Is tax compliance driven only by extrinsic motivations such as deterrence and tax policy or is there also a role for intrinsic motivations such as morals, norms and psychology? Agents may comply based on moral sentiments, social norms, guilt and shame (Andreoni et al. 1998), all of which are non-deterrence driven reasons for compliance. The importance of such intrinsically motivated compliance is hard to study empirically and therefore the least understood. This study uses a unique setting for making progress on this question: the local church tax in Germany. As we show in the paper, tax evaders, compliers, and donors can coexist in the local church tax system and be precisely distinguished from each other. Since there is zero deterrence in the baseline, baseline compliance provides a direct measure of intrinsically motivated tax compliance. Starting from the zero deterrence baseline we use a randomized field experiment to inject deterrence or recognition into the system. This allows us to study if policies aimed at either extrinsic motivation (deterrence) or intrinsic motivation (recognition) have qualitatively different effects on agents who have revealed each of those motivations in the baseline. Our main empirical findings are the following. First, a significant fraction of agents (23%) comply in the zero deterrence baseline where compliance would be zero absent intrinsic motivation, while the remaining 77% evade the tax. Intrinsic motivation is therefore substantial, although the majority behaves as rational, self-interested taxpayers. Second, announcing a zero audit probability (the status quo) has a small and insignificant effect on the compliance rate, showing that there is very little misperception on average. Third, tax salience and deterrence have strong effects on compliance for baseline evaders, but small and insignificant effects for baseline donors. This is consistent with the fact that the enforcement constraint is not binding for the intrinsically motivated and therefore they are naturally unresponsive to deterrence. Fourth and finally, recognition through social and monetary rewards for compliance has fundamentally different effects on baseline donors (who increase their payments) and baseline evaders (who reduce their payments). Hence, whether recognition helps or hurts depends crucially on what motivates taxpayers in the first place, with positive effects on the intrinsically motivated and negative effects on the extrinsically motivated. All of our findings can be explained by a model of tax compliance that unifies the standard Becker-Allingham-Sandmo approach (strengthening extrinsic motives for tax compliance) and the Andreoni (1989, 1990) warm-glow model of pro-social behaviour.
