Her data trick is interesting: if you look at the 1920s and before,
the US data are really poor compared to those of the era after World
War II. Since the data problems were so bad, it's a mistake to
compare, say, the real GDP for the US in 1927 to that of 1957, since
two completely different things are being measured. Most economists
don't want to talk about this. Romer's trick was to say, okay, let's
calculate the 1957 GDP _as if_ all we had was the kind of data of the
type that was available in 1927. Then, since you're using the same
kind of base data to calculate the two GDPs, they are much more
comparable (though of course, data problem remain).

Her article that made the biggest splash suggested that the US economy
after WW2 was just as prone to gyrations as that before 1929 (or at
least that the two periods were hard to separate from each other in
terms of their instability). This surprising result tells us that the
post-war business about macroeconomic stabilization was but an
illusion.

The problem is that she should have included the 1930s as part of the
pre-WW2 years. To leave them out simply slants the results to produce
Romer's surprising conclusion. Also, people don't just care about the
fluctuations of real GDP: it's likely that the period before 1929
involved an economy operating at a higher unemployment rate (on
average) than from World War 2 to 2007 or so.

In any event, by today's degraded standards, Romer's much more liberal
than most macroeconomists.

On Sat, Sep 24, 2011 at 4:18 PM, Perelman, Michael
<[email protected]> wrote:
> Romer made her reputation  adjusting economic statistics in a way that 
> minimized the effectiveness of Keynesian-type policies.  I recall her 
> debating Robert Gordon, who seem to do a good job of minimizing her work.<

-- 
Jim Devine /  "Segui il tuo corso, e lascia dir le genti." (Go your
own way and let people talk.) -- Karl, paraphrasing Dante.
_______________________________________________
pen-l mailing list
[email protected]
https://lists.csuchico.edu/mailman/listinfo/pen-l

Reply via email to