The accuracy of the measured income of family enterprises is a matter of importance in studies of, inter alia, human capital, income distribution, and consumption behavior. Enterprise surveys can measure this income through detailed questions. Household surveys offer a better perspective for a study of living standards and poverty: they capture more of the truly small-scale one-person enterprises; there is a wealth of relevant information about the household; and household surveys allow one to integrate family enterprises into household decisions about labor supply, risk sharing, enterprise start-up, and asset formation. This paper examines three enterprise income values that one may derive from household surveys held in the Cote d'Ivoire and Ghana. The three values vary much and do not correlate all that well. This sobering conclusion implies that relying on self-reported values of sales revenue, expenditures and enterprise earnings is risky. Greater effort should be made to measure the transactions of enterprises carefully. Using worksheets and cross-checking responses in loco should help, but since many enterprises do not use any accounting system, it may be necessary to monitor inflows and outflows either personally or with diaries.
