Policymakers have become increasingly interested in the role entrepreneurs play in the US and other market economies. This is particularly true of ‘transformational entrepreneurs’ — that is, entrepreneurs whose companies introduce major innovations, create many jobs, and disproportionally contribute to productivity growth. These transformational entrepreneurs are contrasted with ‘lifestyle entrepreneurs.’ Lifestyle entrepreneurs, whose numbers are much larger, choose to operate businesses that generate income for themselves and, perhaps, a few family members (Schoar 2010, Hurst and Pugsley 2012). We highlight recent evidence suggesting that the impact of the high-growth, transformational entrepreneurial firms has declined in recent years, and we suggest possible avenues for research to better understand the causes and consequences of this decline.
The decline in high-growth young firms
A growing literature documents a 30-year decline in various measures of business dynamism and entrepreneurship in the US (e.g. Decker et al. 2014, Hathaway and Litan 2014). That trend can be seen in data on young firm activity as a share of employment, as shown in Figure 1. An important question is: Does the decline in entrepreneurship reflect a decline in ‘transformational’ entrepreneurs or a decline in ‘lifestyle’ entrepreneurs? The answer may have important implications for labour markets and economic growth (see Davis and Haltiwanger 2014). Figure 1 cannot answer this question by itself, but as we shall see, the differences in the patterns across sectors offer clues. In particular, the decline in young firm activity in the 1980s and 1990s is dominated by young firms in the retail trade sector, while in the post-2000 period, young firms in the information sector exhibit a sharp decline.
Figure 1. Share of employment at young firms for selected industries
Source: Figure 4 of Decker et al (2016). Note: Young firms have age less than 5 years. Industries are defined on a consistent NAICS basis. Data include all firms (new entrants, exiters, and continuers). Author calculations from the Longitudinal Business Database.
In recent research, we shed light on this question by studying the behaviour of high-growth young firms over time (Decker et al 2016). While it is difficult to identify entrepreneurs with high growth ambitions prior to start up, by using detailed micro-data on the universe of US firms, we can observe the entire distribution of firm growth rates and study the firms with the highest growth. In Figure 2, we report the 90th percentile of the employment-weighted firm growth rate distribution for young and mature firms as well as the economy overall. The 90th percentile line reflects the growth rate of the firm that is growing faster than the firms that account for 90 percent of all US jobs.
Figure 2. High-growth firms by firm age (90th percentile of employment-weighted distribution), continuing firms
Source: Figure 10 of Decker et al (2016). Note: The 90th percentile is based on the employment-weighted distribution of firm employment growth rates. Data are HP trends using parameter set to 100. Data include continuers only. Author calculations from the Longitudinal Business Database.
Several patterns emerge from the data. First, while Figure 1 shows that entrepreneurial activity declined throughout the 1980-2010 period on an economy-wide basis, Figure 2 shows that this decline was not manifest among high-growth young firms until around 2000. The fastest-growing (90th percentile) young firms grew at a steady rate of just under 70% per year throughout the 1980s and 1990s. After 2000, we see the growth rates of the fastest-growing young firms decline sharply. The decline in high-growth young firms coincides with the decline in young firm activity in key sectors like information, which is a core part of the high tech activity in the economy. We now turn to a more in-depth exploration of the high tech sector.
Differences in patterns of dynamism: High tech vs. retail trade
To help illustrate the differences in dynamism between high tech and other sectors, we compare and contrast the pattern for the high tech and the retail trade sectors. We use the approach of Hecker (2005) who identifies the 4-digit NAICS industry groups with the highest concentration of science, technology, engineering, and mathematics (STEM) workers to define high tech. The identified industry groups draw heavily from the Information sector but also from the information technology industries in the Manufacturing sector and from scientific industries in the services sector. In Figure 3, we report the interdecile range (the difference between the 90th and 10th percentiles) of the employment-weighted growth rate distribution of firms in these two sectors. We use this interdecile range as a measure of growth rate dispersion.
Figure 3. 90-10 Differentials for the retail trade and high tech sectors
Source: Figure 3 of Decker et al (2016). Note: The 90-10 differential is the difference between the 90th and the 10th percentile of the employment-weighted distribution of firm employment growth rates. HP filter uses parameter set to 100. Industries are defined on a consistent NAICS basis. Data include all firms (new entrants, exiters, and continuers). Author calculations from the Longitudinal Business Database.
The retail sector exhibits a steady decline in dispersion over the entire period. In contrast, the high tech sector exhibits increasing dispersion during the 1990s and a sharp decline in dispersion after 2000. Further insights are gained by decomposing the interdecile range into the 90-50 gap (the difference between the 90th and 50th percentiles) and the 50-10 gap (the difference between the 50th and 10th percentiles). The difference between the 90-50 and 50-10 gaps is a measure of skewness, indicating the prevalence of high-growth firms if the difference is positive. In Figure 4, we report 90-50 and 50-10 gaps for the same two sectors.
Figure 4. 90-50 and 50-10 differentials for the retail trade and high tech sectors
Source: Figure 11 of Decker et al (2016). Note: Y axis does not start at zero. Solid lines indicate 90-50 differential; dashed lines indicate 50-10 differential. The 90-50 differential and the 50-10 differential are the difference between the 90th and the 50th percentile and the 50th and 10th percentile, respectively, of the employment-weighted distribution of firm employment growth rates. Data are HP trends using parameter set to 100. Industries are defined on a consistent NAICS basis. Data include all firms (new entrants, continuers, and exiters). Author calculations from the Longitudinal Business Database.
Figure 4 shows that the high tech sector is characterised by a high degree of skewness at the beginning of the time series. The skewness largely persists until around 2000, after which skewness steadily declines until it is almost eliminated. Hence, the decline in dispersion in high tech (seen also in Figure 3) is accompanied by a decline in skewness in the post-2000 period. The story in the Retail Trade sector is starkly different. This sector exhibits almost no skewness throughout the time period, with its dispersion decline reflecting equal reductions in the 90-50 and 50-10 gaps.
In Decker et al. (2014, 2016), we show that positive skewness typically arises because of the presence of high-growth young firms. Moreover, they show that the differing patterns of skewness in retail trade and high tech reflect differences in the contribution of high-growth young firms across these sectors. High tech prior to 2000 had a high and growing share of young firm activity with a small but important fraction of high-growth young firms. Since 2000, the share of young-firm activity and the high-growth activity among young firms has declined in high tech. In contrast, the retail trade sector does not show much skewness because high-growth young firms have played only a small role in it.
The declining dynamism agenda
The results in Decker et al. (2016) provide a number of new basic facts that should discipline future research into the declining dynamism phenomenon. First, the character of the decline in entrepreneurship has changed over time. The pre-2000 period exhibits relatively stable activity for high-growth young firms, whereas the post-2000 period shows a decline in transformational entrepreneurship, at least as measured by employment growth. Second, the patterns of the decline in business dynamism did not affect all sectors of the economy in the same manner as reflected in our comparison of the retail trade and high tech sectors. After 2000, however, all sectors joined the decline and convergence trend in dispersion. Third, prior to 2000, the decline in dispersion reflected a somewhat symmetric tightening of the growth rate distribution. Since 2000, the dispersion decline has been accompanied by a decline in skewness, that is, a decline in high-growth young firm activity. These differential patterns across time and sectors suggest that a single underlying explanation for the 30-year decline in dynamism is unlikely.
These facts provide researchers with clues on where to focus in order to understand the causes and consequences of declining business dynamism. The patterns for retail trade are already better understood than those for high tech. Numerous studies (e.g. Foster et al. 2006, Jarmin et al. 2009) document that the business model in retail trade has changed away from single-unit establishment firms (‘Mom and Pop’ firms) toward large, national chains. These studies show that this change has enhanced productivity while accounting for the declining dynamism in retail trade. Establishments of large, national chains are both more productive and more stable than their Mom and Pop counterparts. This structural change is consistent with the patterns shown in Figures 1, 3, and 4. Put in terms of the discussion above, the changes in retail trade are more consistent with a decline in lifestyle entrepreneurs than transformational entrepreneurs.
The post-2000 decline in high-growth young businesses in key innovative sectors like high tech suggests there has been a decline in transformational entrepreneurs in this sector. Why this decline has occurred is an open question. In the post-2000 period, high tech includes fewer young firms, and the young firms that are present are less likely to have high growth. This period of decline in high-growth entrepreneurship in high tech coincides with the decline in aggregate productivity growth in the high tech sectors of the economy as documented by Fernald (2014). Given the important role high-growth young businesses have played historically in the US, especially in sectors like high tech, understanding the causes and consequences of this decline should be a high priority for future research.
Disclaimer: Any opinions and conclusions expressed herein are those of the authors and do not necessarily represent the views of the US Census Bureau or of the Board of Governors of the Federal Reserve System or its staff. All results have been reviewed to ensure that no confidential information is disclosed.
Davis, S J and J Haltiwanger (2014) “Labor market fluidity and economic performance”, NBER, working paper no 20479, (published in the Federal Reserve Bank of Kansas City Jackson Hole Conference Symposium).
Decker, R A, J Haltiwanger, R S Jarmin and J Miranda (2014) “The role of entrepreneurship in US job creation and economic dynamism”, Journal of Economic Perspectives, 28(3): 3-24.
Decker, R A, J Haltiwanger, R S Jarmin and J Miranda (2016) “Where has all the skewness gone? The decline in high-growth (young) firms in the US”, forthcoming, European Economic Review.
Fernald, J (2014) “Productivity and potential output before, during, and after the Great Recession”, Paper presented at NBER Macro Annual 2014.
Foster, L, J Haltiwanger and C J Krizan (2006) “Market selection, reallocation and restructuring in the US retail trade sector in the 1990s”, The Review of Economics and Statistics, 88(4): 748-58.
Hathaway, I and R Litan (2014) “Declining business dynamism in the United States: A look at states and metro”, Brookings Institution.
Hecker, D (2005) “High-technology employment: A NAICS-based update”, Monthly Labor Review, US Department of Labor and US Bureau of Labor Statistics, 128(7): 58.
Hurst, E and B Pugsley (2012) “What do small businesses do?”, Brookings Papers on Economic Activity, 43(2): 73-142.
Jarmin, R S, S Klimek and J Miranda (2009) “The role of retail chains: National, regional and industry results”, in J Dunne and Roberts (eds), Producer dynamics: New evidence from micro data, University of Chicago Press, Chicago: 236-270.
Schoar, A (2010) “The divide between subsistence and transformational entrepreneurship”, in J Lerner and S Stern (eds), NBER Innovation Policy and the Economy.