A time-series analysis of US entrepreneurship: evidence from fractional integration

Abstract

This empirical note employs fractional integration techniques with allowance for seasonality to infer the degree of persistence in US self-employment rate using monthly data from 2000:01 to 2014:05. The results indicate orders of integration in the range of [0.5,1] implying nonstationary, but mean-reverting behaviour.

Keywords:

• entrepreneurship
• self-employment
• fractional integration
• time series

JEL Classification:

• C22
• J23

Notes

¹ For example, see recent research by Faria et al. (2009, 2010), Galindo and Méndez (2014), Gohmann and Fernandez (2014), Yu et al. (2014) and citations therein for the relationship between entrepreneurship with the business cycle and economic growth.

² Preliminary analysis revealed the absence of structural breaks. Results are available upon request.

³ As noted by Congregado et al. (2012, p. 1245), ‘the self-employment rate may represent long-standing industrial and institutional structures within a country rather than the dynamism of an entrepreneurial activity’.

⁴ Incorporated self-employment data from the US Bureau of Labor Statistics are only available from 2000 forwards.

⁵ Robinson (1978) and Granger (1980) justified the use of these models based on the aggregation of heterogeneous autoregressive processes. Gil-Alana and Hualde (2009) present an updated review of fractional integration and its applications in economic time series.

⁶ Though not reported in Panel B of Table 1, the seasonal AR coefficient is 0.237.