Winning the Oil Lottery: The Impact of Natural Resource Extraction on Growth

Abstract

Standard-Nutzungsbedingungen:Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden.Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen.Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in AbstractThis paper studies the effect of oil discoveries on economic growth in Brazilian municipalities for the period from 1940 to 2000. It uses a unique identification strategy which exploits data on the drilling of approximately 20,000 oil wells in Brazil since oil explorations began in the country. We argue that oil discoveries are randomly assigned conditional on geological characteristics. The quasi-experimental outcome from drilling generates the treatment assignment: municipalities where oil was discovered during drilling constitute the treatment group while municipalities with drilling but no discovery are the control group. In our preferred specifications we find that oil discoveries increase per capita GDP by 25% over the 60-years period compared to the control group. Importantly, oil extraction has positive spillovers to other sectors of the economy. Services GDP per capita is estimated to increase by roughly 20% and urbanization by over 4% points. We show that the increase in services GDP per capita is mainly due to an increase in labor productivity. In line with intuition, these spillovers are present for onshore but not for offshore discoveries.Among other potential channels, the results are consistent with an increase in local demand for non-tradable services driven by the oil producing firm and highly paid oil workers.

Dutch disease and the oil boom and bust

“…Caselli and Michaels (2009) exploit differences in oil endowment across Brazilian municipalities, finding little effect on non-oil GDP. Cavalcanti et al (2016) also find that oil discoveries in Brazilian municipalities have positive spillovers for manufacturing and services. Allcott and Keniston (2018) use an approach similar to this paper and find that oil booms increase manufacturing output in oil-rich United States counties.…”

“…Cavalcanti et al. () also find that oil discoveries in Brazilian municipalities have positive spillovers for manufacturing and services. Allcott and Keniston (2018) use an approach similar to this paper and find that oil booms increase manufacturing output in oil‐rich United States counties.…”

Understanding the resource curse (or blessing) across national and regional scales: Theory, empirical challenges and an application

“…Although studies econometrically evaluating the impacts of mining on regional economic development are still not abundant, some researchers have provided important insights on this matter (e.g. Black et al 2005;Marchand 2012;Partridge et al 2012;Reeson et al 2012;Caselli and Michaels 2013;Cavalcanti et al 2014;Weber 2014;Fleming and Measham 2015a;Paredes et al 2015).…”

The Impact of Natural Resource Discoveries in Latin America and the Caribbean: A Closer Look at the Case of Bolivia

“…Caselli and Michaels (2013) and Monteiro and Ferraz (2012) show that the fiscal windfall from oil tends to inefficiently spent. On the other hand, Cavalcanti et al (2016) show that the pure market effect (abstracting from the fiscal channel) of having an oil sector is beneficial for Brazilian municipalities and leads to structural transformation over the long-run due to the positive demand and investment associated with it. Aragon and Rud (2013) show that a large gold mine in Peru lead to positive spillovers to the local economy, while Loayaza and Rigolini (2015) find that gold mines in Peru reduce local poverty.…”

“…Like the now popular synthetic control method, entropy balancing implicitly makes a strong linearity assumption, however. As a robustness test we also use alternative ways to improve overlap between the treatment and control group.Tables are available on request.25 See, for example,Loayza et al (2013) andCavalcanti et al (2016) for related identification strategies to retrieve the impact of natural resource extraction on local socio-economic outcomes.26 Recall that the share in manufacturing employment in Bolivia as a whole fell between 2001 and 2012 and the results here are relative to other Bolivian municipalities.©International Monetary Fund. Not for Redistribution…”