whose outputs highly depend on the legal department without expanding the latter.

This is exactly the reason why this second rule-of-thumb works for allocating basic education investments. If we think of Indonesia’s regions (provinces or districts) as economic units, the strategy of putting basic education investments in regions with lowest levels of per-capita education is optimal.

But I argue that this works less well with allocating higher-education scholarships, particularly the (highly-specialized) doctorate scholarships. I describe a variant of this problem in this post, but here, I’ll try to connect it directly with the problem of scholarship allocation.

As critical readers might notice, while it may be intuitive, the assertion that “the scarcest production factor is the most valuable” only applies for production processes whose productivity depends on the interactions of all three types of factors, i.e., human capital, physical capital, and labour. This describes most manufacturing processes, in which case, investment in the relatively scarce production factor is most optimal.

However, if the output can mostly depend on the interaction within one factor of production only (say, human capital only), than investing in just that one factor (and not the scarcest of them) is most optimal. This phenomenon is known among economists as “increasing returns to scale”.

This explains why the second rule-of-thumb is not a good guide in allocating PhD scholarships. How? Think of the main “outputs” of PhDs. Two that merit publicly-funded scholarships are teaching and research. A critical feature of both of these activities is that the quality of their outputs depends a lot on more human capital in similar study areas.

A PhD graduate in biochemistry working as a teacher-researcher is likely to produce more outputs working around PhDs in biochemistry than around PhDs in political science, and vice versa. Hence, two PhDs in biochemistry generate more social returns than two PhDs, one in biochemistry and another in political science. In other words, there is an increasing return to scale in higher-education investments.

What’s the implication? Abstracting from politically-motivated decisions, the sense that I got from the way (Fulbright) doctorate scholarships are allocated is that the selection committee still utilised the second rule-of-thumb in its allocation decision. That is, not only does it prioritise “backward” regions, it also seems to prioritise subjects that Indonesia has the least number of PhDs in.

I think this is not an optimal strategy. This strategy is only optimal for subjects that have immediate practical relevance (i.e., those that allow linkages with the private sector). For more research-oriented ones, increasing-return plays a significant role and hence, choosing candidates in “scarcest” subjects might give sub-optimal social returns.

Instead, a better strategy might be to choose candidates in subjects where there are quite a number of research-oriented (and teaching-oriented) PhDs already – such as, umm… I don’t know, economics, perhaps?