« Annals of Nature | Main | Why Oh Why Are We Ruled by These [Censoreds]? »

June 22, 2005

Mark Thoma Is Very Unhappy with Robert Samuelson

He writes:

Economist's View: Time to Toss Samuelson: Where do I start? I will agree with Samuelson on one point. One of us does not understand macroeconomics:

Time to Toss the Textbooks, by Robert J. Samuelson, Washington Post: Our ideas for explaining trends in output, employment and living standards -- what we call 'macroeconomics' -- are in a state of disarray.... Let me give you three examples. We once thought we understood consumer spending, the economy's mainstay. For decades, disposable income and consumption spending advanced in lock step. Americans spent a bit more than 90 percent of their after-tax income and saved about 8 to 10 percent.... But since 1990, consumer spending has changed. It has consistently outpaced income growth.... The main cause is the 'wealth effect.' In the 1990s higher stock prices caused Americans to spend more; now higher home values... are doing the same. So consumer spending increasingly depends on 'asset markets' -- stocks and homes -- and not just income...

So [Samuelson] says we don't understand consumer spending and then he explains it... [by] mak[ing] consumption a function of wealth, something we've been doing since the 1950's....

Next:

We can't determine 'full employment.'' Economists call full employment the 'natural rate of unemployment' -- the lowest rate consistent with stable inflation. Go lower and tight labor markets trigger a wage-price spiral. Unfortunately, we don't know what full employment is. The Congressional Budget Office now puts it at 5.2 percent. But past estimates have been too high and too low, because the 'natural rate'... constantly changes. It's influenced by population changes (younger workers have higher unemployment rates) and government policies, among other things....

So what's the point? Because it's hard to estimate exactly (even though he presumes to list the important factors that ought to go into a model of the natural rate) we shouldn't even try? That instead we should operate with no information at all about the target level of output?... [This isn't] a theoretical failing; this is a problem of measurement. We can't measure the ex-ante real interest rate exactly either (because it depends upon expectations), but that doesn%'t imply theory is wrong. Not at all.

Posted by DeLong at June 22, 2005 03:58 PM