≡ Menu


There was some confusion about my recent post on what economics is good for. I think it’s good for a lot of things. When I said it isn’t like physics I agree that that isn’t the end of the world. In one sense, who cares if it isn’t like physics? The problem is that too many economists and others treat it as if it were like physics. And most of these applications are silly, non-scientific and meaningless. Today’s Washington Post provides a good example:

According to Zandi’s analysis, the payroll tax cut will boost economic growth by 0.5 percent for the year, with most of the growth occurring in the first half. Low- and middle- income households are expected to benefit the most, as the payroll tax is only assessed on the first $106,800 in earnings. The tax cut is likely to create 1 million new jobs, he said.

That sounds scientific. A forecast of growth (using a decimal point!) and an estimate softened by the word “likely.” But it is a meaningless statement. It is scientism, the use of the language and tools of science to reach a conclusion that is not merited. The cut in the payroll tax could actually lead to more concerns about the deficit, discourage investment and risk-taking, and reduce employment in the future. Which view is right? This pessimistic view or Zandi’s? My argument is that there is no way to know. This is an abuse of economics. Here is a more accurate statement that Zandi might have made:

My model of the macroeconomy which has proved to be very inaccurate in the past, predicts a million jobs will be created. I am confident that I have made enough corrections to the model to insure that this forecast is accurate even though there is no way after the fact to assess the accuracy of my forecast.

I am sure Zandi’s forecast of job creation has some confidence interval around it. That is why he adds the work “likely.” That changes nothing. In a way it makes it worse–more scientific-sounding lingo like “confidence interval” when in fact there is no more science and no confidence either.