Forecasting unemployment: a small business-survey model; Small business plans and expectations provide accurate and timely forecasts of national employment

Business Economics, Oct, 2004 by William C. Dunkelberg, Jonathan A. Scott, William J. Dennis, Jr.

The labor force grew by 4.1 percent (1.4 percent per year) from 1998 to 2000 but only 2.2 percent (0.7 percent per year) from 2001 to 2003. Absent this unusually slow post-recession growth in the labor force, the full model suggests that unemployment rates in the 2001 recession could have rivaled those experienced in the 1990-91 recession. The exuberance of the late 1990s encouraged labor force participation as the Internet-dot.com craze drove hiring and spending to record levels. It was common for a company that had just completed an initial public offering to rent 50,000 square feet of space and hire workers before it had any sales. But the rush to hire pervaded the entire economy, not just technology sectors. Plans to expand employment (net of plans to reduce employment) were also pervasive among small firms, reaching a quarterly record 19 percent (net of those planning reductions--Figure 2) in 2000. (12)

The build-up of accumulated labor was substantial going into 2000, creating an excess stock of labor by the second half of 2000, when GDP growth slowed substantially. Figure 3 shows the difference between the actual employment level and the predicted level of employment based on GDP levels. (13) Many firms employed labor (with IPO capital etc.) but never produced output prior to 2001. Income, including capital gains, grew faster than GDP estimates based on measured output. (14) Firms planned to create more jobs than they were able to. The NFIB job creation plans rose to a record 19 percent of all firms, but much of this planned employment energy was unsuccessful at hiring new employees. Instead, the percent of small business firms with unfilled job openings rose to a record 35 percent of all firms (Figure 2), with the unemployment rate reaching 3.9 percent and the fraction of the total labor force with a job reaching a record 64.6 percent of the population age 16 and over.

In comparison, the peak percent of firms planning to create new jobs in the 1980s was 6 points lower at 13 percent and the peak percent of owners reporting hard-to-fill job openings was 25 percent, ten points lower than the most recent period. Had employers been successful in hiring enough employees in the late 1990s to keep the job openings statistic nearer the historical average, excess employment would have been even larger as the economy moved into recession in 2000. This disequilibrium is a major contributor to the sluggish recovery of employment since the 2001 recession ended as, according to the employment model, there were about 2.5 million unneeded workers employed in 2000.

Having hired workers faster than GDP growth required through mid-2000, the onset of the recession triggered a prolonged period of adjustments (similar to that observed in the early 1990s, but not as large) to employment and to the labor force (the latter having expanded at abnormal rates--probably in response to strong growth in compensation). (15) Some of this excess was absorbed by modest economic growth after the end of the recession. Also, unusually slow labor force growth as workers left the labor force absorbed a large part of the excess labor inventory. Finally, productivity estimates were likely biased upward, as firms shed employees while the reduction in output during the recession was modest and economic growth resumed in the fourth quarter of 2001.