VoxEU Column Development Frontiers of economic research

Climate, ecosystem resilience, and the slave trade

The slave trade continues to shape modern Africa. This column analyzes environmental shocks to the supply side of the trans-Atlantic slave trade and their long-term effects. During warm periods, African ports exported fewer slaves because lower agricultural productivity raised slavers' costs. These temperature fluctuations had long-run impacts, and ports that experienced a warmer period during the decades when the slave trade was most active appear more developed today.

The slave trade is important if we want to understand Africa’s modern development. Higher participation in the trade is associated with worse modern economic outcomes, including lower GDP per capita, lower trust, and greater ethnic fractionalisation (Nunn 2008; Nunn and Wantchekon 2011; Whatley and Gillezeau 2011). Despite the numerous studies linking modern outcomes to participation in the slave trade, there has been very little empirical analysis of the mechanisms affecting the supply of slaves from Africa.

In recent research (Fenske and Kala 2013), we examine the impact of environmental factors on the number of slaves exported from African ports. We find that higher temperatures that depressed African agricultural productivity also reduced slave exports. Specifically, we estimate the impact of annual port-level temperature shocks on slave exports from 134 ports during the period from 1730 and 1866. We match slave exports from the Trans-Atlantic Slave Trade Database of Eltis et al. (1999) and historical gridded temperature data from Mann et al. (1998), and match them to the closest African ports. Our analysis considers the Atlantic slave trade, which comprised the majority of African slave exports between 1400 and 1900.

Testing the importance of temperature shocks on slave exports

On average, ports in our data exported roughly 450 slaves annually. In years when exports were non-zero, average exports were 2,500 slaves. We find that a one degree increase in temperature leads to a reduction of roughly 3,000 fewer slave exports. This is comparable to the mean number of slaves exported from an active port. Further, including long-run climate (30-year average temperature) in the analysis, we find that short-run temperature shocks and long-run trends both have the ability to explain slave exports. The impact of a one-degree increase in the climate trend has the same direction as a short-run shock. That is, fewer slaves are exported in warmer periods. The magnitude of the effect of climate is much larger. This is in keeping with the literature that finds that long-run changes to climate such as multi-decadal droughts cause depopulation on a large scale through mortality and out-migration, and may cause societies to collapse (Barrios et al. 2010, Haug et al. 2003, Weiss and Bradley 2001).

Two important mechanisms explain our results:

  • First, warmer years are years of lower agricultural productivity in Africa.

This increases the costs of producing slaves, since provisions needed to be secured for the both slaves and the armed groups that enslaved them. Further, areas of slave supply became more disordered during periods of depressed agricultural productivity. To test this hypothesis, we examine the heterogeneous impact of higher temperatures across agro-ecological zones in Africa. These are zones defined by their similar climate, soil characteristics, and growing seasons. Drier zones respond more strongly to the shocks in our data. Previous research has identified these as less resilient to temperature increases (Seo et al. 2009). That is, zones where agricultural returns fluctuate most with temperature are those in which the marginal impact of temperature on slave exports was greatest.

  • A second, complementary explanation for our result is that higher temperatures are years of greater mortality.

This is caused directly by a higher disease burden and indirectly by the impact of lower agricultural productivity on food scarcity. Both these links limit slave exports in warmer years and imply that these impacts will be greatest in the least resilient ecological zones.

We examine three case studies of ports that were statistically influential in our data – Benguela, Mozambique, and Whydah. We show that the historical evidence emphasises the strong complementarity of the agricultural sector with the slave trade. Agriculture was an economic base for local rulers who raised armies and conducted raids for slave capture. It was a source of provisions for slave ships.

Using additional data, we rule out several competing explanations that would interpret temperature as a demand-side shock. For example, warmer years could also be years of greater natural hazards for ships, lower wind speeds, or longer voyage lengths. Our data allow us to show that natural hazards increased in colder years, that the correlation between temperature and wind speed is small, and that ships spent took more than one year in Africa. In addition, including the temperature from the nearest New World ports of slave demand or region-specific slave prices do not change the substance of our results. We are also able to rule out the possibility that any particular region or the use of ports as units of observation drive our result.

Thus, we are able to show that, unlike several modern settings in which environmental stress exacerbates conflict, periods of historic environmental stress during the slave trade reduced the incidence of violent slave raiding. The dynamics of the slave trade, then, followed a logic similar to the model of Besley and Persson (2011): greater state revenues encouraged greater repression. This mirrors other cases in which institutions lead resource rents to be captured by elites to perpetuate conflict (Mehlum et al. 2006, Ross 2004).

Modern impact of historic temperature shocks

Finally, we examine how temperature fluctuations in the past predict current economic outcomes. Warmer years were years of lower agricultural productivity, but also years in which fewer slaves were exported. Because historic participation in the slave trade is associated with worse development outcomes, it is possible that the long-run effects of reduced slave exports offset temporarily depressed productivity in the long run. 

Following Michalopoulos and Papaioannou (2013), we use modern night-lights data in the areas surrounding our ports in order to measure modern development. We show that whether warmer years during the peak of the slave trade, measured relative to modern temperatures, imply lower modern luminosity in the cross-section of ports. This shows that the effect of temperature on slave exports dominates the long-run effect of depressed agriculture. Ports that experienced a warmer period during the decades when the slave trade was most active are more luminous today.


Barrios, S, Bertinelli, L, and Strobl, E (2010), “Trends in rainfall and economic growth in Africa: A neglected cause of the growth tragedy”, Review of Economics & Statistics, 92(2), 350–366.

Besley, T and Persson, T (2011), “The logic of political violence”, The Quarterly Journal of Economics 126(3), 1411–1445.

Eltis, D, Behrendt, S D, Richardson, D, and Klein, H S (1999), The trans-Atlantic slave trade: a database on CD-ROM, Cambridge University Press.

Fenske, J and Kala, N (2013), “Climate, ecosystem resilience and the slave trade”, CEPR Discussion Paper Number 9449.

Haug, G, Günther, D, Peterson, L, Sigman, D, Hughen, K, and Aeschlimann, B (2003), “Climate and the collapse of Maya civilization”, Science, 299(5613), 1731.

Mann, M, Bradley, R, and Hughes, M (1998a), “Global-scale temperature patterns and climate forcing over the past six centuries”, Nature, 392(6678), 779–787.

Mann, M, Bradley, R, and Hughes, M (1998b), “Global six century temperature patterns”, IGBP PAGES World Data Center.

Mehlum, H, Moene, K, and Torvik, R (2006), “Institutions and the resource curse”, The Economic Journal, 116(508), 1–20.

Michalopoulos, S, and Papaioannou, E (2013), “Pre‐Colonial Ethnic Institutions and Contemporary African Development”, Econometrica, 81(1), 113-152.

Nunn, N (2008), “The long-term effects of Africa’s slave trades”, The Quarterly Journal of Economics, 123(1), 139–176.

Nunn, N and Wantchekon, L (2011), “The slave trade and the origins of mistrust in Africa”, The American Economic Review, 101(7), 3221–3252.

Ross, M L (2004), “What do we know about natural resources and civil war?”, Journal of Peace Research, 41(3), 337–356.

Seo, N, Mendelsohn, R, Kurukulasuriya, P,Dinar, A, and Hassan, R (2009), “Differential adaptation strategies to climate change in African cropland by agro-ecological zones”, Environmental & Resource Economics 43(3), 313–332.

Weiss, H and Bradley, R (2001), “What drives societal collapse?”, Science 291(5504), 609.

Whatley, W and Gillezeau, R (2011b), “The impact of the transatlantic slave trade on ethnic stratification in Africa”, The American Economic Review 101(3), 571–576.

4,199 Reads