This paper empirically examines firm owners' gender difference in labor demand. We estimate the average treatment effect (ATE) of female ownership on employment of the firm using the 2007 Survey of Business Owners (SBO) Public Use Micro Sample (PUMS), provided by the U.S. Census Bureau. Because female microentrepreneurs potentially demand more labor so as to allocate time for household production, we hypothesize a condition under which female microentrepreneurs employ more, and that is, if they are free from financial constraints. We show first that the estimation of the ATE for female ownership can have a downward selection bias that may yield negative ATE estimates, and this downward selection bias comes from male owners being less financially constrained than female owners. We then perform the two-stage least squares (TSLS) estimation using two sets of instrumental variables (IVs), which are indicator variables for i) inheritance; and ii) loans from bank or family/friend. The estimation results present that the female owner effect on labor demand as local average treatment effect (LATE) is identified and consistently estimated by using the IVs. From the main model estimation, we find a positive and statistically significant female owner effect that female owners hire more employees than male owners by about 25.8%.