George Bush and the Labor Market: Like Father, Like Son?

The first President Bush failed to be re-elected in 1992 because of the state of the economy. Odd, because the business cycle troughed in 1991 and the economy was expanding steadily right through the run-up to the election. That did not stop the Democrats from chanting "It's the economy, stupid"—and that idiotic slogan worked like a charm.

Unfortunately for the incumbent, the expansion of 1991-92 was a "jobless recovery." The now-familiar term was coined, in fact, to describe that period. As output expanded, growth in employment was far weaker than it had been in previous business cycles. This is surely worrying for the second President Bush, because job growth in the current economic recovery is even weaker than it was in 1991-92. Partly, this is because productivity growth in the current expansion is stronger. In itself, of course, the growth is good news: Ultimately, higher productivity means faster growth in living standards. The Democrats, if they are wise, will be cautious about claiming that the economy is in really awful shape. Nonetheless, the sluggish growth in jobs exposes the White House to a damaging line of attack.

All the more so if the Democrats can mingle anxiety about employment with a stiff jolt of xenophobia. No sooner said than done. The current alarm about the flight of service-sector jobs to India and other developing countries fits right in. It makes America's weak labor market explicable to public opinion—growing demand is creating jobs, all right, but overseas, not at home. And it points to a policy prescription that the Democrats can at least affect, for the moment, to embrace: protectionism of one vaguely specified kind or another.

In a previous column, I tried to argue that the concern about the migration overseas of service-sector jobs is greatly overdone. The numbers are tiny, set against the background of normal turnover in the jobs market. And if companies succeed in cutting costs this way, the American economy as a whole will benefit, exactly as Gregory Mankiw, chairman of the president's Council of Economic Advisers, pointed out recently, to shrieks of horror—some feigned and some genuine—from cynics and economic illiterates on Capitol Hill. Looking farther ahead, the faster the developing countries like India grow, the sooner they will be attractive markets for America's high-value exports. The whole "Exporting America" scare is, in fact, a populist scam. Many of the pundits promoting it know this very well, and they should be ashamed of themselves.

But if the flight of jobs overseas is not the cause of the current labor-market malaise, what is? And, to state the crucial question so far as the election is concerned, at what point will growth in jobs start to match the economy's impressive growth in output?

Employment was flat for 12 months after the trough of the 1990-91 recession, and then it picked up. In typical recoveries, net job creation starts sooner than that, within a quarter or two of the trough, and then it pulls strongly away. In the current recovery, well over two years past the trough in 2001, employment is only starting to revive and still stands below what it was at the trough.

Research published last summer by the Federal Reserve Bank of New York (downloadable from asks why the path of employment during the recoveries of 1991-92 and 2002-03 was so different from that typical of earlier business cycles. The answer is that the two most-recent recoveries involve unusually large elements of structural change.

"Cyclical adjustments," say the New York Fed's economists, "are reversible responses to lulls in demand, while structural adjustments transform a firm or industry by relocating workers and capital." Ordinary cyclical downturns cause workers to be laid off. When demand subsequently recovers—thanks to improving business sentiment, or renewed confidence among consumers, or fiscal stimulus (from tax cuts or increases in public spending), or monetary stimulus (from lower interest rates or a depreciated dollar), or all of the above—then workers are hired back into essentially the same jobs as before.

In a structural downturn, in contrast, jobs are permanently destroyed. And in a structural recovery, jobs are created that did not exist before. As a result, when the demand for labor increases, the economy struggles to marry the skills and experience of the workers displaced from the jobs that have disappeared to the new employment opportunities. This delays the labor market's revival. It may also skew the pattern of wages. If the skills needed in the new jobs are in short supply, the wages they can command will be bid up. Conversely, if the skills required in the old jobs are no longer in such demand, the wages they command will fall. Politically, this matters: The losers tend to feel more strongly than the winners.

One measure of structural as opposed to cyclical change is temporary layoffs. The New York Fed's researchers reason that the bigger the role played by temporary layoffs in raising unemployment as the economy declines, the smaller the structural component in changing employment. The thinking is simple. When a company makes temporary layoffs (often helping these workers to claim unemployment insurance, rather than seek new jobs, thus positioning them to be easily rehired in due course), it is expecting its business to be much as it was before, once demand revives. Permanent layoffs signal the opposite.

Tracking the last six recessions, the researchers found that in the four downturns before 1990, temporary layoffs rose sharply as demand contracted and then fell sharply during the subsequent recoveries, thus accounting for a big share of the total net movement in employment. The pattern on either side of the trough of 1991, and even more so on either side of the trough of 2001, is very different. Temporary layoffs hardly budged during either period.

Next, the New York Fed's economists looked at the direction of job flows before and after the recessions. The idea is that if job losses in certain industries persisted after the recession, or if job gains in certain industries were under way before the recession, then structural, as opposed to cyclical, forces were at work. The results point the same way as the temporary layoffs.

In the pre-1990 recessions, industries typically saw job losses during the downturns but then recouped those declines as the economy recovered. This "procyclical" pattern suggests that structural change was not a big factor. The pattern around the recession of 2001 was very different. The industries that lost jobs during that downswing—industries such as telecommunications, electronics, and securities and commodities brokering—were still losing jobs two years into the recovery. Conversely, industries that created jobs during the recovery—including other kinds of financial services—were creating them during the downswing, too. In pre-1990 recessions, the researchers estimate, about 50 percent of American workers were employed in industries undergoing structural change; in the recession of 2001, the figure was nearly 80 percent.

The president will find little consolation in knowing that structural change is good—because it is good in the medium and longer term, not right now. The ferocious pace of economic change within the country is the very dynamism that the United States has long been known for, only more so. It is merely another expression of what the country's remarkable productivity numbers continue to show: Prospects for raising living standards in the future have dramatically improved over the past few years.

But this process, driven chiefly by technological change at home, and much less by the migration of jobs overseas and the forces of globalization, is a painful business for many in the short term. More people than usual who lose their jobs will fail to find new ones—not primarily because of any failing of government policy, but as a result of the sheer pace of technological change. Whatever its long-term implications, that looks and feels like a sick labor market. It is something the Democrats can point to with advantage.

Things may yet improve for the White House. Bush can plausibly hope that the pace of productivity growth will slow in the coming months, and, so long as demand remains buoyant (which the Federal Reserve will strive to ensure), more of the present brisk recovery in output will then express itself as demand for labor. Layoffs are slowing down, and productivity growth is already overdue for its typical midrecovery cooling off. But the stickiness of unemployment, and especially long-term unemployment, implied by the structural character of this expansion is likely to give the Democrats a useful target between now and November.