Work in Progress

The Presource Curse: Anticipation, Disappointment, and Governance after Oil Discoveries
Job Market Paper   Slides (20 Minutes)   Slides (75 Minutes)   World Bank Development Impact Blog

Abstract Resource discoveries are often followed by long delays and heterogeneous production realizations. Post-discovery uncertainty creates challenges for governance: policymakers may alter present behavior in anticipation of future revenues or struggle to adapt to disappointed expectations. I explore the dynamics of local governance after offshore oil and gas discoveries in Brazil. I exploit quasi-experimental subnational variation in discoveries and subsequent production realizations to identify causal effects of news and revenue shocks on municipal public finances, public goods provision, and political competition, selection, and patronage. Using a forecasting model of offshore oil production, I decompose post-discovery impacts across places where production meets expectations and places that are left disappointed. Relative to never-treated controls, places that experience discovery announcements but never receive windfalls suffer significant declines in per capita investment and public goods spending after ten years. In contrast, places where discoveries are realized enjoy significant growth in per capita revenues and spending, but do not invest in economic diversification or improve public goods provision. Electoral competition increases after discovery announcements and less-educated candidates run for and win office. My findings identify how local governments and politicians respond to shocks across time. Methodologically, I highlight the importance of accounting for dynamic treatment effects and heterogeneity in production outcomes after discovery announcements.

Labor Reallocation, Human Capital Investment, and ‘Stranded Careers’: Evidence from an Oil Boom and Bust
(with Dominic Parker and Steven Poehlekke)
Working Paper Coming Soon!   Slides (60 Minutes)

Abstract How are workers' careers shaped by exposure to a volatile sector? Drawing on employer-employee linked panel data covering the universe of formal workers in Brazil between 2003- 2017, we identify workers who are poached or newly hired into oil-linked sectors (direct, upstream, and downstream) during periods of oil boom and bust. We estimate dynamic effects of exposure to the oil sector on wages, employment, promotion, and lifetime formal earnings, relative to (extremely) closely matched counterfactual workers who are poached or newly hired into other sectors at the same time. Results show that timing of entry into oil matters: workers hired into oil-linked sectors at the beginning of Brazil's oil boom earn significantly more than matched workers hired into other sectors; workers hired immediately prior to or during busts suffer significant earnings and employment penalties. Furthermore, exposure to oil exerts dramatically heterogeneous effects on workers of differing education levels. High education workers earn more during boom years and are retained by firms during busts. Lower education workers never enjoy signifcant gains during the boom and are laid off during busts. Exploiting a complementary dataset on oil-linked graduates from higher education programs in Brazil, we document that the oil boom period was accompanied by rapid growth in oil-linked human capital investment, driven by growth in private-sector technical training programs.

Agricultural Elites, Special Interest Politics, and Deforestation in the Amazon
(with Fanny Moffette)

Abstract When economic incentives to clear land for agriculture run up against centralized efforts to combat deforestation, do local politicians intercede to promote their own economic interests or those of their campaign supporters? How does the behavior of local politicians and special interest groups change when centralized enforcement efforts are relaxed? By combining comprehensive land registries for the Brazilian Amazon, data on local political candidates and campaign donors in five municipal elections between 2000- 2016, and satellite data on deforestation and land use, we construct an original panel dataset of annual tree cover loss on properties belonging to political candidates and campaign donors in the Brazilian Amazon between 2000-2020. Drawing on this dataset, we estimate event studies around close municipal elections and find that deforestation decreases on landowning candidates’ properties while they are in office, but increases on the properties of their campaign donors. Our findings inform debates over the decentralization of resource governance, elite capture, and resource conservation policy

Can Natural Resources Promote Industrialization? Firms, Competition, and Spillovers from an Industrial Policy

Abstract Industrial policies are often hotly debated, but empirical evidence of their efficacy and underlying mechanisms is thin. I evaluate a common type of industrial policy–a local content requirement (LCR)–which requires multinational firms to source a percentage of their inputs from local suppliers. Using firm-level panel data from Brazil, I measure whether an LCR for the oil sector increased manufacturing firm growth, innovation, and productivity among upstream input-suppliers, or instead led to rent-seeking and inefficiencies. Competition is considered a primary mechanism underlying successful industrial policies. I measure whether targeted firms in more competitive subsectors exhibit higher productivity growth relative to firms in less competitive subsectors after introduction of the policy. Another justification of industrial policies is their potential to create positive spillovers. By measuring supply-chain linkages and geographical distance between targeted and non-targeted firms, I estimate spillover effects of the LCR on the broader manufacturing sector. To disentangle mechanisms, I build a structural model of demand and supply for oil inputs and estimate counterfactual firm outcomes had the LCR never been introduced. Finally, I leverage data on campaign donations made by LCR beneficiary firms and firm owners to explore the role of special interest politics in sustaining the LCR.


The Relation Between Labor Productivity and Wages in Brazil: A Sectoral Analysis
Erik S. Katovich and Alexandre Gori Maia
Nova Economia, Vol. 28, No. 1, pp. 7-38 (2018)
PDF   Published Version

Abstract Labor productivity is a crucial long-run determinant of real wages. Nonetheless, wage and productivity dynamics often diverge in practice due to a range of economic and institutional factors. This study analyzes the relation between the dynamics of labor productivity and wages in Brazil from 1996 to 2014, and adopts a sectoral perspective to account for divergent trends among economic sectors. Analyses are based on pooled data drawn from the National Accounts and the Pesquisa Nacional por Amostra de Domicílios, and hierarchical data models are estimated to assess the impacts of state- and sector-level factors on individuals’ wages. Results indicate that productivity is significantly positively associated with wage levels for all economic sectors, but that institutional factors such as labor formalization and minimum wage exert equally significant impacts, suggesting that wage growth over the 1996-2014 period was as much the result of institutional changes as of transformation of Brazil’s productive structure.

Are GMO Policies “Trade Related”? Empirical Analysis of Latin America
Pamela J. Smith and Erik S. Katovich
Applied Economic Perspectives and Policy, Vol. 39, No. 2, pp. 286-312 (2017)
Published Version

Abstract This paper empirically examines whether GMO policies are “trade related” for countries in Latin America (LA). First, we use the Balassa index to assess the “revealed comparative advantage” of LA countries. We find that LA countries have a revealed comparative advantage in GMO industries relative to the world, and that intra-regional trade in these industries is modest relative to external trade. Second, we estimate the Gravity model to examine the effects of importers’ GMO policies on Argentina and Brazil’s bilateral exports of soybeans and maize. We find that strong GMO policies in importers have a negative effect on Argentina’s bilateral exports of soybeans (an industry and country with historically high GMO content). Further, we find that past GMO policies are a strong determinant of Argentina’s future bilateral exports, and that the negative trade effects of strong GMO policies are increasing over time. In contrast, we find a weaker relationship between the GMO policies of importers and Brazil’s bilateral exports (consistent with Brazil’s more recent increases in GMO content). These findings for Argentina and Brazil provide a benchmark for other developing countries that are looking for guidance on servicing trading partners with diverse GMO policies.

Other Projects

What do we really know about the impacts of improved grain legumes and dryland cereals: A critical review of 18 impact studies
Erik S. Katovich, Andrew W. Feist, Karl Hughes, Kai Mausch, and Michael Hauser
ICRAF Working Paper Series, No. 295, World Agroforestry-Nairobi (2019).