Tax effort revisited : new estimates from the Government Revenue Dataset

Attention on domestic resource mobilization-particularly in developing countries- has increased significantly in recent years. This stems from, among other things, recognition in the Sustainable Development Goals that further domestic funding is required for development needs, and the Addis Tax Initiative, which aims to foster fairer and more effective domestic resource mobilization. And whilst there is a recognition that many low- and middle-income countries could be collecting more in tax revenues, the answer to the question of just how much more is unclear. So-called tax effort studies have attempted to shed light on this question for many years and recent estimates suggest that many developing countries are not performing anywhere near their potential (i.e. exerting enough effort). This study makes two significant contributions to the tax effort literature. First, we find that the stochastic frontier approaches used in many recent studies are sensitive to empirical specification and the resulting tax effort scores are strongly influenced by outlying input observations. We employ the True Random Effects approach and find that tax effort scores are, on average, higher and more tightly distributed than previous studies suggest. Second, we enjoy improved data coverage due to using the most recent version of the UNUWIDER Government Revenue Dataset, alongside governance indicators from the V-Dem dataset. This allows us to present tax effort estimates for 50 per cent more observations than the next most complete study.