Oil Prices and International Conflict: Why Low Oil Revenue May Not Pacify Petrostates¹

Abstract

This article explores how declining oil revenue might shape the amount of international conflict initiated by major oil producers (petrostates). We analyze four potential mechanisms through which variation in oil prices could affect petrostate conflict initiation: emboldenment, battling over a smaller market, signaling strength, and diversionary conflict. The empirical findings suggest that higher oil prices are associated with lower rates of petrostate conflict initiation. From one standard deviation below the mean oil price to one standard deviation above it, the predicted number of militarized interstate disputes declines twofold, from .025 [95% CI: .016–.034] per petrostate per year to .012 [.007–.016]. Moreover, the evidence suggests that petrostates are more likely to target other petrostates when oil prices are low. This suggests that the energy transition may not be a boon for international peace among petrostates, and for a time, it may even prove to be the opposite.

RESUMEN

Este artículo explora cómo la disminución de los ingresos petroleros podría influir en la magnitud del conflicto internacional iniciado por los principales productores de petróleo (petroestados). Analizamos cuatro mecanismos potenciales a través de los cuales la variación en los precios del petróleo podría afectar el inicio de un conflicto entre petroestados: envalentonamiento, lucha por un mercado más pequeño, señalización de fuerza y conflicto de distracción. Los hallazgos empíricos sugieren que los precios más altos del petróleo están asociados con tasas más bajas de inicio de conflictos petroestatales. De una desviación estándar por debajo del precio medio del petróleo a una desviación estándar por encima de él, el número previsto de disputas interestatales militarizadas se reduce a la mitad, de 0,025 [IC del 95%: 0,016-0,034] por petroestado por año a 0,012 [0,007-0,016]. Además, la evidencia sugiere que es más probable que los petroestados apunten a otros petroestados cuando los precios del petróleo son bajos. Esto sugiere que la transición energética puede no ser una bendición para la paz internacional entre los petroestados y, durante un tiempo, incluso puede resultar lo contrario.

RÉSUMÉ

Cet article explore la manière dont la baisse des revenus pétroliers pourrait influencer l’ampleur des conflits internationaux déclenchés par les principaux producteurs de pétrole (pétroétats). Nous analysons quatre mécanismes potentiels par lesquels la variation des prix du pétrole pourrait affecter le déclenchement d’un conflit pétro-étatique : l’enhardissement, la lutte pour un marché plus petit, le signal de force et le conflit de diversion. Les résultats empiriques suggèrent que des prix pétroliers plus élevés sont associés à des taux plus faibles de déclenchement de conflits pétroétatiques. D’un écart type en dessous du prix moyen du pétrole à un écart type au-dessus, le nombre prévu de conflits interétatiques militarisés diminue de deux fois, passant de 0,025 [IC à 95 % : 0,016 à 0,034] par pétro-État et par an à 0,012 [0,007-0,016]. De plus, les données suggèrent que les pétro-États sont plus susceptibles de cibler d’autres pétroÉtats lorsque les prix du pétrole sont bas. Cela suggère que la transition énergétique n’est peut-être pas une aubaine pour la paix internationale entre les États pétroliers, et que, pendant un certain temps, elle pourrait même s’avérer être le contraire.

Keywords:

• Conflict
• energy transition
• oil
• petrostates

Disclosure statement

Funding for this research was provided by the Research Council of Norway (grant number 324628).

Notes

2 Though there is a body of literature on the link between oil prices and civil war (see Andersen, Nordvik, and Tesei [2022]).

3 Ross (2004) undertakes a similar enterprise for the study of natural resource wealth and civil war by testing several competing mechanisms by which resource wealth might affect civil war.

4 The phrase “Battling over a Smaller Market” is based on a similar one used by Van de Graaf (2018).

5 See also the literature on the dynamics of power transitions and declining states (Organski 1958; Gilpin 1981; Levy 1987; Lemke 2002).

6 Correlates of War Militarized Interstate Dispute Data, Version 5.0, November 3, 2020.

7 Tables A20–A23 also report results using a binary dependent variable estimated with a linear probability model and logistic regression model. The results are consistent with the main results, but are weaker when country fixed effects are included due to the limited variation in the dependent variable in some country-decades.

8 Correlates of War, Militarized Interstate Disputes (v5.0), https://correlatesofwar.org/data-sets/mids/.

9 Table A24 reports results using three- and five-year moving averages of the change in the oil price, which gives a better sense of the trend in oil prices over a longer period in a way that is less volatile than the annual change. The results indicate that when oil prices decline, petrostate conflict initiation increases.

10 Note that Colgan’s (2012) Revolutionary Leader Dataset (v1.1), only extends through 2004, which is why we do not include it in the main analysis. As shown in Tables A14–A15, when the analysis is limited to this time period, inclusion of the revolutionary government variable does not change the results.

11 GDP and GDP per capita data are from the World Bank (2022) and are in constant 2012 USD. Polity score is from Version 5 of the Polity project (Marshall, Gurr, and Jaggers 2020). Table A13 excludes GDP from the model specification, as oil prices and GDP are likely to covary in petrostates in ways that might cancel each other out. We thank an anonymous reviewer for this suggestion.

12 Estimates for Figures 1–4 are created using Stata 16’s margins and marginsplot commands.

13 The results are robust to using net oil exports in place of oil production (Table A8), and to using each country’s time-invariant average level of production (Table A9).

14 For MIDs with hostility level of 2+, 91.13% of country-years initiated zero, 7.69% initiated one, 0.91% initiated two, and 0.27% initiated three. For MIDs with hostility level of 4+, 94.14% of country-years initiated zero, 5.22% initiated one, 0.55% initiated two, and 0.09% initiated three.

15 The results are robust to controlling for the number of other petrostates each country bordered as well (Table A16).