A two-part model is applied to a unbalanced panel data consisting of 39 not-for-profit hospitals from 2000 to 2010, in order to estimate the relationship between board composition and probability of donation revenues, and association with board and donation revenue amount. The first part of the two-part model used logit analysis to predict the probability of donation revenues and the board composition, the empirical result is insignificant. The second part of the linear regression model indicated that both the board size and CEO on board are negative associated with the donation revenue. CEO on small size board has positive relationship with the donation revenue. The result indicated that the board structure and composition were the determinants of hospitals’ donation revenue, however, were irrelevant to the probability of receiving donation revenues.
2012 ARNOVA’s 41th Annual Conference, Association for Research on Nonprofit Organizations and Voluntary Action