The recession probably ended months ago but employment has not yet started to recover. Is America’s jobs machine broken?

America's labour market

Something's not working
The recession probably ended months ago but employment has not yet started to recover. Is America’s jobs machine broken?
Apr 29th 2010 | From The Economist print edition

IN THE summer of 1984, Ronald Reagan declared “morning in America”. Running for re-election after a tumultuous first term during which unemployment rose to a postwar record of 10.8%, the president needed a hopeful message to take to the American people. When Reagan’s television advertisement ran, nearly two years after the official end of recession, unemployment was still high, at about 7.5%. But the president could accurately claim: “Today more men and women will go to work than ever before in our country’s history.”

In the spring of 2010, a similar dawn is some way off. Economic recovery probably began ten months ago, but employment has scarcely risen from its trough (see chart 1). About as many Americans are working as in the autumn of 1999—in a population that is larger by 28m. If Barack Obama wants to repeat Reagan’s boast in his own re-election campaign, by which time the recovery will be three years old, the economy will need to add jobs at a rate of more than 3m a year. That has not happened in over a decade.

It is well known that America shed jobs at a furious rate as recession took hold. The rise in the unemployment rate between boom and bust, of nearly six percentage points, was the largest since the Depression. Adjusted for the age distribution of the population, the jobless rate hit a postwar high. So did the number of long-term unemployed, reaching 6.5m, or nearly half the total (see chart 2).

What is less well understood is that the American jobs machine has stalled badly. Granted, employment has been rising—non-farm payrolls were up by 162,000 in March—and may gain strength in the months ahead as cyclical joblessness falls. But the recovery has not yet brought jobs for the millions of workers idled by recession. This is at odds with what economists refer to as Okun’s law, a rough but empirically regular connection between changes in GDP and changes in unemployment. Over the past 12 months unemployment has not fallen as quickly as economists have come to expect. The hitherto tight link between economic activity and job creation looks disturbingly slack.

To see how this has come about, it helps to divide the recession into three phases. In the first, in 2008, workers were laid off rapidly, particularly during the second half of the year. At the same time, both hiring and vacancies plummeted. Most people who lost their jobs during this period had little hope of finding new ones. The second phase began early in 2009. Layoff rates peaked and declined, but firms still had little interest in hiring. Vacancies continued to shrink in the first half of the year, as companies adjusted to falling demand.

By the middle of last year, a surprising third phase had started. As the economic weather brightened, firms began advertising new vacancies. But despite the large pool of available unemployed workers, hiring did not rebound. The pattern has continued into 2010: 15m Americans remain unemployed, and the jobless rate continues to hover near 10%.

Long slogs out of recession are not unfamiliar to Americans, who have become used to jobless recoveries. Employment did not regain its previous peak for 23 months after the 1991 recession, or for 39 months after that of 2001. But there are signs that this time the delay is of a different nature. After the 2001 recession, hiring began to rise nine months before the number of new vacancies did. Now vacancies are already rising, even though hiring is not. For one reason or another, available jobs are not being filled.

Why aren’t more people being hired? In the first and second phases of recession, the answer clearly lay with demand. Real output shrank by 3.7%, the deepest decline of any postwar recession (a record shared with the downturn of 1957). Facing a collapse in global trade and consumer demand, firms slashed production and cut their payrolls to match. Employers are still cautious and weak demand will continue to limit new hiring. No one expects, say, a repeat of the blazing recovery of Reagan’s re-election year, when GDP rose by 7.2%. The Federal Reserve forecasts that GDP will grow by a bit more than 3% in 2010 and nearly 4% in 2011. The IMF thinks 3% will be a stretch in each of the next two years.

Nonetheless, GDP has been recovering. Were demand all that mattered, employment would be stronger. But its weakness, taken with the divergence of hiring and vacancies, suggests that other explanations are needed. It looks increasingly likely that America’s labour market has developed structural problems that may explain why it is struggling to respond.

One is that the recession hit certain industries, such as manufacturing and construction, especially hard. Jobs lost there will return slowly if at all, and people turned out of factories and building sites may be poorly suited to openings in growing fields such as health care and education. Critics of this theory (including Christina Romer, head of Mr Obama’s Council of Economic Advisers) point out that the rate at which unemployed workers get new jobs has fallen across all industries. But other measures, such as sectoral differences in stock returns, suggest that the downturn’s impacts were historically uneven. Adjustments across industries will be needed, and skills mismatches will slow them down.

So will geographic factors. Ten percentage points separate the states with the highest and lowest unemployment rates. But the property crash is making it much harder for Americans to move to where the jobs are. A quarter of mortgage borrowers owe more than their houses are worth. Many people are stuck in places with poor employment prospects, unable to leave for cities where their skills may be in demand. Although the economy is starting to create new and often highly remunerative jobs, they are out of reach to those who cannot move.

An unprecedented extension of jobless benefits may also be to blame. Congress has voted several times to continue providing emergency unemployment benefits, and in many states an unemployed worker can receive jobless benefits for nearly two years. This could encourage workers to pass up potential job opportunities. (It could also encourage people to register as unemployed, without affecting underlying labour supply.) Most studies suggest that the extensions can be blamed for an increase in the unemployment rate of about a percentage point, although a new report from the Federal Reserve Bank of San Francisco says the effect is much smaller, just 0.4 percentage points.

Complicating matters is the rapid increase in long-term unemployment (ie, lasting 27 weeks or more). The long-term unemployed take longer to find work, because their skills and job connections erode and employers tend to be sceptical of their abilities and work habits. The authors of a new paper published by the Brookings Institution estimate that the rise in long-term unemployment alone could cause labour-market recovery to take twice as long as after the 1982 recession.

The risk is that hiring will be so slow that long-term unemployment becomes permanent, and the jobless recovery becomes sclerosis. So far America is avoiding this fate. The rate at which people exit unemployment is about four times faster than was typical in continental Europe over the past few decades, according to the Brookings paper. (That said, Europe is probably better than it was at getting people back to work.) But the unemployment rate may still be above “normal” by the middle of this decade.

A lasting blight
The long-term effects of all this are hard to predict, especially when the recession ended so recently. But history provides some guide to the unpleasant shifts that lie ahead for many Americans.

Lots of laid-off workers will never regain their former earning power. A study of those who lost stable jobs in 1981-82 found that their earnings fell at once by 30%; they had made up less than half the gap 15-20 years later. New graduates entering the labour force will earn less than they might have done a few years ago, although the effects will fade during their careers. A recent study by Yale’s Lisa Kahn estimates that each percentage-point increase in the national unemployment rate produces a loss in initial wages of 6-7%. New graduates take less prestigious jobs and stay in them longer, stunting upward mobility.

For those aged 16-24 the effects are even greater. Some have chosen to carry on studying, which should lead to increased productivity and pay. Enrolment rates at high schools and colleges have increased by several percentage points since the recession began. But many other young people have become idle, neither in work nor in school, and can expect low pay when they do find a job. Idleness in this age group has risen most among males, and especially black males—to 24%, from just under 18% before the recession. That this trend has not coincided with an increase in crime has surprised many analysts. Rates of violent and property crimes ticked up in the early 1980s and in 1991, but tumbled in 2008 and 2009. Larry Katz, an economist at Harvard, has mused that video games and online diversions may have absorbed the attention of young, idle people.

Older people have less time to make up for lost earnings. Some have put off retirement plans in order to rebuild lost wealth but half as many again, says a recent estimate, have retired early because job prospects are poor. These people face a longer retirement on reduced means, especially because early drawing of Social Security (ie, state pension) benefits means a lower payout. The shift will place extra strain on public finances, because older people will pay less in taxes and collect more in benefits. By reducing payroll-tax revenues, the recession has hastened the arrival of an annual deficit in Social Security, once forecast for 2016 but now expected this year.

Income inequality is likely to widen. Educated people were less likely to lose their jobs during the recession. For workers without a high-school diploma, the unemployment rate peaked at 15.6%; for those with one, it was more than four percentage points lower. A college degree knocked off another six points. The educated may be favoured by future job growth too. That would place upward pressure on wages that are already above average, while the incomes of less skilled workers stagnate. Richer households will also find it easier to adjust to structural obstacles in the labour market. They can more easily afford to send children to university, absorb property losses or move to another city.

The knock-on effects from these shifts are hard to predict. Research indicates that recessions in early adulthood can shape attitudes toward success and government intervention. If these people vote, they may demand more redistributive government policies. They may also prefer economic nationalism to free trade.

Some may see in this the beginnings of a shift in the American economy towards a pattern more familiar to Europeans. High unemployment, especially of the long-term variety, sets in. That calls forth demands for a more generous social-safety net, which in turn exacerbates the structural flaws in the labour market.

Old-world thinking
It is much too early to reach such a gloomy conclusion. America seems to be in little danger of repeating the mistakes Europe made in the 1970s and 1980s in encouraging early retirement and protecting employment. Its politicians have largely resisted calls to follow Europe’s lead in the latest recession, too. Germany’s Kurzarbeit programme shared declines in labour demand across workers by curtailing hours, for example. That has limited unemployment but may contribute to longer-term problems should recovery require more of a structural adjustment.

Nevertheless, America’s jobs machine is sputtering. The economic engineers who work on it should preserve and if possible enhance labour-market flexibility. Hiring subsidies may help mop up cyclical unemployment. Investments in training and (more likely) broader education could begin to address structural problems. But adjustments to global economic shocks do not happen overnight. What is important is that the American government avoids costly mistakes. In facing the recession it heeded the lessons of the 1930s. In the aftermath it should remember those of the 1970s and 1980s.