1 BUSINESS DYNAMICS STATISTICS BRIEFING: Where Have All the Young Firms Gone? May 2012 Sixth in a series of reports using data from the U.S. Census Bureau’s Business Dynamics Statistics John Haltiwanger University of Maryland Ron Jarmin U.S. Bureau of the Census Javier Miranda U.S. Bureau of the Census
2 Business Dynamics Statistics Briefing: Where Have All the Young Firms Gone? About the Business Dynamics Statistics The Business Dynamics Statistics (BDS) is a product of the U.S. Census Bureau that measures business openings and closings, startups, job creation, and job destruction by firm size, age, industrial sector, and state. The U.S. economy is comprised of more than six million establishments with paid employees. The population of these businesses is constantly churning—some businesses grow, others decline, and yet others close. New businesses constantly replenish this pool. The BDS monitors this activity to provide a picture of the dynamics underlying aggregate net employment growth. More information about the BDS can be found at http://www.census.gov/ces/dataproducts/bds/index.html. ©2012 by the Ewing Marion Kauffman Foundation. All rights reserved.
3 Business Dynamics Statistics Briefing: [ 1 ] Where Have All the Young Firms Gone? Where Have All the Young Firms Gone? 1 by John Haltiwanger, Ron Jarmin, and Javier Miranda 1. University of Maryland and NBER, U.S. Census Bureau, and U.S. Census Bureau. The analysis in this paper uses only public domain data from the Business Dynamics Statistics (BDS). We thank E.J. Reedy for providing useful comments on an earlier draft.
4 Business Dynamics Statistics Briefing: [ ] 2 Where Have All the Young Firms Gone? 4 same period from March 2009 to March 2010, The Census Bureau’s Business Dynamics Statistics the 2 net job creation from all U.S. private sector firms was (BDS) provides data on business dynamics for U.S. 3 -1.8 million jobs. Without the contribution of business firms and establishments with paid employees. This startups, the net employment loss would have been briefing highlights some key features of the most substantially greater. recent BDS update, which now has data through 2010. As the most complete public-use dataset Previous work using the BDS has highlighted the allowing for the analysis of business dynamics in the critical contribution of startups to job creation (see, United States, the BDS is a key source of knowledge e.g., Haltiwanger, Jarmin, and Miranda [forthcoming]). about the changing state, as well as the national, However, a potentially troubling trend identified economy. from earlier BDS releases is that the pace of business startups has exhibited a long-run decline that dates The new BDS data release shows that, in 2010, back to the 1980s (see, e.g., Haltiwanger, Jarmin, 394,000 startups created 2.3 million jobs (these 5 and Miranda  and Litan and Reedy ). were not simply establishment openings but new firms whose establishments also were new to the The newly released BDS shows that this trend has economy). This reflects substantial job creation in a continued through 2010. Figure 1 shows the long time of anemic overall economic activity. Over the decline in the pace of overall job creation in the United Figure 1 Annual Job Creation in U.S. Private Sector– Overall and from Startups 10 20 19 9 18 8 17 7 16 6 5 15 4 14 Job Creation Job Creation Startups 3 13 (Percent of Employment) (Percent of Employment) 12 2 11 1 10 0 1981 1991 2001 1988 1983 1989 1982 1984 1986 1987 1996 1997 1992 1999 1998 1993 1994 1990 1980 1985 1995 2004 2006 2007 2008 2002 2003 2010 2009 2005 2000 Job Creation Job Creation Startups (Right Axis) 2. The BDS was developed at the Census Bureau’s Center for Economic Studies, with support from the Census Bureau and the Ewing Marion Kauffman Foundation. The current update also received support from the Small Business Administration. Statistics on business dynamics are provided at an economy-wide level and by firm size, firm age, sector, and state. Starting early in 2012, the BDS is released annually. For the first time, business dynamics also are provided by establishment size and establishment age. 3. The BDS does not include non-employer firms and, as such, this brief does not speak to job creation from non-employer businesses. 4. In the BDS, net and gross flows are measured from March to March. The net growth rate of employment from March 2008 to March 2009 in the U.S. private sector was -4.9 percent, and was -1.6 percent from March 2009 to March 2010. 5. Another recent study that uses the BDS to explore the role of startups for job creation is Strangler and Kedrosky (2010).
5 Business Dynamics Statistics Briefing: [ ] 3 Where Have All the Young Firms Gone? States accompanied by a declining rate of job creation for by young firms (defined here as firms aged five or less) has steadily declined over time. In addition, from business startups. It is evident that the decline consistent with the patterns for startups, there has in job creation from startups accelerated in the recent been acceleration in the decline of activity accounted recession. for by young firms in recent years. Figure 3 shows the The declining job creation from business startups percent of young firms, their employment, and their reflects a declining firm startup rate. Figure 2 shows job creation as a share of the total number of firms, that the startup rate of firms has declined from as economy-wide employment, and overall job creation. high as 12 percent to 13 percent (as a percentage of From the early 1980s, the share of young firms has all firms) in the 1980s to 7 percent or 8 percent in declined from close to 50 percent to less than 35 recent years. Figure 2 also shows that there has been percent in 2010. This decline reflects both a long-run 6 no discernible trend in the average size of a new firm. secular decline and accelerated decline in the recent Figure 2 shows that the average firm startup is well recession. Figure 3 also shows an accompanying under ten workers. decline in the share of employment and share of job A consequence of this drop in the pace of creation accounted for by young firms. The share of job creation from young firms has fallen from above business startups is that the share of activity accounted Figure 2 Declining Pace of Firm Startups, U.S. Private Sector, BDS 14 8 13 7 12 6 11 5 10 4 9 3 Number of Workers 8 2 Startup Rate (Percent of All Firms) 7 1 6 0 1991 2001 1987 1983 1984 1986 1988 1989 1998 1999 1994 1997 1992 1993 1996 1981 1990 1980 1985 1982 1995 2006 2002 2003 2004 2008 2007 2010 2009 2005 2000 Firm Startup Rate Average Size of Startups (Right Axis) 6. This pattern might appear to contrast with recent findings from Reedy and Litan (2011), who emphasize the apparent smaller declining size of new establishments over time, but we note that our focus here is on new firms and not new establishments. There is a high pace of establishment openings from existing firms. For example, Haltiwanger, Jarmin, and Miranda (forthcoming) show that job creation from new establishments from existing firms is about as large as job creation from new firms. So distinguishing between firms and establishments is critical for both conceptual and measurement reasons. Note that Reedy and Litan (2011) also show that there is a discrepancy between Census and BLS evidence on the time series patterns of the average size of new establishments. BLS evidence shows a more evident pattern of decline in the average size of new establishments. These discrepancies in establishment patterns are an open area of research. Also, note that Reedy and Litan (2011) also show what we show in Figure 2—no discernible trend in the size of new firms from the BDS (although this is not their emphasis).
6 Business Dynamics Statistics Briefing: [ 4 ] Where Have All the Young Firms Gone? Figure 3 Declining Share of Activity from Young Firms (Firm Age Five or less) U.S. Private Sector, BDS 60 40 55 35 50 30 45 25 40 20 Percent Percent 35 15 30 10 25 5 0 20 1991 2001 1989 1983 1984 1986 1987 1988 1999 1998 1997 1996 1994 1993 1992 1981 1980 1985 1990 1982 1995 2006 2002 2003 2004 2007 2008 2010 2005 2009 2000 Share of Firms that are Young Share of Job Creation from Young Firms Share of Employment from Young Firms (Right Axis) 40 percent in the 1980s to around 30 percent in Addressing these questions is beyond the scope of this brief. However, it is useful to examine how recent years. The share of employment accounted for widespread these changes are across regions and by young firms has fallen from more than 20 percent states. Narrowing in on the decline from the national in the 1980s to as low as 12 percent in 2010. to the state-level data, which also are available in the Examining the contributions of young firms to BDS through 2010, Figure 4 shows that the declines business activity and job creation has been highlighted have been ubiquitous, but with varying magnitude of in recent studies that show the disproportionate decline across states. Comparing the changes from contribution of both startups and high-growth young the 1987–1989 to 2004–2006 business cycle peaks, firms to job creation (see Haltiwanger, Jarmin, and the biggest declines at the state level ranged from Miranda [forthcoming]). Figures 1 through 3 highlight a decline of 7 percentage points to 14 percentage that the business activity and job creation from points, while the smallest declines were a more startups and young firms has declined substantially modest 2 percentage points to 5 percentage points. from the 1980s through 2010. These patterns While there is no distinct regional pattern, there is raise a variety of questions about the underlying a common pattern to those states with the largest forces shaping these changes, but, more generally, declines—they are the states that had some of the raise questions about whether the United States is highest initial shares of business activity accounted becoming less entrepreneurial given the lower pace of for by young businesses in the 1980s. That is, there startups and the smaller share of activity accounted for is a strong inverse correlation (about -0.55) between the initial share of activity accounted for by young by young firms.
7 Business Dynamics Statistics Briefing: [ ] 5 Where Have All the Young Firms Gone? Figure 4 Change in the Percent of Employment Accounted for by Young Firms by State: Peak (1987–1994/2004–2006) Biggest decline to smallest decline (percent) -11.8 -14.3– -11.8– -7.1 -7.1– -5.8 Source: U.S. Census Bureau Business Dynamics Statistics -5.8– -4.5 Note: Cutoff points are determined by the quintiles of the distribution. -4.5– -2.4 businesses and the secular decline. In that respect, in young firms declined from 14.4 to 12.0 percent there has been a marked pattern of convergence in (a decline of 2.4 percentage points). Using the state- the activity accounted for by young businesses across level data, Figure 5 shows that the cyclical declines in states. But the convergence has taken the form of young business activity have been shared by all states. all states becoming less entrepreneurial (in the sense However, in terms of this latest cyclical decline, there of a smaller share of activity from young firms) with are discernible regional patterns. The largest declines the most entrepreneurial states exhibiting the largest are concentrated in the West, Southwest, and South. declines. These are all regions hit especially hard during the recession, so, in that respect, these regional patterns Looking more closely at the 2006 to 2010 period, are not surprising. The correlation between the change Figure 2 shows that the decline in the pace of business in the share of employment accounted for by young startups at the national level accelerated in the firms and the change in the net growth rate of the recession and has not yet recovered. Figure 3 shows state over the 2006 to 2010 period is 0.63. That is, the that, at the national level, the share of employment
8 Business Dynamics Statistics Briefing: [ 6 ] Where Have All the Young Firms Gone? Figure 5 Change in the Percent of Employment Accounted for by Young Firms by State: 2006–2010 Biggest decline to smallest decline (percent) -11.8 -14.3– -11.8– -7.1 -5.8 -7.1– -4.5 -5.8– Source: U.S. Census Bureau Business Dynamics Statistics -4.5– -2.4 Note: Cutoff points are determined by the quintiles of the distribution. [forthcoming] and Haltiwanger ). Much evidence states with the greatest net contractions in the Great Recession also exhibited the largest decline in the share shows this dynamism has contributed substantially to of activity accounted for by young businesses. U.S. productivity and job growth. Historically, startups have contributed substantially to job creation by It is an open question why the United States has themselves. While many young businesses fail, the experienced the secular and cyclical decline in the young businesses that have survived have exhibited activity and job creation from startups and young high average growth rates (with the latter dominated firms. Without knowing the sources of these declines, by the most rapidly growing young businesses). In it is, of course, impossible to draw inferences about addition, the evidence shows that the high-growth the implications of these declines for U.S. economic surviving young firms have contributed substantially performance. However, we note that a hallmark of the to productivity growth. While we don’t yet know the U.S. economy often cited by analysts and policymakers reasons, the evidence presented here on the secular is the high pace of business dynamism. A core aspect and cyclical declines implies that, if nothing else, of that dynamism has been a robust pace of business the contribution of such entrepreneurial activity has startups accompanied by a rich “up or out” dynamic of declined. U.S. businesses (see Haltiwanger, Jarmin, and Miranda
9 Business Dynamics Statistics Briefing: [ 7 ] Where Have All the Young Firms Gone? References Haltiwanger, John, Ron Jarmin, and Javier Miranda (forthcoming), “Who Creates Jobs? Small vs. Large vs. Young,” Review of Economics and Statistics . Haltiwanger, John, Ron Jarmin, and Javier Miranda (2011), “Historically Large Declines in Job Creation from Startup and Existing Firms in the 2008–09 Recession,” Kauffman Foundation, http://www.kauffman.org/uploadedFiles/ bds_report_3-22-11.pdf. Haltiwanger, John (2012), “Job Creation and Firm Dynamics in the U.S.,” Innovation Policy and the Economy . NBER/Chicago Press. Reedy, E.J., and Robert Litan (2011), ”Starting Smaller; Staying Smaller: America’s Slow Leak in Job Creation,” Kauffman Foundation, http://www.kauffman.org/ uploadedFiles/job_leaks_starting_smaller_study.pdf. Stangler, Dane, and Paul Kedrosky (2010), “Neutralism and Entrepreneurship: The Structural Dynamics of Startups, Young Firms, and Job Creation,” Kauffman Foundation, http://www.kauffman.org/uploadedFiles/firm-formation- neutralism.pdf.
10 Business Dynamics Statistics Briefing: [ 8 ] Where Have All the Young Firms Gone?
11 Business Dynamics Statistics Briefing: Where Have All the Young Firms Gone?
12 4801 ROCKHILL ROAD KANSAS CITY, MISSOURI 64110 816-932-1000 www.kauffman.org