EVENTS

Date:	Thursday, July 28, 2005
Time:	2:00 PM -- 4:00 PM
Location:	Wohlstetter Conference Center, Twelfth Floor, AEI 1150 Seventeenth Street, N.W., Washington, D.C. 20036

July 2005

The Fluid Character of U.S. Labor Markets

The U.S. economy creates more than 7 million new jobs every quarter, and millions of workers quit their jobs each month. What do these numbers mean for our understanding of the labor market? What impact does a fluid labor market have on productivity growth? How does the pace of job creation, departures, and layoffs differ by industry and employer size? These and other questions were the subject of a July 28th AEI panel discussion where AEI visiting scholar Steven J. Davis elaborated on the findings of his recent study, "The Flow Approach to Labor Markets."

Kevin A. Hassett
AEI

It used to be the case that economists saw the labor market as the schoolroom where everyone sat at their desks plugging away, but now they are starting to see things more like the playground with the kids wandering from activity to activity. To the extent that our perceptions of the labor markets have changed over time, the members of today’s panel are responsible for that.

Steven J. Davis
AEI

Payroll employment rose by 146,000 jobs in June and 600,000 in the last three months. This is the headline that you see in the news each month. It is often what moves financial markets, and it is the number that was quite prominent in the presidential election. In the BED (Business Employment Dynamics), you can see that the United States creates 8 million new jobs each quarter. JOLTS (Job Opening and Labor Turnover Survey) shows that, in a typical month, there are 4.5 million new hires--not what you see in the payroll data.

The net change in employment essentially represents jobs created minus jobs destroyed, or hires minus separations. The new BLS (Bureau of Labor Statistics) data allows you to separate out how these changes in employment occur. The BED data goes through a sample of employers to find out how the shift in the employment situation.

You can also look from the point of view of workers: Where did they come from? Where did they go? The BLS and census bureau’s new statistics give us a great new window into what the economy is doing. In 1990–2003, 8 percent of jobs were newly created during the last three months. The job destruction number (7.7 percent) is also huge, but the net shows employment growth. Annually, 13–14 percent of jobs disappear in an average year--and that is not even in a recession. This is a tremendously fluid process. 3.2 percent of workers join their employers each month, and 3.1 percent leave each month.

The working-age population is split into those employed, those unemployed (i.e., who do not have a job and would like one), and those out of the labor force (i.e., people who do not have a job and do not want one). In the average month, 8 percent of working age adults are moving between employers. That is an astonishing amount of fluidity and flexibility in the U.S. labor market.

The 1991 recession was characterized by no change in job creation and a spike in job destruction. The 2001 recession showed a drop-off in the rate of creation of new jobs. There is a big difference between the characters of both downturns. This new data is a great tool to understand what is going on out there.

The expansion of Wal-Mart is a story of clustering and gradual outward spreading – not one of the chain moving to the big markets right away. In 2004, there were more than 3,000 Wal-Mart stores, and more than a million employees. On a net basis, Wal-Mart created 1 million new jobs since 1962, but we will break this down into job creation and destruction.

Wal-Mart is the leading example of the transformation of the U.S. retail sector. Wal-Mart expanded into more retail markets and so you can look at what has happened to prices and productivity in these local retail markets.

Emek Basker’s recent research into the economic effect of Wal-Mart’s expansion into local retail markets provides the key information on this. There is an average net employment gain of 100 jobs per store, in the first year, although there is still some job destruction; over the next few years, the net gain falls to 50. There is also a loss of jobs at suppliers, since Wal-Mart is vertically integrated. On the other hand, some retailers benefit from the entry of Wal-Mart, since it might draw in new business. The openings of stores like Wal-Mart, Best Buy, and Borders in recent years have triggered both widespread job creation and destruction.

Wal-Mart transformed the landscape in the late 1990s, driving the retail sector to produce 10 percent annual productivity growth, as its stores displaced less productive competitors. Sears and Target have also sought to emulate Wal-Mart’s pricing and inventory management practices, and so the competitive discipline of Wal-Mart has pushed up productivity elsewhere. The dramatic retail productivity gains of the 1990s can largely be explained by the entry of large national retail chains displacing older independent outlets, as this creative destruction process has been responsible for America’s wider impressive productivity record.

It’s difficult to say exactly how the creation and destruction of jobs in the United States compares with job creation and destruction in other countries because comparative data is hard to find. The United States has more of a churn [what does that mean?] than other countries have, and it is known that the U.S. labor market is more fluid--but it is hard to determine exactly how much more. There is higher productivity growth and churn in the service sector than in manufacturing because the former allows for more creative destruction and market fluidity.

A lot of people have underestimated the dynamism and resilience of U.S. labor markets. Many had not realized the extent to which the U.S. economy has been able to absorb millions of jobs destroyed and generate millions of new ones.

Jared Bernstein
Economic Policy Institute

Why has the rate of job creation stayed so low in this recession and recovery? In a sector like manufacturing, quits are low relative to layoffs; in government, quits and layoffs are both low. It might be interesting to look at the role of Unions in this. One would also like to learn more about the role of wages in these shifts. Do high-productivity industries show different elasticity dynamics than low-productivity industries show?

Job losses in some industries are much more costly than losses in others. You are more likely to lose your job if you are highly educated, but will likely find new work sooner than if you were not. In low-skill industries, displacement means that your earnings are likely to fall dramatically.

The highest rates of job destruction and job creation occur with the fastest net growth and contraction of firms. Most establishments are small, but most people work at large ones--so, large firms dominate the results. If General Motors, Hewlett-Packard, and IBM undergo significant contractions, we will see huge destruction--regardless of what small firms do.

When comparing statistics of employment volatility, the age distribution of the workforce must be controlled for. Increased temporary employment may reduce volatility since it might not show up in the statistics.

So, does this mean there is really less volatility or that the volatility is more concentrated in the temporary worker periphery? The rate of displacement against unemployment indicates that there have been more layoffs in a tighter job market, and increased displacement shows greater volatility. It is firms that are growing that hire, and firms that contract that fire. With the very serious challenge that we face with job creation, we need to make sure that the benefits of growth are more widely shared.

Why is the U.S. job creation machine jammed in a low gear? This is not a concentrated recession due to a demand shock held back. The recession was mild, consumption did not ever really fall and yet employers lack confidence. The economy does not work like tennis players who need a rest after a long match, but bubbles take a lot longer to clear up than we once realized.

The IT industry is still in the doldrums. It has not come back at all. Job creation in IT has also collapsed since the bubble. Once a certain number of establishments start firing on net, there are not enough others to compensate with job creation.

A shift toward just-in-time inventory employment practices may lead to lower rates of job creation, and there is a shift toward less cyclical industries. The problem of manufacturing decline multiplies beyond manufacturers.

Diana Furchtgott-Roth
Hudson Institute

Steve’s analysis shows how useful the BLS and Department of Labor data has been. Several decades ago, it was common for U.S. workers to spend their entire lives with one or two employers. Yet, those born between 1957 and 1964 held an average of 9.6 jobs from the ages 18–36. People are moving more frequently and getting better matches for their skills. According to the seasonally adjusted JOLTS survey, there were 54 million new hires and 51 million total new separations in 2004. This does not show job creation jammed in low gear. 2004 and 2005 were very strong years. Many of the separations were people moving to take better jobs. Some industries experienced greater flows than others. Quit rates indicate worker confidence in finding new jobs.

JOLTS looks at the types of separations, while the BED looks at job gains. The JOLTS data break-down by region also helps us see the differences in growth and employment. The South shows the higher turnover, since it has a higher proportion in leisure and construction, though there are also fewer unions.

From 1992 to 2004, the rates of gains and losses declined, but the same sectors and types of people had the same ranking of turnover. Higher pay does not lead to lower rates of turnover. Turnover is also related to confidence of finding a new job in the industry. Construction enjoyed high turnover due to high levels of opportunities, whereas manufacturing saw declines. Declining employment does not necessarily mean declining quality.

Sectors like construction might work on a project by project basis. It is hard to get comparable data for other countries. The United States has the highest turnover and lowest job stability of the members of the Organization for Economic Co-operation and Development (OECD). In the United Kingdom, the desire to leave jobs was driven by dissatisfaction with current jobs. In the United States, it was the hope of finding newer, better jobs.

Job security provisions and EPL (Employment Protection Legislation) inhibit labor market flexibility. Hiring and firing laws are less stringent in the United States and turnover is higher. Countries with high EPL have high unemployment rates. In the United States, 12 percent of the unemployed will be out of work for only twelve months; in Germany, over 58 percent will. This is serious, as skills decay over time when a worker is out of employment.

Turnover is associated with people seeking better paying and more appropriate jobs. This is good for the economy. Regular turnover means a better and higher job match.

AEI research assistant Chris Pope prepared this summary.