Man Without Qualities

Wednesday, September 18, 2002

Swedish Massaging by and of the Numbers

"So the key comparison number is really what a person can earn per hour in each country."


So if each worker in Country-S works, say, (0.1) hours a week at an hourly rate (with or without government subsidies and taxes) of $100, for a weekly income of $10, then those people are better off than the workers of Country-U who work 40 hours per week at an hourly rate of $20, for a weekly income of $800?

And the reason this is the "key number" is that the workers in Country-U "work far more hours per year than people in Country-S, so any increased annual income is 'bought' by giving up the vacations and holidays that the people in Country-S enjoy."

Is it true that one variable - what a person can earn per hour in each country - tells the largest part of the comparison story even where many more labor decisions in one country are taken from the individual and made at the government level? There's plenty of reason to think decisions at the governmental level are - and are even intended to be - less economically efficient than those made at the individual level. Does it matter whether or not the individual workers in Country-S would like to work more? Does it matter whether people in Country-S have smaller families than they would otherwise have because of perceived adverse economic prospects? Does it matter that one well-known problem with international economic comparisons is the tendency of such comparisons to grossly overstate the value of government services? The answer to these last three questions seems to be: Apparently not to Mr. Newman. Unless I missed it, Mr. Newman's methodology does not seem to account for any of that. And that old bugaboo about governments being obligated to make sure everyone had as much work as they wanted is now - presto! - transformed into the government making everyone "enjoy" all that work they don't have to do! Forget about economies being judged by whether they create jobs, the real measure of an economy is how much down-time they gin out. By this measure, think of how much richer the cronically un- and under- employed countries of the third-world are than they understand.

So why stop with a mandated 35 hour work week the way the French did and now reportedly widely regret and are set to relax? Why stop with (.01) hours of work per week suggested by the above example? Why not pass a law requiring everyone in County-S to "enjoy" 365 days of vacations and holidays per year and make everyone infinitely happy!

O frabjous day! Callooh! Callay!

If one really wants to isolate one single variable that gives the best idea of how much real wealth an economy is delivering to each of its citizens, isn't per capita GDP the place to start? [With respect to which, see below.] But, of course, no single variable tells all - or even most - of the story.

Further, Mr. Newman purports to make the difference by factoring in public benefits. But as one of InstaPundit's posts already summarizes, this point was promptly made by a second article, which attempted a "debunking" of the original study. The second (debunking) article pointed out, among other things:

The "differences in government and welfare services" are precisely the factors which the above studies failed to take into account. If you live in a country earning $100,000, but paying $99,000 in rent and health insurance, are you really better off than in Sweden?

And this "debunking" article was already itself addressed by yet a third article, which purported to take a position between the first two articles. All three articles are summarized in Instapundit - and they include many points. But Mr. Newman does not address those points - his approach is sweeping and employs rather charged language, which is fine, but makes rather clear that he is hardly personally and scientifically disinterested. Moreover, InstaPundit already linked to a purchasing power parity site, which Mr. Newman seems to think has not been done.

Further, some of the parameters under discussion already take government benefits into account. For example, the first article makes a big point that:"International Monetary Fund data from 2001 show that U.S. GDP per capita in dollar terms was 56 percent higher than in Sweden while in 1980, Swedish GDP per capita was 20 percent higher." Also, the first study already notes that: "If Sweden were a U.S. state, it would be the poorest measured by household gross income before taxes." Now, welfare benefits of all kinds are transfer payments (except to the extent the government runs a deficit). With respect to determining such average results it is not necessary to consider transfer payments - since one household's gain is another's loss. Mr. Newman is correct to note (although he is not the first) that differences in household size will skew the per capita results here. On the other hand, it is simply absurd to think that two people (for example) living separately but with the same income live as well as if they lived together and pooled resources in a single household. Isn't that one big reason that international economists look at variables on a household and a per capita basis? Of course, if household size were being set by efficient market forces, none of this would matter as much, since one might assume that some value was added in the choice to live in a smaller household for those that made such a choice. But one would have to be exceptionally blind to modern social realities not to note that household size can largely be determined by state welfare policies, which hardly count as efficient market forces. Indeed, while one cannot be too facile with international comparisons, the United States' experience with its 1996 Welfare reform Act strongly suggests that the efficient family structure is a world-wide casualty of many state welfare systems. And Sweden's is bigger than most.

Moreover, in some cases, per capita calculations are just wrong. For example, absolute per capita unemployment (in which all stay-at-home spouses are deemed "unemployed") doesn't mean much in most cases - but household employment is important, as is the employment rate of people who consider themselves to be in the job market.

Ultimately, it is some notion of "worker productivity" that must account for wealth production, and Mr. Newman's confidence in his sweeping approach and that a simple citation to some DOL figures will settle this issue is not shared by many international economists. As I pointed out in a prior post:

Another consequence of the odd subsidies the European way creates is its artificial enhancement to European “worker productivity.” As a subsidy to unions and established employee interests, European law imposes very high costs (compared to United States law) on an employer’s decision to hire a worker – especially young, inexperienced workers. One consequence is that an employer will only hire and retain workers who are productive enough to justify such high costs – other workers will become unemployed. Since “productivity” figures generally reflect the output of those who are employed, the European subsidy system tends to artificially enhance putative “productivity.” ...[B]ut the failure of higher productivity to result in growth and employment concerns European social planners, some of whom noted, for example:

“Faster economic growth will be critical in creating jobs and assuring the success of EMU. Historically, Europe successfully created jobs when economies were strong. Fast growth between 1986 and 1990 (3.4% annually) added 8.1 million jobs in the EU. Unfortunately many of them disappeared in the slow growth of the 1990s, (1.5% annually). …[W]e believe that EU economies must grow more than 3 percent annually in the first 4 years of EMU: in order to handle growth in the labor force, to accommodate productivity improvements, and in order to reduce the unemployment rate. Anything less would mean a definite risk that the public would blame the EMU for failing to address unemployment.”

International comparisions of worker productivity are tricky. For example, "Gross Domestic Product Per Hour Worked" is a measure of productivity less subject to the artificial enhancement described above. Such figures exist. However, such reports with which I am familiar are considered to be "experimental," carry warnings that they are to be used for detection of trends and not reliable for cross border comparisions, and require acceptance of some rather difficult conclusions. For example, ...the figures linked above rank the French economy as very slightly "more productive" (on a GDP/Hour basis) than that of the United States. But those figures also rank the Japanese economy as vastly less efficient than that of France or the United States. For example, the United Kingdom is defined on that productivity index to be 100, and for the year 2000 France and the US both show up at about 125 - while Japan appears at 90! Perhaps this is true. But one does not easily accept the suggestion that an hour worked by a French worker was almost 40 percent more productive than that of an hour worked by a Japanese worker in 2000. But then that study specifically warns that it is not to be used for cross-border comparisions.

Some of Mr. Newman's points seem rhetorically excessive, which does not encourage reliance on his sweeping conclusions. For example, his column rather strongly suggests that InstaPundit endorsed the first article (which got the whole issue rolling) - where InstaPundit clearly states that he "personally tend[s] to agree with the third article." Moreover, InstaPundit has already pointed out that he has not been discussing poverty rates - but Mr. newman (apparently, but not clearly, purporting to debunk InstaPundit) says: "When discussing the horrors of the Swedish welfare state, conservatives will bemoan the poverty and low wages of the Swedes."

I don't suggest that anything appearing above settles any particular point in the Sweden-Mississippi issue. For one thing, all three articles are in Swedish - which I do not understand. So the Man Without Qualities has no dog in this hunt. But the methodology of Mr. Newman's analysis exists at a rather peculiar level of sophistication, and his brief, undetailed column purports to be the first to address rather obvious points already addressed and counteraddressed in the very articles he purports to definitively supersede without any detailed analysis of those articles. His rhetoric is not disinterested, which is fine, but also signals his partisanship. Even his partisanship wouldn't matter if he had taken the time and effort to address the points raised by all three articles in reasonable detail. So it seems wise at least for the moment to hold in abeyance any decision that the matter has been put to rest - or even any decision that Mr. Newman has made a material contribution.

Comments: Post a Comment