July 30 , 2003
Finally, the National Bureau of Economic Research has
decided that the past recession was one of the shortest and mildest on record.
For those still losing their jobs, the November 2001 ending to the
recession seems as ludicrous as declaring that peace exists in Iraq.
All of which raises the question of whether this business
cycle dating process makes sense anymore.
The economy has changed dramatically since Wesley Mitchell
and Arthur Burns concluded that economic cycles had consistent characteristics
that would allow economists to determine when an expansion was in jeopardy and
when a rebound was beginning. They
correctly noted that orders and building permits changed before employment and
Households and companies tended to accumulate liquid assets
before they made purchases. When
companies thought credit was getting easier, they borrowed short term.
When it was tightening, they tried to borrow long term.
By establishing a series of indicators that captured this
behavior, Mitchell and Burns believed that a barometer predicting economic
changes would be as effective as the weather barometer predicting weather
changes. This is how the whole
process of business cycle dating came about.
However, in those times, production and manufacturing were
dominant. The hourly workweek
changed before workers were hired or fired.
Inventories changed before production lines were opened or closed.
There was a pace of business that readily was reflected in the economic
That pace of business still exists for goods producing
industries. However, most economic
activity today is service oriented. Indeed,
we are even changing who qualifies for overtime hours.
Just think how much less useful that workweek measure is becoming.
Now, I do believe that new indicators may surface to
capture the mood changes by businesses and consumers that precedes their changes
in spending patterns. Adding
consumer expectations and the interest rate spread between short and long term
rates tries to capture some of that shifting process. Nevertheless, the current indicators simply are not as close
to prevailing economic conditions as the old indicators were, and they probably
never will be again.
So, are we wrong to date the timing of downturns anymore.
Just as we may suffer more casualties in Iraq after the large fighting is
over, we are suffering large job loss after the recession is over.
Just as Iraq is not yet peace, this is not yet recovery.
Perhaps we need another word to name this process.
Nevertheless, some useful information comes from the dating
of the recession. For instance,
construction spending begins falling quickly when a recession begins.
It then slows its rate of decline but continues to drift lower until
after employment growth returns. Normally,
this process takes about two years to thirty months.
We are approaching thirty months from the beginning of the
downturn, and private construction appears to be following this normal cyclical
Other patterns that are following normal cyclical behavior
are productivity changes, reduced inflationary pressures, interest rates (they
normally fall for a full year after the recession is declared over.
This time, it looks like the decline is about one and a half years, close
enough for government work).
Aside from employment changes, abnormal behavior is seen is
corporate profits, stock prices, and state and local revenues.
Profits fell on queue, but they plunged by 21 percent while the recession
was mild. In other words, profits
took the brunt of the economic weakness, just as they took most of the gains of
the previous expansion.
Stock prices lagged the downturn by a month and seriously lagged the upturn. As a rule, stock prices are a leading indicator, but they were very slow this time. When they decided to move upward, however, they moved up much more strongly than earnings. Normally, the ratio of stock prices to earnings stalls after an initial surge and remains below normal relative to historical experience until the expansion reaches above normal rates of growth.
This time, stock prices jumped much more quickly than
earnings, although they waited until earnings actually began to rise and
continue to grow much faster than earnings are growing.
Perhaps the shift to a service based economy has created a
new business cycle that requires an entirely new interpretation.
Alternatively, the old cyclical forces may still be at work, but we are
being pummeled by more external shocks than normal cycles endure.
We will need more information to know what explanation is more relevant
for today's economy.
In the meantime, those looking for a job should know that rising long term interest rates when short term rates are below normal and a rising stock market remain signals of improved activity ahead. If no more external shocks arrive, their job search persistence should begin paying off.