The revenue management system engenders non-rule-based tax collection. The strongest concern they raise is not corruption, but the conversion of what would have been compliant taxpayers into “semi-compliant” ones, i.e. compliant with the dictates of government employees rather than the requirements of law. I lay out a conceptual framework for analyzing the revenue management system, captured by “atomistic coercion.” “Atomism” means that (i) instead of a government organization, the authorities that taxpayers obey are individuals; and (ii) tax collectors acquire taxpayer information that they need not share with the state apparatus. “Coercion” conveys the idea that taxpayers perform compliance activities only because they are monitored by government authorities, not because of some social norm of complying with the law. Coercion secures compliance through government observation and intervention. Intervention can secure compliance along behavioral margins that are accurately observed. But as the state’s powers of observation and intervention are bounded, compliance can prove very low along the vast portions of taxpayer behavior not observed by the state or where the state’s capacity for intervention is constrained. This is illustrated by several examples, most importantly the phenomenon of taxpayers formally registering with tax authorities while dodging most specific compliance obligations.

Chapter 8 explores the implications of the book’s narratives for the study and practice of taxation in developing countries. China has adopted taxes that are designed to be administered through self-assessment, while minimizing reliance on self-assessment. This essential modification of the modern tax paradigm has radical implications. First, it suggests substantial revisions to economic theories of tax administration and compliance. In particular, because self-assessment is quintessentially about complying with legal norms, modern taxation is deeply intertwined with legal systems: the Chinese case demonstrates how profoundly different a tax system is when severed from the legal system’s support – an idea that economic theories have ignored. Second, the evolution of Chinese tax administration both affirms and enriches a theory in political economy, namely that, to properly study tax and development, one must consider not only state capacity in tax collection, but also view such state capacity as endogenous with politics. Finally, one may query whether Chinese taxation will converge with more familiar tax paradigms. I argue that, to the contrary, technological developments could strengthen the tendency of China’s tax system to continue on its recent path; one can even ask whether other countries will converge to China’s model.