The economy is front and center in the presidential campaign, and candidates will not only disagree on what the facts mean: They'll disagree on the facts. To help make sense of the debate, I asked two former presidential economic advisors — one Democrat, one Republican — to answer a simple question, "Is the jobs market back? Would you say it has now recovered from the financial crisis? What are the indicators you would use to judge?"
This opening piece in the series finds two nationally known labor economists debating the topic. For the Democrats: Princeton economics professor Alan Krueger, chairman of President Barack Obama's Council of Economic Advisors from 2011 to 2013. For the Republicans: Stanford economics professor Ed Lazear, CEA chairman under President George Bush from 2006 to 2009.
This time, Krueger wrote first and Lazear responded. We'll reverse it for the next installment.
To answer your question one needs to have a working definition of "recovered." A good analogy is to a person recovering from a heart attack. We would probably say the patient has recovered if his functioning returns to the same level that it was in the period before the heart attack struck. Can he work a full day without getting tired? Can he walk up a flight of stairs? Can he play tennis as well as he did before the heart attack? Of course, even if he has recovered in this sense, he still would not have made up for the suffering and lost time during the heart attack and subsequent bypass surgery, insertion of a pacemaker and whatever other treatments were required to restore his health. We also wouldn't necessarily say he is in good health — he may have been restored to the same mediocre state of health he exhibited before the heart attack. But his doctor would say that, as far as the heart attack goes, he has recovered to where he was before the heart attack.
This very much describes the position the labor market is currently in — in some respects recovered, in some respects still on the road to recovery, and in some respects in stronger health than it was before and in some respects in weaker health. By most measures — including the unemployment rate, average work week, real average hourly earnings — the labor market is performing about where it was during the last recovery (see chart). The unemployment rate is slightly below where it was in the average month in the last business cycle expansion (2001-07); short-term unemployment is below where it was at the end of the previous recovery and long-term unemployment is above it; the broader U-6 measure, which includes workers who are part time for economic reasons, is still elevated. Adjusted for inflation, average hourly earnings have risen at a moderate pace since the peak of the last recovery. The 2.5 percent rise in nominal hourly wages in the last year suggests that the labor market is getting close to full employment.
The labor force participation rate dropped 3.6 percentage points after the Great Recession struck, and has recovered 0.6 percentage point in the last six months (and then dropped 0.2 percentage point in April, according to last Friday's report). This is obviously not fully recovered, but the reason for at least half of the decline is simply demographics and more baby boomers reaching retirement age. In addition, the labor force participation rate fell in the last recovery, so forces were putting downward pressure on labor force participation before the Great Recession. And the labor force participation rate and employment-to-population rate of men were both on a downward trajectory since the early 1950s, so by the definition set out above they never recovered from any recession, including the boom years of the 1990s.
Problems of sluggish median wage growth and rising inequality, which have plagued the job market since the early 1980s, are also still with us, so in these important respects, the labor market has been restored to the same mediocre health it was in prior to the financial crisis.
If our heart attack patient returned to the same risky behaviors that caused the heart attack to build up in the first place, we would worry that his recovery was transitory. Fortunately, the passage of Dodd-Frank Wall Street Reform Act and more vigilant financial regulatory enforcement is likely to reduce the odds of another financial crisis in the near term.
Alan and I are in agreement for the most part, but I am somewhat more negative in my reading of the data. I discount the unemployment rate data heavily and focus more on the employment rate and other indexes, which paint a much less positive picture, even when correcting for demographics. As a result, I conclude that although we have made significant progress, the labor market has not yet recovered.
For the most part, I look at the same data as those presented in Alan's graph, but I think it is useful to look at a few more series and to note the entire time series rather than pre- and post- averages.
Like Alan, I start with the most important fact, which is that the labor market has recovered much of the way from where it was during the depths of the recession. By any measure, the labor market is in much better shape than it was six years ago. Employment is up, unemployment is down, hiring is up and the ratio of quits-relative-to-layoffs is up from where it was during early 2010. We may disagree on whether we are all the way back, but we surely agree that much of the road home has been traveled. I believe that is the main point made by Alan's graph.
The indicator that most argues in favor of full recovery is the unemployment rate. Not only is 5 percent equal to the average during the 2000-07 period, but it is also close to what might be the typical economist's estimate of the unemployment rate at full potential output. (see chart below). By this measure, the labor market is completely back to normal.
I believe that we are not fully recovered for a number of reasons. Most of the series that I track suggest otherwise and the most important indicators make clear to me that the labor market is still not back to normal.
The most important summary statistic is the employment rate, which is the ratio of the number working to the number in the working-age population. Most labor economists prefer it to all other indicators because it cuts through the issue of discouraged workers and those who are slow to enter the labor force. The employment rate peaked at almost 63.5 percent in late 2006 and early 2007. It fell to a low of 58.2 percent in mid-2011 and is now back up to about 60 percent. Part of the difference between the previous peak and the current rate can be explained by demographics. As a larger proportion of the population is in what would normally be the retirement years, that rate will fall. As Alan points out, perhaps half (he says at least half) can be explained by this factor. Many have noted that the labor-force participation rate has fallen (see graph) and a portion of that fall is clearly attributable to demographics. But demographics is not the whole story, which is best seen by examining the employment rate of those in the 25-54 year old group.
Employment rates for this group remain too low. If anything, changing demographics should imply higher rates of employment for these workers than in the past because as older workers retire, the demand for the prime-age workers' labor rises, which should induce more of them to work. The opposite has occurred for prime-age workers. Their employment rates are still substantially below the prerecession peak and even the early 2000s recession trough. It is clear that the labor market is not back for these individuals.
A positive indicator, but one that suggests we are short of fully recovered is job creation, which is occurring at the healthy pace of slightly over 200,000 per month, even after Friday's report. Ironically, this pace is too fast to be consistent with full recovery. When we are in recovery equilibrium, job growth should just keep up with population growth. To keep pace with a growing population, about 140,000 per month are required. The fact that it is higher than that means that we are still gaining jobs on net, which means we are still recovering, not fully recovered. When a patient has completely recovered from a disease, his health is stable, not improving. When the labor market is improving rather than stable, it is not fully back to normal.
There is significant additional corroborating evidence that labor market recovery is not complete. I am fond of the Bureau of Labor Statistics' (BLS) JOLTS data (Job Openings and Labor Turnover Survey). JOLTS reports the number of hires and separations each month. The number of hires far outstrips net job creation because hires and separations tend to move together. But hires are a good indicator of the robustness of the labor force. At the last peak in late 2006, the hiring rate, defined as the numbers of hires relative to employment, was at 4 percent per month (amazingly high). Today, that rate is 3.7 percent. But even this may overstate the recovery because the hiring rate looks at hires relative to those employed, not relative to the working age population. If labor-force participation and employment are too low for some reason, the hiring rate might look good because the denominator, employment, is too low. Just as the employment rate considers all those who could possibly work, so too could a hiring rate consider all those who could possibly be hired.
Another measure is the number of hires relative to the working-age population. It peaked at 2.4 percent in late 2006 and now stands at 2.1 percent. This arguably better measure tells the same tale, but with a bit more force. Hiring is much better than it was at the depths of the recession, but not back to full-recovery levels. Once again, the current rate is more like the rate at the trough of the early 2000s recession than it is like the peak of that recovery.
That wages have not grown is no surprise. A necessary condition for wage growth in the U.S. has been productivity growth. Except for a brief period at the end of the recession, productivity growth has been low. I believe that the weak recovery is best explained by low investment. Incentives for capital have not been strong. Increased taxes on capital, additional regulation, government suits and fines against companies especially in the financial sector have not encouraged capital formation. These factors, coupled with large increases in the debt, are not conducive to strong growth in productivity and as a result, place a cap on wage growth. I am not confident that Dodd-Frank has done much to make the economy more secure. Instead, I believe it treated the symptom rather than the cause.As I have argued elsewhere, the financial crisis was brought about by systemic factors, not by the failure of a few key firms. Dodd-Frank focuses too heavily on the latter rather than the former.
Finally, one can ask whether the lingering deficiencies are structural or cyclical. In 2012, James Spletzer and I argued that most of the lagging employment was a cyclical rather than a structural problem. However, if one includes those aspects of policy mentioned above as part of the structure, then it is possible that we will move to a regime of lower employment, lower productivity growth and lower wage growth that is more in line with countries that have adopted similar growth-impeding policies.