Perspectives

Also see our Economic Commentary which supplements the Laufenberg Economic Quarterly and primarily focuses on more recent economic developments.

This page is designed to offer you my perspective on economic fundamentals, ranging from business cycles to yield curves. It will provide more detailed economic analysis than is generally available in the Quarterly. For the most part, the essays provided here will attempt to discuss timely fundamental issues related to the forecast but are expected to have a longer shelf life than the content on either the Quarterly or Commentary pages. I hope that over time you will consider the information on this page as a source of reference when debating economic issues in the future.

Daniel E. Laufenberg, Ph.D.
LaufenbergQuarterly.com

bar

Is the great American job machine finally broken?
(December 6, 2011)

The U.S. economy was a remarkable “job-creating machine” in the 1990s, producing a whopping 24 million new payroll jobs during the decade long expansion of that period and pushing the unemployment rate below 4.0 percent for the first time since the late 1960s. What made this performance even more remarkable was the slow start to job growth early in the expansion. A similar scenario unfolded during the economic expansion from 2001 to 2007; that is, the labor market started slow again before improving more dramatically down the stretch. However, job growth during the 2001-2007 expansion was not nearly as stellar as the previous one, as businesses increased their payrolls by only 7.3 million and the unemployment rate bottomed at 4.5 percent. Of course, with the unemployment rate currently at 8.6 percent and 13 million people unemployed, 7.3 million new jobs and an unemployment rate of 4.5 percent would be very welcomed.

So what is going on in the labor market today? Will the U.S. economy become a job-creating machine again or have structural changes in the labor market made it impossible for the jobs machine to restart this time? My conclusion is that it is too soon to declare the U.S. jobs machine permanently broken. As shown in Chart 1, the stickiness downward in the unemployment rate in the current recovery is not unusual when compared to each of the two previous recoveries. The difference is that this time the unemployment rate was much higher when the recovery began than in either of the two earlier periods. This slow start to the recovery in the labor market seems to be the outcome of several factors limiting jobs, both cyclical and secular in nature.

chart 1

First, I would like to discuss the cyclical factors that most likely have contributed to the sluggish recovery in the labor market. Included among these factors frequently mentioned by market participants are the lack of new hires, massive layoffs, heavy reliance on part-time employment, and an army of discouraged workers. Although businesses are justifiably cautious, the reluctance of business managers to add to their payrolls seems to be overstated. As shown in chart 2, gross hires so far during the current economic recovery actually have improved much sooner than they did during the previous recovery (the data series only goes back to 2001). Following the recession of 2001, gross hires did not bottom until 16 months into the recovery, whereas gross hires bottomed 3 months after the end of the more recent recession.

A major difference this time is that gross hires started from a much lower level than they did in the prior recovery. Indeed, it was the deadly combination of plunging hires and aggressive separations during the recession that drove the unemployment rate higher. However, the two series have behaved about as expected since the recession ended. The problem is that the gradual improvement in gross hires started from such a low level. This low level of gross hires may also help explain why the improvement in the unemployment rate during the current recovery has been disappointingly slow. In other words, gross hires have tracked rather closely with separations, suggesting that most of the jobs have been replacement jobs resulting in only a small net gain in employment but a gain nevertheless.

chart 2

A note of caution about the analysis above is warranted. In particular, the data series are relatively new and therefore do not provide very many observations to consider. Although the series provides nearly eleven years of monthly data, it includes only one complete business cycle (2001-2007) and only two recessions (the extremely mild recession of 2001 and the extremely severe recession of 2007-09). A statistician would complain that there are insufficient degrees of freedom to test any hypothesis about cyclical or structural differences.

As shown in the dashed-green line in Chart 3, most of the delay in the decline in separations in early 2009 was due to a surge in layoffs over this period. Again, the data series only goes back to 2001, so the only comparison is very mild 2001 recession. Also shown in Chart 3 (the solid blue line), the number of people who quit their jobs during the last recession slowed dramatically (which may not be unusual in the late stage of a severe recession), but not enough to offset the spike in layoffs (which also may not be unusual). Indeed, since early 2010, the levels for quits and layoffs have been similar, which was far different than what happened in the aftermath of the 2001 recession when the gap between quits and layoffs narrowed but never turned negative like it did in late 2008 and for all of 2009. In fact, one could argue that separations have remained low because quits are still relatively low and not because layoffs are unusually high.

More recently, quits have started to rise again, suggesting that employees are starting to feel better about the labor market than they did, but still not anywhere near as optimistic as they were prior to either of the last two recessions.

chart 3

Another cyclical item that has been mentioned by some analysts as a reason for concern about the labor market is the increased reliance on part-time workers. The concern is that people looking for full-time work have to settle for a part-time job, which is more a case of underemployment than unemployment. After all, an employed person, regardless of whether they work full-time or part-time, are considered employed when calculating the unemployment rate. Nevertheless, part-time work for economic reasons as a percent of total employment increased sharply in 2009 when compared to what happened to part-time employment for economic reasons during the 2001 recession. The good news about this data series is that it has more history to review than hires, separations, quits and layoffs, which allows us to compare the current employment situation with more than just the recovery from a considerably mild recession.

For example, as shown in Chart 4, part-time employment for economic reasons as a percent of all jobs (the dashed-green line) actually was slightly higher during the recession of 1981-83 than it was during the recession of 2007-09. The difference is that it retreated much quicker during the recovery in 1983-84 than it has so far in the current recovery. The implication is that a retreat is probably in order, but it is just a question of when. As the economy continues to improve in 2012, I expect this percentage to decline to about 4.0 percent.

chart 4

Moreover, a mildly interesting aspect of the employment story is that working part-time for other than economic reasons as a percent of total employment (the solid line in Chart 4) has actually slid lower over the last decade or so. In other words, part-time employment as a choice is slightly less widespread than it was earlier in the decade but still quite high by historical standards. More importantly, as mentioned above, part-time work for economic reasons probably will retreat over the next few years due to an improving economy and further gains in gross hiring, whereas part-time employment for non-economic reasons most likely will increase roughly in line with overall employment. As a result, part-time jobs as a percent of total employment should remain at about 13 percent over the forecast horizon.

Another apparent cyclical feature of the employment data is the surge in the number of discouraged workers from late 2008 to late 2010, as shown in Chart 5. The concern of many analysts is that despite the improvement in real output since mid-2009, the level of discouraged workers remains stubbornly high by historical standards. Of course, the historical data for this series is very limited as well, starting only in 1994. This means that the extremely mild recession of 2001 was the only recession available for comparison with the very severe 2007-09 recession. It may be that discouraged workers were just as prevalent in the severe recessions of 1973-75 and 1981-83, but we do not have data to compare. Nevertheless, even if workers were discouraged in roughly the same magnitude then as now, I doubt that the level of discouraged workers remained as high as it has this time. In that sense, the current situation may be different. The question is will these workers be encouraged again to return to the labor force. In most cases, probably not without an upgrade to their skill set. And that is difficult to do without a job to learn, assuming they know how to learn. In other words, the army of discouraged workers may remain high for a considerable time; that is, they may represent a structural change to the labor force more so than a cyclical phenomenon.

chart 5

Second, I contend that there are other secular changes underway that may also contribute to the slow recovery in the labor market. After all, the U.S. jobs machine must operate within the limits dictated by demographics. And the demographics favor solid gains in gross jobs without much improvement in net new jobs. Gross hires are likely to continue to improve but separations are likely to move higher as well. This is not unusual given that as the job market improves, an increasing number of workers are more willing to change jobs than when times are tough. The difference this time will be that separations will be exacerbated by retirements, and it should start in earnest in December of this year.

As shown in Chart 6, other separations, which include retirements, are available only on a not-seasonally adjusted basis but it is quite clear that there is a substantial seasonal element to the data. In particular, other separations tend to spike at the end of the year, with a secondary spike mid-year. These are most likely to two times of year that most people retire. Also, the trend in other separations from 2001 to the end of the last recession has been downward as shown by the first trendline in Chart 6. Since then, however, the trend has been upward, albeit from a very low level. For that reason, I expect a sharp jump in other separations this month (December), reflecting the aging population combined with an improved economy. Of course there are a couple of factors that may suppress retirements, including very low interest rates and a struggling housing market. Nevertheless, the demographic effect will be the overwhelming factor, causing the upward trend in other separations to continue over the forecast horizon.

The demographic effect matters. One aspect of the current labor market that has surprised many analysts has been the steady decline in the overall labor force participation rate in recent years (see Chart 7). Although economists often discuss a cyclical influence on the participation rate, it tends to be dominated by other factors. As shown in Chart 7, the participation rate for men has been in a steady decline on average since World War II, whereas the participation rate for women increased steadily until the mid-1990s. Since then, the participation rate for women has joined the downward trend in the participation rate for men. The explanations for this decline in the participation rate also have importantly implications for the unemployment rate going forward.

chart 7

In this case, I have been struggling for about two months trying to come up with a way to demonstrate this idea in a more understandable way. The result of that struggle is Table 1. In this table, I break the participation rate down into various age groups according to their participation rate for that age group as well as the relative share of that age group in the overall population. This is an attempt to quantify the contribution to the overall participation rate from the aging of the population. The labor force participation rate was 64.2 percent in October 2011, down sharply from 66.8 percent in October 2001.

Let me start by stating my conclusion that the aging of the population alone is estimated to have detracted about 1.4 percentage points of the 2.6 percentage-point drop in the labor force participation rate from 2001 to 2011. To calculate this estimate, I held the participation rates constant over the two periods and calculated what would have happened to the overall labor force participation rate if the population among various age groups was the only thing that changed over the last decade. The result is that the overall participation rate would have fallen to 65.4 percent in 2011 rather than to 64.2 percent. This makes sense given that the participation rate for the 55 years and older population, which is the fastest growing segment of the population, is lower than it is for the young adult segment. And this is still true despite recent increases in the former and declines in the latter.

The implication is that the aging population explains a big part of the decline in the participation rate but not all of it. Indeed, one aspect of the data in Table 1 that jumps out is that the labor force participation rates for younger cohorts actually fell over the last decade, whereas participation rates for older workers increased. For the most part, there may be an explanation, including historically high unemployment rates for the younger cohorts, the apparent end to the upward trend in the participation rate for women around 2001, the hit to household balance sheets during the last recession forcing older cohorts to delay retirement, and the relatively lower unemployment rate for women.

For example, the labor force participation rate for teenagers fell to 34.6 percent in October 2011 from 49.3 percent in October 2001, whereas the participation rate for persons aged 65-69 years increased to 32.3 percent in October 2011 from 25.4 percent in 2001. In the first case, the unemployment rate among teenagers was 23.7 percent in November 2011, down from 24.1 percent in October but still very high by historical standards. As a result, there may be very little incentive for teens to look for work (to participate in the labor force) if there are no jobs available to them. In the second case, the age at which older workers receive full retirement benefits has increased from 65 years of age in 2001 to 66 years of age in 2011. Hence, given the hit to household balance sheets during the last recession (home equity in particular), it should be no surprise that more seniors are working until full retirement age rather than opting to take early Social Security benefits. As a result, there are incentives for seniors to delay retirement (to participate in the labor force).

Finally, it is important to note that the move into full retirement age for baby boomers is just beginning. I know. I am among the first boomers to reach age 66 (next month), my full retirement age with regard to Social Security benefits. The implications for the labor market are important. In particular, the number of net new jobs needed to push the unemployment rate lower is far fewer than most analysts estimate. I contend that the unemployment rate is on the verge of a sharp decline, even though net new payroll jobs will be disappointingly weak by historical standards. The great American job machine is not broken, it is just not as young as it use to be.

bar

red tri

For the current economic forecast, as well as other analysis and commentary, please visit the Stonebridge Capital Advisors website.

Dlaufenberg@stonebridgecap.com

red triGo Here for more information about Dan Laufenberg

The views expressed here reflect the views of Daniel Laufenberg as of the date referenced. These views may change as economic fundamentals and market conditions change. This commentary is provided as a general source of information only and is not intended to provide investment advice for individual investor circumstances. Past performance does not guarantee future results.


Current Perspective

2011 Economic Perspectives

Is the great American job machine finally broken?
-- December 6, 2011

Recession fear, not fact
-- September 4, 2011

Core inflation: a policy guide more than a policy target
-- May, 2011

Help not wanted?
-- Feb, 2011

Special Report: Health Care Legislation
Essay No.1 - Costs
Essay No.2 - Revenue
Essay No.3 - Budget Deficits

2010 Economic Perspectives

Causes of the financial crisis revisited
--November, 2010

A less robust U.S. Economy longer term
--August, 2010

Greece: A test of Europe's resolve to remain united
--May, 2010

The U.S. jobs machine restarting at a slower pace
--February, 2010

~ ~ ~ ~ ~

HOMEABOUTCOMMENTARY FORECASTspacerPERSPECTIVESLINKS