A Convenient Untruth

It is by now a non-controversial economic fact of history, first described by Robert Solow in the 1950s, that economic growth and a society's level of welfare is symbiotically linked to increases in productivity.

Unfortunately, productivity is as hard to pin down precisely as it is important. As a result, its all-too-convenient use in political rhetoric — as is prevalent in the U.S. presidential campaign — is fraught with considerable danger.

Indeed, one of the standard economic ways to measure (multi-factor) productivity is to not measure it at all. Instead, this particular approach tries to account for everything else — the tools, raw materials and labor that go into generating a nation's goods and services. Then, one looks to see what the residual that one cannot explain in the other categories comes out to be.

This is subsequently taken as an expression of just how productively workers have used their tools and raw materials to produce the goods and services that they did.

In a sense, then, what both John McCain and Barack Obama are really saying is this: "Well, compared to other countries, there seems to be more of the stuff that American workers make that we can't really explain where it comes from. So we have to assume they are better than others at making it!"

A key question in all economic data-gathering is of particular importance here, too: In which denomination should we measure these results? How much of precisely what are we talking about?

One frequently used measure that would intuitively seem to correspond relatively closely to the pontifications of U.S. presidential candidates on the subject would be to say: "What is the value of the output produced by each American, relative to that generated by people in other countries?"

Usually, this question is answered by referring to GDP per capita, adjusted by the Conference Board for purchasing power. A look at the latest data shows that U.S. levels in 2007 were at $45,547, which was significantly above the levels of Japan ($33,683), Canada ($38,500), Germany ($34,008), France ($33,344), Britain ($35,193) and Mexico ($12,356).

It therefore seems appropriate to say that U.S. productivity levels by this measure are significantly (by about 30%) above those of most countries with which the United States generally likes to compare itself.

However, Luxembourg's GDP per capita was $79,774 in 2007. So given these numbers, it would seem that people in Luxembourg were fully 75% more productive than U.S. workers — who, in other words, are actually not the most productive in the world.

Yet recall, these are "country averages" — and there are big differences among countries. Hence, being 75% less productive than Luxembourg (a city-state of fewer than 500,000 people with very liberal tax laws and banking secretly targeted to German tax dodgers) is not necessarily such a bad thing for a big, diverse country like the United States. It furthermore seems likely that when the data for 2008 come in, Luxembourg — which is dependent on the financial sector — will be rather more affected than the U.S. national average.

Incidentally, the latest data available from the Bureau of Economic Analysis concerning GDP in U.S. metropolitan areas show that several U.S. metropolitan areas — like the hedge fund capital of Stamford and Silicon Valley — are more productive than the city-state of Luxembourg. At the same time, of course, financial centers like Stamford will likely also look less shiny when the 2008 data is available.

However, GDP per capita is just an average across the entire population. What we are really interested in here is how productive the "American worker" is, relative to workers in other countries.

We don't want any distortions from, say, young people in school, the unemployed or retirees. Therefore, it is perhaps better to ask the question: "What is the value produced by each American worker, relative to others?"

Here, the data reveal that GDP per U.S. worker employed in 2007 was $92,934. This compares to $66,942 in Japan, $74,917 in Canada, $70,692 in Germany, $80,600 in France, $74,986 in Britain and $30,589 in Mexico. That is safely ahead of other major economies, but once again less than Luxembourg, where each employed worker generated over $115,000 of GDP in 2007.

Maybe another approach is in order. For example, the whole world knows that the French are such lazy people that they never work more than 35 hours a week and have at least a month of vacation every year.

Yet, each employed French person — putting in a lot fewer hours in the office each year — still generated 90% as much as each employed American. That indicates that the French are really more productive than Americans — when they actually do work, that is.

The way to sort that out is to determine the GDP per hour actually worked in 2007. Indeed, the data show that each American worker generated on average $52.31 for each hour on the job. Lo and behold, though, each Belgian, Norwegian and Luxembourger generated $53.41, $69.48 and $71.29, respectively, per hour worked — and, as such, can reasonably claim to be more productive than their American counterparts.

Americans are, of course, still richer — but that's only because they work much longer hours — not because they are more productive!

Based on the evidence gathered so far, it seems as if the U.S. presidential candidates are just trying to flatter their constituents — rather than telling them the truth about how well they really measure up against other nations' workers.

Well, wait another minute. All these GDP data are, of course, averages. And, as always, with averages, if one can exclude those least valuable, one gets a higher average. The European countries listed above all have far fewer people working as a share of the total population — and they work fewer hours.

Given that far more Americans are employed for far more hours per year than is generally the case in Europe, this means that far more hours are included in estimating the U.S. average than in the European ones. That makes the United States look relatively worse and less productive per hour.

So maybe an ultimate attempt at adjusting the numbers to the Americans’ advantage would be in order. How productive would the average European worker be per hour worked — if as many Europeans worked as many hours as Americans do? And how would that compare with the United States?

Amazingly, it is possible to make such an estimation — of the so-called "structural productivity levels" — comparing similar levels of labor input. That allows one to adjust for the fact that a lot fewer (typically low-productivity) younger or older workers and women are employed in Europe.

At long last, it indeed turns out that, with such adjustments making other societies work as much as Americans, then U.S. workers are in fact more productive per hour — or so it seems. Alas, even that is with the exception of Luxembourg (and, benefiting from their offshore oil wealth, Norway).

However, making such a statement to their voters — that Americans are in fact more productive at being Americans than anyone else — isn't likely to win presidential candidates as many voters on the stump.

Which is why the whole world can count on the fact that they will go on reciting an obvious untruth — that Americans are the most productive workers in the world.