In community settings in low-income developing countries better forest management depends on collective action (CA), but if CA really offers better incentives than open access, we should observe behavioral differences across CA levels. In this paper we examine one potential farm-level behavioral effect by trying to isolate and understand the effects of community forest CA on households’ incentives to invest in trees located on their own farms. Using a household level analytical model, we find that more stringent forest CA should create incentives for private tree planting as a substitute for overusing community forests. We test this hypothesis using detailed measures of highland Ethiopia forest CA attributes taken directly from the rich CA literature and a variety of empirical specifications. Though we are unable to draw firm conclusions due to the nature of our data, we do find robust evidence across specifications that more effective forest collective action causes households to plant more trees on their farms.