Abstract This paper presents an alternative perspective to a longstanding empirical puzzle: that most entrepreneurs persevere despite persistently low earnings. Because entrepreneurial earnings are notoriously difficult to measure, I approach the question from an expenditure angle. I look at how switching into self-employment corresponds to changes in reported earnings versus expenditure. Using 45 years of longitudinal data, I find that individuals report earning 27.7% less in self-employment, while spending 3.8% more. This household expenditure premium accrues with entrepreneurial experience and is not offset by lower savings or longer work hours. The results hold in highly educated and incorporated business owner subsamples.