Comparative well-being of the self-employed and paid employees in the USA

“How Millennials Want to Work and Live”, Gallup Report, 2016.; Anna Bahney. “What Millennials Really Want at Work”, CNNMoney, https://​money.​cnn.​com/​2017/​12/​29/​pf/​millennials-work/​index.​html.

Reagan, Brad. “What the Silver Screen Tells Us About Entrepreneurship.” Wall Street Journal, December 2, 2013.

Author calculations using the subsample of the respondents from the USA (N = 3001) included in the publicly available Flash Eurobarometer 354 survey, which was carried out in a total of 40 countries in 2012. The survey publication is available here: http://​ec.​europa.​eu/​commfrontoffice/​publicopinion/​flash/​fl_​354_​en.​pdf

Despite the reported income gap between the self-employed and those in paid work (e.g., Hamilton, 2000), there is greater heterogeneity in the income of the self-employed. Moreover, Sorgner et al. (2017) document that 34% of the 500 wealthiest people in Germany earned their money by running their own business compared with 8% who were employees and 58% who inherited their wealth.

The major provisions of the Affordable Care Act came into force in 2014, but a significant number of additional people were covered during 2016 at the end of our sample period.

https://​www.​sba.​gov/​blogs/​franchise-fees-why-do-you-pay-them-and-how-much-are-they; https://​www.​thebalancesmb.​com

See more here: https://​www.​kauffman.​org/​kauffman-index/​about/​kiea-microdata

See more here: https://​www.​gemconsortium.​org/​

An average of 1000 households were surveyed daily across the USA until 2012. Since 2013, an average of 500 households are surveyed daily.

These estimates are roughly comparable with those reported on self-employment by the US Bureau of Labor Statistics (10.1% of all US workers including part time vs. 8.95% among full-time workers in our sample in 2015), Kauffman Entrepreneurship database (9.4% of full- and part-time workers), and Bureau of Economic Analysis (county average nonfarm proprietors employment including full- and part-time workers is 12.1% vs. 11.7% among those working full-time in our sample), even though each database measures self-employment slightly differently.

Using the General Social Survey, Kahneman and Krueger (2006) find that non-response rates for both evaluative and hedonic well-being measures are very low, below the 1% rate (in comparison with 17% non-response to questions on income). This low non-response rate is a characteristic of the Gallup survey as well, with only 0.22% of responses missing for the question on evaluative well-being. Testing whether individuals locate themselves on the same point on the scale measuring overall life satisfaction, Lucas et al. (1996) find that retesting individuals 4 weeks later leads to a 0.77 correlation between responses, reasonably near to the 0.90 correlation for most demographic questions that factually have not changed (such as education and income). A number of other scholars find that responses to life satisfaction track robustly with other indicators of well-being across countries (Graham, 2009; Sacks, Stevenson, and Wolfers, 2010; Helliwell, Layard, and Sachs, 2013). Neuroscience has also confirmed that self-reports of happiness and prefrontal cortex activity associated with positive affect correlate to a high degree (Urry et al., 2004).

It should be noted that the oversampling of the elderly is a common problem in surveys, particularly for the Gallup Poll. For example, the median age in Kauffman Survey is 43 and 56 in Gallup versus 37.9 for the US population overall. Luckily, our specific findings are less influenced by this problem, because we restrict the sample to full-time workers. Source for the US population data can be found here: https://​www.​cia.​gov/​library/​publications/​the-world-factbook/​fields/​2177.​html

Available upon request from the authors.