December 2001

Peaks and Troughs

Larry DeBoer
Professor
Department of Agricultural Economics
Purdue University

Here's a link to all the Capital Comments columns.

We’d been agonizing for months about whether we were in a recession or not. Doing worse than last year? No doubt. Layoffs in manufacturing? A bunch. Playing havoc with the state’s budget? Absolutely. But a recession? Maybe yes, maybe no.

Until a few weeks ago, that is, when the economists at the National Bureau of Economic Research declared that, not only is this a recession, but it started last March! Who are these people, and how come they get to mark the starting and ending points of recessions?

The NBER is a non-profit research group in Cambridge, Massachusetts. It was founded in 1920, and made its reputation especially for work on the business cycle, the succession of expansions and recessions that afflict our economy.

If you’re going to study expansions and recessions, you need to know when they happened. So the NBER combed through economic records and picked dates for the starting and stopping months of recessions, all the way back to 1854. The start of a recession is called a “peak,” when economic activity stops climbing and starts falling. The end of a recession is a “trough,” when activity stops falling and starts climbing again.

For example, the peak of the big one—the Great Depression—is marked in August 1929 (yes, before the stock market crash in October). The trough is marked in March 1933. We usually think of the Depression lasting throughout the 1930s, but the NBER puts the trough in 1933. That’s because the time from peak to trough is a “contraction,” when economic activity is falling. The economy sank so low during those 43 months that it took the rest of the decade to get back to where it had been. While it was doing so, though, it was growing, so the NBER calls most of that period an expansion.

Before March, the last trough they’d marked was in March 1991, the end of the last recession. The expansion of the 1990s lasted exactly ten years, according to the NBER. It was the longest expansion in U.S. history, more than a year longer than the expansion of the 1960s.

One really useful thing to do with peak and trough dates is to compare them to measures of economic activity. Take the “inventory-sales ratio,” for example. Retailers order products months in advance, trying to anticipate how much consumers will want to buy. Sometimes sales don’t meet expectations, and the inventory of goods doesn’t sell. Sales go down, inventories go up, so the ratio of inventories to sales increases. Eventually, retailers cut back on their orders from factories, not wanting to order more goods until their inventory is sold. Factories produce less, and employees are laid off. Retailers don’t order more until they’ve sold the excess inventory.

If you compare the inventory-sales ratio to the peak and trough dates, you’ll see that the ratio usually starts rising before the peak. That’s before the recession starts. The ratio starts falling during the recession, after the peak, but before the trough. Makes sense—manufacturers don’t start producing more products and hiring more workers until retailers sell the old inventories.

Looks like the inventory-sales ratio peaked in June, and has been falling since. Retailers are selling off old inventory. That usually happens before the recession trough, so maybe we’re on our way. On the other hand, a measurement called the index of consumer expectations usually falls before peaks, and rises before troughs. Consumers have to get optimistic before they start buying. We haven’t seen much evidence of rising consumer expectations yet.

We’ve been in recession since March. That’s bad news, but maybe there’s a silver lining in this economic cloud. It took the NBER eight months to identify the peak date. The average recession since World War II has lasted 11 months. If this recession is average, the trough will be in February 2002. That’s a glint of silver. By the time we know we’re in the average recession, it’s almost over.