From recent news reports, the American economy might appear to be in great shape. The estimated rate of growth in GDP in the third quarter was a very strong 8.2%, and the Dow Jones climbed back above 10,000 last week. The economic recovery now seems genuine. Furthermore, forecasts are for growth to continue strong, in the vicinity of 4%, into 2004.
This is very significant politically, of course. Past history suggests that the outcome of presidential elections is heavily influenced by the rate of change in the economy during the year preceding the election, rather than by either economic performance over the entire preceding four years, or longer-term prospects.
The truth, however, is that the economy does not look nearly as good from a longer-term perspective.
Democrats emphasize that the economy has lost well over 2 million jobs since President Bush took office, more than under any president since Herbert Hoover. Democratic presidential candidates may be hoping that continued high unemployment will deprive the second George Bush of a second term in 2004 as it did to the first George Bush in 1992.
In response to the jobs situation, Bush supporters take one of two very different tacks. The first is to emphasize that employment has begun to rise over the past five months. The hiring has just come with a lag after firms began to increase output two years ago. As in the recovery from the 1990-91 recession, firms have waited to hire until they are sure that the expansion is well-established, as it now is. The second response is to argue that the divergent paths of output and employment represent an economically beneficial acceleration of long-run productivity growth. The merit in each of these arguments is, however, limited.
Contrary to some recent reporting, the recent expansion of employment has been anemic. Even measured relative to the lowest point in June, the average increase in payroll jobs has averaged a mere 54,000 per month. To be sure, this is an improvement over the preceding steep loss of jobs. But by any other benchmark, it is a poor performance. During the eight years of the Clinton administration, the growth in employment averaged 240,000 per month.
In October, Treasury Secretary John Snow set a benchmark of 167,000 jobs per month to be created by the election. The recent numbers are not enough to keep pace with the rate of growth of the population, let alone to make up for the previous decline.
Some say that the standard payroll employment numbers, which come from the non-agricultural establishment survey of the Bureau of Labor Statistics, substantially understate job growth. They say that we should instead look at the household survey –– which shows a faster rate of growth in employment. Their argument is that the government statisticians are behind the times, failing to recognize the importance of new start-up firms, which are missed in the survey of business establishments.
Unfortunately, this argument is wrong.
The establishment numbers probably remain the more reliable of the two sets of statistics. The Bureau of Labor Statistics has made progress in recent years in adjusting the numbers for the effects of new start-up firms. It is the critics who are out of date. The “joblessness” of the recovery to date is genuine.
The other line of defense is that the jobless expansion represents an acceleration of long-run productivity growth, which is good for the national economy in the aggregate, even if it is hard on those who are laid off. It is true, as a matter of definition, that if GDP grows more rapidly than employment, then productivity is growing. It is also true, in the long run, that productivity growth is the means whereby countries are able to raise their standard of living from one generation to the next.
But an alternative interpretation of recent productivity numbers is possible. During the 1990s, corporate executives and stock market analysts were extravagant in their reporting of contemporaneous corporate profits, and overly optimistic in their forecasts of future profit growth. Ever since the corporate accounting scandals began in 2001, executives have been under pressure each quarter to perform more closely to those promises, by delivering profit growth that is strong in reality. Under this alternative explanation of the employment numbers, firm managers have been desperately cutting labor costs in any way that they can, a squeeze that is not sustainable in the long run.
How can one tell which explanation of the jobless recovery — short-term cost-cutting or long-run productivity growth — is the right one? If we were seeing an acceleration of the productivity boom that began in the 1990s, we would see workers sharing in the benefits, in the form of higher wages, as they did in the late 1990s. Instead, real wages have been almost flat in recent years. The failure of firms to share productivity gains with workers supports the hypothesis of desperate short-term cost-cutting, and not the hypothesis of an improved long-term productivity trend.
Employment is bound to recover substantially before long, given how far it fell in 2001 through 2003. If one is looking for the adverse implications of Bush economic policies, it is better to look five or 10 years out than to look at the coming year. In fact, good economic logic does not support the idea that Bush fiscal policies caused the weak economy of the last three years. Good economic logic supports, rather, a causal link between Bush fiscal policies and the next recession. The future downturn is likely to be far worse than the recent one.
The problem is the remarkably ill-conceived composition of the Bush tax changes legislated in 2001 and 2003. It is almost as if some aspects of the tax cuts were composed to minimize bang for the buck. Such measures as the decision to abolish estate taxes in the year 2010 and the reduction in dividend taxes, without offsetting future revenue gains, flunk most tests for an intelligent tax cut. The tax changes were not designed to have most of their effect in the present, when we needed the boost. Moreover, they were designed to redistribute from poor to rich and from consumption to saving; thus they have provided only a limited stimulus to consumption. Worse, the dividend and estate tax measures were designed to produce large revenue losses later in the decade. Thus they minimize expansionary bang and maximize the long-run buck.
A sensible response to the 2001 recession would have included some tax cuts, but would have had a mix much better targeted to stimulate the economy, without sacrificing long-term fiscal discipline. A well-designed plan would have avoided setting the future economy on a long-term path of rapidly rising national debt. As it is, interest rates are likely to rise for the remainder of this decade, crowding out private spending.
It is impossible to say when the next recession will come. But when it does, it is likely to be worse than the 2001 recession. Why? Precisely because we will enter it at a time when the budget deficit, national debt and international borrowing are already alarmingly high. Thus when the next recession hits, we will not have the luxury of being able to cut taxes and increase spending as the current President Bush has done. If anything, we are more likely to be in the midst of raising taxes, as the elder President Bush had to do in 1990.
This will exacerbate the downturn, but we will have no choice. The resulting pain will make today’s economic travails pale in comparison.
This story "Bush’s Jobless Economic Recovery: Workers Pay Price for Productivity" was written by Jeffrey Frankel.