Is there adaptation to predictable climate change along the temperature-conflict nexus? Evidence from the El Niño Southern Oscillation

ABSTRACT

A growing body of research connects short-run deviations in weather with violence. Less well understood is the scope for agents to adapt to medium and longer-run climate fluctuations. We follow existing research and use the existing climactic forces of the El Niño Southern Oscillation (ENSO) to analyse climate change – exploiting the fact that in 1989 published forecasts of these fluctuations became available. In a generalized differences framework, we contrast the impact of ENSO in affected areas of the globe relative to unaffected areas before and after 1989, finding no robust evidence that adaptation efforts are successfully mitigating conflict or civil unrest occurring as a result of these fluctuations.

KEYWORDS:

• Adaptation
• climate change
• El Nino
• civil conflict and unrest

JEL CLASSIFICATION:

• O1
• Q5
• D70

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

¹ For broad summaries see Burke, Hsiang, and Miguel (2015) and Dell, Jones, and Olken (2014).

² Climate specific research has predominantly focused on estimating the impact of temperature and rainfall deviations on factors such as civil conflict, interpersonal crime, and economic outcomes such as agricultural yields (Burke, Hsiang, and Miguel 2015; Dell, Jones, and Olken 2014). Within this research, analysis of social and political unrest, particularly contemporary unrest, is relatively understudied. Several historical studies provide evidence that climate change may heavily influence behaviour (Dell, 2012; Chaney 2013; Kung and Ma 2014; Jia 2014; and Iyigun, Nunn, and Qian 2017). These analyses are representative of the impact of longer-run climactic variation among historical populations which may have had less scope for recognizing climate changes and potentially fewer resources and technologies available for adaptation.

³ Alternatively, some changes in the frequency or intensity of shocks may be harder for individuals to internalize as abnormal when they occur over longer periods of time.

⁴ Shrader (2017) is, to our knowledge, the first paper in economics to detail the ENSO forecasting process, utilizes the precise forecasts in a careful microeconomic analysis, and should be credited with highlighting the timing of forecast availability as a potential event study for the field.

⁵ The measure of El Niño follows Hsiang, Meng, and Cane (2011). It measures the anomalies in sea surface temperature over the NINO3 region in the Pacific. Positive values refer to hotter than normal temperatures (El Niño) and negative values refer to cooler than normal temperatures (La Niña).

⁶ It may have taken time for forecasts to become accurate and for agents to become widely aware of their availability. While plausible, this does not appear to be driving our results, as we obtain similar findings if we exclude the first few years after forecasts become available from the analysis (results available upon request).

⁷ For example, Cai et al. (2014) suggests that the process of global climate change may as much as double the incidence of extreme El Niño events.