昨日紹介したデロングの記事は、スティーブ・ワルドマンの直近ブログ記事でリンクしていたものだが、そのエントリでワルドマンは興味深いことを書いている。なお、これはModeled Behaviorのカール・スミスのエントリを受けて書かれたものである。

Smith is very right that we do not measure prosperity by consumption, but by production. We get wealthy by producing what someone values to consume. Our wealth takes either the form of current consumption or (more importantly in a deeply uncertain world) claims on others’ production whenever we might want or need it. We should never, ever measure prosperity in terms of consumption, but always in terms of production, weighted, importantly, by the strength of financial arrangements that allow us to convert unconsumed production into credible options to consume later. [*]

However, we do not measure prosperity in terms of how much work people are doing. That is a terrible error and the most vulgar form of Keynesianism. Make-work is not a path to prosperity, and effort is not production. Prosperity is a matter of the rate at which goods and services that are highly valued can be produced, whether the production is labor intensive or capital intensive, however much or little people are working. Employment is a more complicated issue than output. Under current (foolish) institutional and ideological arrangements, for most people, employment is our main source of claims on current or future production, and also a measure of our respectability and value as human beings. Holding institution and ideology constant, unemployment is harmful far beyond its effect on output. But let’s not confuse objectives. Employment and output are historically correlated, because it takes work to produce output, and also because the measured value of output is dollar weighted and there are few consumers to confer value without widespread employment. But the degree to which the production of any given thing is tethered to traditional employment is variable and technology dependent, and the ability to confer value through purchase could be distributed via means other than employment. Employment and prosperity are different things. (Please don’t confuse average and marginal values or invite ridiculous linear extrapolations by pointing to measured labor productivity. That statistic tells us nothing about the output we would get from further employment.)

We measure prosperity by production, not by work, and we measure production by value, by what people are willing to pay for what is produced.

[*] The caveat about financial arrangements is important with respect to Germany’s apparent prosperity. Germany, like China, is less prosperous than it seems, because its surplus production is geared to sale for claims that cannot credibly be redeemed for what the country’s citizens would want should they exercise their option to consume.




[*] 金融的手段についての注意書きは、現在のドイツの外見的な繁栄にとって重要である。ドイツは、中国と同様、見掛けほど繁栄してはいない。というのは、その余剰生産力は、国民が消費の選択権を行使した際に欲するだけのものを得られる保証の無い権利を見返りとした販売に向けられているからである。

コメント欄でAndy Harlessが上記引用部の最後の段落に対して異論を唱えている。

What people are willing to pay depends on the supply and demand for the numeraire good. If you’re taking money as the numeraire, the problem with this point of view should be obvious: inflation does not constitute an increase in production. I would suggest taking labor as the numeraire: it’s as good a numeraire as any other, and it represents something real and tangible that can’t be inflated. If we take labor as the numeraire, then work is production, and we can measure prosperity by work.

Now you probably will want to apply a sophisticated notion of value, where we use revealed preference to ascertain some (necessarily vague) kind of subjective value. I would suggest that this is a very difficult approach which ends up being no better than measuring value in terms labor. Once you start using subjective value, you open the door to questions such as distribution and non-pecuniary value, and the evil spirits begin to fly out of Pandora’s Box. If we measure value in terms of labor, it conveniently correlates with the distributional effects of wide employment and the non-pecuniary value of work.