Legal positivism as a jurisprudential theory is often associated with the project of understanding 'law as it is' rather than 'law as it ought to be.' H.L.A. Hart, widely regarded as the most influential legal philosopher of the 20th century and the founder of contemporary legal positivism, characterized his project as 'descriptive sociology,' an attempt to provide theoretical insight into the workings of actual modern legal systems. Since most people, as well as most lawyers, would say that the law’s enforceability through sanctions, whether criminal or civil, is a key feature of law, it is surprising that Hart rejected the view that coercion is essential to the nature of law. It is equally surprising that Joseph Raz, the most influential positivist of the last quarter century, agrees with this rejection. Indeed, according to Hart, coercion is merely ancillary to the main functions of law in society, and according to Raz, coercion is not an essential feature of law because it plays no necessary conceptual role in elucidating and distinguishing law as an institutionalized normative system.In this article, I argue that Hart and Raz are both mistaken here - i.e., that coercion is essential to the nature to law. In particular, with respect to Hart, I argue that (1) in order to understand his account regarding the role of coercion in a theory of the nature of law, we must analyze his criticisms of John Austin’s and Hans Kelsen’s sanction-centered accounts of law; (2) Hart’s critique of Austin and Kelsen reveal a fundamental role for the concept of obligation; (3) Hart’s critique of Austin and Kelsen reveal a fundamental role for an explanation of the experiences of law-subjects; (4) Hart rejects sanction-centered accounts because they could not adequately account for legal obligation and obscured how law functions in society; (5) Hart’s theory of obligation, however, requires an explication of the legal sanction in order to differentiate legal obligation from other forms of obligation; and (6) Hart’s focus on the experience of law-subjects undermines his claim that the internal point of view is primary for understanding the nature of modern legal systems.With respect to Raz, I argue that: (1) in order to understand his account of the role of coercion in a theory of the nature of law, we must first understand his account of institutionalized normative systems; (2) his account of such systems apparently implies that coercion is unnecessary for the existence of a legal system; (3) this account, however, fails because it does not provide a basis for distinguishing a legal system from other institutionalized normative systems within the same society (specifically, religious systems and popular morality); and (4) his conceptual methodology implies a different conclusion than the one offered.