Are recessions good for the economy?

Question from Reader E-mail: I have come to believe that a recession is good for an economy because it culls away weak businesses and teaches the strong to survive by cutting the fat that is not needed. I was wondering, though, do you think a depression is good for an economy and why?
A: I don't agree that even recessions are good for the economy. Joseph Schumpeter passionately argued in his 1942 book Capitalism, Socialism and Democracy that recessions are a necessary evil in capitalist societies. The idea that recessions are a necessary evil is still around today. Mark Rostenko, the editor of the Sovereign Strategist wrote the following in an editorial titled The Dips Don't See a "Double-Dip":

The "job" of a recession is to clean the "fat" out of the system, mop up excess, and pave the way for the next expansion. Until that process is complete, there isn't much from which a legitimate expansion can arise.

Recessions put weak companies out of business. In so doing, resources (skilled workers, capital) are freed up to be deployed more efficiently elsewhere. For example, Wall Street analysts who touted bankrupt Internet stocks are redeployed at local fast food restaurants to serve people in a capacity for which they are much better suited.

Stronger businesses that have used the contraction to firm up their bottom lines and grow more efficient are able to take advantage of these resources during the ensuing expansion. The economy emerges from a recession leaner, more efficient and in good shape for the next wave of growth and progress.

While the logic seems sound, it doesn't seem to match the data. If recessions were necessary to "clean the fat out of the system", we'd expect there to be a lot of bankruptcies and firm closures during recessions and little during booms. The data, however, does not support this as you can see in the table on the bottom of the page.

I have data for five different years, 1990, 1995, 2000, 2001 and 2002. The only year in the chart that overlaps with a recession is 1990, as the National Bureau for Economic Research indicates that the United States had a recession from July 1990 until March 1991. For the five years here, the GDP growth rate was positive in each year, from a high of 3.8% in 2000 to 0.3% in 2001.

Notice how little firm closures differ between these five years. We do not see great differences in firm closures between periods of high growth and periods of low growth. While 1995 was the beginning of a period of exceptional growth, almost 500,000 firms closed shop. The year 2001 saw almost no growth in the economy, but we only had 14% more business closures than in 1995 and fewer businesses filed for bankruptcy in 2001 than 1995.

Rostenko is correct when he claims that firm closures are a necessary part of capitalism since it allows "resources (skilled workers, capital) [to be] freed up to be deployed more efficiently elsewhere." When we look at the data, though, we see that we do not need recessions for this to occur; firms do not close that much more frequently in busts than in booms. So at least in this regard, recessions are not necessary at all.