|
1. The document was written in 1990, so the conclusions are 20 years old. The last data used are from 1985, so the 20-year-old conclusions apply to the state of affairs 25 years ago. That, alone, renders the post stale. But for the purpose of Saturday entertainment, let's not stop there.
2. The entire premise is squirrelly. They don't look at actual expenditures. They look at expenditures as a percentage of available resources. What their premise actually does is rank the the relative importance or ranking of education versus other things. They acknowledge this in saying that they show the relative "commitment". You have to read down to their "Conclllsion" section to find that. Of course, their "conclllsions" have some caveats, many of which they point out--such rankings aren't fully compatible. Why aren't they? They point out differences in centralization and heterogeneity of student populations. They don't say that an important caveat is that the relative wealth or poverty of a country matters. In that context, the year chosen, 1985, matters, simply because some of Europe was having a tough time (with national incomes down) while the US's GDP was increasing and deficit spending was sharply up.
Let's perform a roughly similar analysis on diet between us and a hypothetical family down the street. "We" are husband/wife/kid/cat, "they" are also husband/wife/kid/dog. We spend $900/month on food--a lot of chicken, milk, vegetables. They spend $400/month--a lot of beans, high-fat ground beef, rice. However, we conclude that we are much worse nourished than they are, because we spend only 22.5% of our take-home income on food, while they spend 38% of their take-home income on food. They spend 1.7 times more on food than we do, when adjusting for family net domestic product. Now, if you take the base not to be "take-home income," but gross family domestic income, then their expenditures fall to 36%, while ours fall to 17%, so they actually spend 2.1 times as much as we do. These researchers should predict that my kid would be malnourished. (And given that the increased income allows for far more crappy food they might well be right, making sense of the truth table entry that F --> T is a true inference; on the other hand, F --> F is also a true inference).
My "roughly similar" analysis is the same until we get to "gross family domestic income." Then it goes a bit astray. The point is that it assumes that a "unit of return"--education, nutrition--is independent of the amount of money actually spent but must instead be relativized to the particular resource set that the student might have access to. For my kid, since our income is higher a "nutrition unit" must actually require the same percentage of income than it does for a poorer family; the family down the street, with higher income, has to pay even more for the same nutrition. Yet when you go to the store, my $900 dollars, a smaller proportion of my family's income, still buys more nutrition than the down-the-street family's $400 does.
Similarly, to educate a kid in a country with a large GDP requires more money per "education unit" than in a poor country. However, at some point the delta ROI decreases to nearly nothing. I keep pointing out that my school district built a lot of new schools in the last 5 years--with wifi, stone and brick interiors, vaulted ceilings and really nice lighting. They replaced overcrowded, run-down schools. Lots of cash spent--and the only increases in provable educational outcomes is 100% accounted for in an analysis of variance by demographic changes as more Latinos moved into the neighborhoods. In other words, the base assumption--a commonplace in the '80s--is false. If you graph educational outcomes against dollars spent (or % GDP) the slope decreases to near zero. We're there for the most part. So's Europe. What's left in many places is gilding.
3. They continue with some other odd choices.
A. They astutely point out that nations running deficits can appear to be overspending (if you're just comparing expenditures with GDP). They like comparing education spending with total national spending. They then act like that matters. (The net effect is to drive down US education spending in the ranking; perhaps that alone justifies this approach in the researchers' minds.)
B. They point out that exchange rates influence the relative ranking, if you try to express things in a single currency is problematic. Then they point out that the dollar in 1985 was overvalued, so looking at US expenditures in terms of dollars overstates expenditures (thus accomplishing what the researchers want: Showing that the doesn't spend nearly as much as it should). Exchange rates are a recurring problem since seldom will anybody say that all currencies are properly valued. There's an index for this problem, purchasing parity power. This has problems, of course (nothing's perfect). Still, it's an attempt to get around a problem that the researchers don't like, and was a well-established methodology long before 1990.
C. We can quibble over the choice of countries used. They say as much. On the other hand, they specifically look at wealthier countries with rather large GDPs. They need to: They're looking at relative commitments to K-12 spending and in poor economies the relative commitments aren't made freely but under compulsion: You simply must spend so much on certain things, even if it crowds out education spending. Moreover, many weren't "free", in the sense of democratic. Still, it's no coincidence that Australia and Ireland rank below the US. The first has large military expenditures. In the '80s Ireland saw a burst in population growth (more kids, lower per-student expenditures), a budget crisis (insufficient growth) and a large deficit. Adjusting expenditures based on student cohort size makes sense.
The assumptions are perfectly justified and admirably chosen in order to reach utterly sound, iron-clad, absolutely incontrovertible conclllsions. As for actual conclusions, well, not so much.
There is a glimmer of conclusion there, though: We have expenditures based on what we're committed to. If families aren't committed to education, presumably the government won't be committed to education and won't spend as much money on education (usefully spent or not). Still, this has to be understood in context: Why spend $30k if you can get the same product for $1k? Moreover, you have to look at purchasing parity power. Most importantly, you have to make sure you're comparing apples and apples.
|