In her gracious and incisive review, Miriam R. Lowi raises three main questions: 1) whether Russia is indeed a case of private domestic ownership (P₁); 2) whether there is a fundamental difference between state ownership with control (S₁) and without control (S₂); and 3) whether we adequately take into account alternative explanations for the variation in ownership structure (OS) across petroleum-rich states over the twentieth century.

Lowi's doubt as to the coding of Russia's initial OS is understandable. It lies in our acknowledgment that although the entire petroleum sector was privatized in the early 1990s, the oil sector was dispersed into several private, vertically integrated companies (VICs), whereas the gas sector was privatized as a single entity (Gazprom) that enjoyed monopoly status and close ties to the government. Erika Weinthal and I thus describe the latter as nominally (or partially) privatized. While the distinction complicates our analysis, it provides an opportunity to further test our hypothesis that P₁ promotes strong fiscal regimes, in part by establishing clear boundaries between the main claimants (i.e., enterprise owners and governing elites) that foster mutual incentives to build institutions that set effective limits on the state's ability to extract and spend. Simply put, both sides can capture the gains from more predictable revenue streams, including more efficient tax collection and budgetary stability, which in turn affects investment capacity. In Russia's gas sector, however, the boundaries between the main claimants were conspicuously blurred. We thus expected not only the preferences of the oil and gas sectors to diverge but also their trajectories.

Their respective responses to the 1998 financial crisis serve to illustrate. Soon after, privately owned VICs in the oil sector and the Russian government entered into negotiations that produced a formal broad-based tax code, which enabled the VICs to undergo internal restructuring and eventually expansion. In contrast, Gazprom's managers displayed little interest in curtailing the tacit agreement with the government to provide huge gas subsidies to domestic consumers in exchange for enriching themselves personally, leaving the company mismanaged and undercapitalized.

Lowi is also correct that the impact of S₁ and S₂ on fiscal regimes is indistinguishable prior to the 1990s. Our argument is that the prevalence of the obsolescing bargain during the 1970s and 1980s generated incentives for weak fiscal regimes under both forms of OS, and thus, the poor developmental outcomes associated with the resource curse. Previous scholars have therefore conflated them. Yet the subsequent shift in relative bargaining power away from host governments and the emergence of corporate social responsibility (CSR) meant that foreign investors (FIs) could exert a positive influence on fiscal regimes under S₂. We are careful not to overstate this claim. Rather, we specify two conditions under which FIs are most likely to foster taxation and spending policies that are qualitatively different and substantively better from the perspective of the population: They are strongly committed to CSR and able to enforce the terms of their contracts.

Lowi's concern that international factors may be driving the variation in OS over the twentieth century is also valid, particularly given the temporal trends in the data. Yet our statistical results (based on an original cross-sectional data set) suggest that the global trends in the structure of ownership over time are more accurately attributed to the coincidence of domestic constraints across petroleum-rich countries. Controlling for international market conditions (including technology and difficulty of extraction), as well as pressures for policy diffusion (including OPEC), we find strong support for our argument that the variation in OS cannot be explained without incorporating the degree to which access to alternative sources of revenue (AR) and the level of distributional conflict (DC) constrained state leaders' choice of OS. We do not claim to explain every case, but to provide a more complete aggregate explanation. Kuwait and Saudi Arabia serve to illustrate: Lowi is correct that they faced the same domestic constraints (↓ AR, ↓ DC) in the 1960s and 1970s as the other Gulf States, and yet adopted S₁. The other five, however, chose S₂, as expected.

In sum, Lowi's review is a welcome challenge to the alignment between the general theory and the analysis and coding of specific cases. I appreciate the chance to respond and clarify.