![Sample our Global Development journals, sign in here to start your access, latest two full volumes FREE to you for 14 days]({"type":"image" , "src" : "/sda/5929/TOC-Subject-Sample-2021-Global-Development.jpg"})
ABSTRACT
We show new evidence that the consequences of historical warfare for state development differ for Sub-Saharan Africa. We identify the locations of more than 1,600 conflicts in Africa, Asia, and Europe from 1400 to 1799. We find that historical warfare predicts common-interest states defined by high fiscal capacity and low civil conflict across much of the Old World. For Sub-Saharan Africa, historical warfare predicts special-interest states defined by high fiscal capacity and high civil conflict. Our results offer new evidence about where and when ‘war makes states’.
We thank Idalia Bastiaens, Robert Bates, Catherine Boone, Peter Carroll, Christian Davenport, James Fearon, Nahomi Ichino, Eliana La Ferrara, James Morrow, Torsten Persson, James Robinson, Frances Rosenbluth, Kenneth Scheve, Peter Schram, Guido Tabellini, Warren Whatley, and seminar participants at APSA 2015, Bocconi, Caltech, EPSA 2015, George Mason, MPSA 2015, Michigan, Queen Mary, Stanford, UCL, and WEHC 2015 for helpful comments. We thank Heather Elliott, Maiko Heller, and Eric Payerle for research assistance, and the Department of Political Science at the University of Michigan for financial support.
ORCID
Mark Dincecco http://orcid.org/0000-0003-2860-7069
James Fenske http://orcid.org/0000-0001-6003-629X
Massimiliano Gaetano Onorato http://orcid.org/0000-0002-4574-5853
Notes
2 In the paper’s title, we follow a convention in the literature (e.g. Herbst Citation2000) that employs the term ‘Africa’ as shorthand for ‘Sub-Saharan Africa’. Throughout the paper’s text, however, we always explicitly distinguish between Sub-Saharan Africa and North Africa.
3 Our results support the evidence in Osafo-Kwaako and Robinson (Citation2013), who use the Standard Cross-Cultural Sample (Murdock & White Citation1969) to study political centralization in pre-colonial Sub-Saharan Africa relative to the rest of the world. Like us, they find that the logic of historical state development differed for this world region. Namely, there is a positive and significant correlation between warfare and political centralization for the whole world, but there is no such correlation for pre-colonial Sub-Saharan Africa.
4 By contrast, Osafo-Kwaako and Robinson (Citation2013) find no such correlation. They analyse pre-colonial Sub-Saharan Africa relative to the rest of the world.
5 By contrast, new labour market access under colonialism may have helped reduce the political power of traditional political elites (Meier zu Selhausen et al. Citation2018).
6 Boone (Citation2014) argues that land-related conflicts in modern-day Africa can actually be an outcome of state-building efforts. Heldring (Citation2014) finds that greater state capacity led to more conflict in 1990s Rwanda.
7 This definition is: ‘An occurrence of purposive and lethal violence among 2+ social groups pursuing conflicting political goals that results in fatalities, with at least one belligerent group organized under the command of authoritative leadership.’
8 Similarly, we lack accurate enough data to code the modern country of origin for each belligerent participant. To the extent that participants in raiding wars in Sub-Saharan Africa were located within the borders of the same modern country, however, then our coding scheme should accurately reflect such historical conflicts. For example, both the Kingdom of Kongo and the Ambundu Kingdom were located (at least in part) within the modern country of Angola, and, according to Brecke, the 1514 conflict ‘Kongo-Ambundu (northwest Angola), 1514’ took place there.
9 This approach is similar to dividing continents into square grids (e.g. 100×100 km). As described, an advantage of using modern borders is that far more covariates are available at the country level than at the grid cell level.
10 Large conflicts may lead to greater fiscal reforms than small conflicts. To measure conflict intensity, one could use casualty totals (Dincecco & Prado Citation2012), but these data are only available for about one-third of Brecke’s conflicts. A second possibility is to incorporate conflict durations in days or months. However, specific start and end dates are not available for over 70% of the Brecke data.
11 The main results ahead remain robust if we code historical conflict as log(1 + Conflicti) to reduce the influence of outliers, or control for historical conflicts fought in neighbouring countries (not shown to save space).
12 Besley and Reynal-Querol (Citation2014) focus on conflicts between 1400 and 1700. We extend this periodization to include the eighteenth century. However, as we show ahead, our main results are robust if we restrict the historical conflict data to 1400–1700 () or 1400–1600 (Appendix Table A9).
13 These statistics use the conflict start variable to avoid double-counting, since some conflicts spill over from one century into the next.
14 This version of the paper uses updated fiscal data relative to the previous one (i.e. Dincecco et al. Citation2014). Thus, the results for fiscal capacity have changed somewhat, though the overall interpretation remains similar as before.
15 As we discuss ahead, the main results remain robust if we employ the UCDP/PRIO Armed Conflict Database (Citation2018) to re-compute this variable for alternative periodizations such as 1960–2014 (Appendix Table A10).
16 For ease of exposition, we henceforth refer to them as ‘continental’ fixed effects, even if North Africa is obviously a region rather than a continent.
17 For robustness, we use two non-fiscal alternatives in Appendix Table A2. The first is the government anti-diversion score according to the International Country Risk Guide (Citation2010). This measure averages the index scores in 1997 for the following categories: law and order, bureaucratic quality, corruption, risk of expropriation, and government repudiation of contracts. The second is the Brookings Institution’s state weakness score according to Rice and Patrick (Citation2008). In each case, the coefficient for Conflicti is positive and significant, while the coefficient for Conflicti × Africa is not significantly different.
18 For robustness, we estimate the specification in column 1 of for a similar sample as our main fiscal capacity variable (the civil conflict and fiscal capacity variables overlap for 106 out of 110 total observations). The results are very similar in magnitude and significant to the reported results (not shown to save space).
19 This result is consistent with the first-stage result in Dincecco and Prado (Citation2012), who find that greater wartime participation between 1816 and 1913 predicts larger fiscal capacity today.
20 Similarly, in Table A4 of the appendix, we account for ethnicity in two ways. First, we add the control for ethnic fractionalization from Alesina et al. (Citation2002). Second, we control for ethnic dominance in terms of whether a single language is spoken by at least half of the country’s population according to data taken from the Joshua Project. The main results in and are robust to both controls.
21 To the extent that the particular form of colonial rule influences post-independence interventions by past colonizers (e.g. the relationship between Benin and France), then the colonizer dummies should account for this possibility. For robustness, we control for two other features that help proxy for the autonomy of newly-independent nations in Appendix Table A5. To account for Cold War alliances, we control for vote affinity with the United States across roll-call votes in the UN General Assembly between 1946 and 1989 according to Strezhnev and Voeten (Citation2013). To proxy for leadership quality, we control for the share of years between 1946–2000 for which a nation’s leader is highly educated according to Besley and Reynal-Querol (Citation2011). The main results in and remain robust.
22 In Appendix Table A6, we account for the historical role of indigenous slavery in Sub-Saharan Africa in two ways. First, we control for the historical presence of the institution of indigenous slavery according to Bezemer et al. (Citation2014). Second, we control for log slave exports according to Nunn (Citation2008). For simplicity, we code this variable as zero for all non-African sample nations. If slavery was broadly more important in Asia or Europe relative to Africa, then continental fixed effects should help capture such differences. The main results in and are robust to both controls.
23 For robustness, we control for two other geographic variables in Appendix Table A7. Iliffe (Citation2007) suggests that border zones between forests and savannas in Africa could be prone to more conflict. To proxy for ecological diversity, we compute one minus the Herfindahl index of the different ecological zones in each country according to GAEZ (Fischer et al. Citation2000, Plate 55). To further control for natural resource wealth, we include average oil production between 1980 and 2012 according to the EIA (Citation2013). The main results in and are unchanged in both cases. Finally, to the extent that geography (e.g. terrain ruggedness) influences the type of colonial independence movement (Garcia-Ponce & Wantchekon Citation2017), then the geographic controls account for this possibility.
24 As an alternative way to control for unobservables, we include fixed effects for 14 macro-geographical regions according to the UN Statistics Division in Appendix Table A8. The main results in and remain qualitatively similar.
25 Our results suggest that conflict locations in Sub-Saharan Africa persist from the pre-colonial period to the present. Ideally, we want to know whether the same groups that fought in the past continue to fight today. To proxy for migration patterns, we control for a country’s share of foreign migrants in total population in 1960 as compiled by Ashraf and Galor (Citation2011). The main result in is robust to this control (not shown to save space).
26 In Appendix Table A9, we compute yet another alternative periodization for our main historical conflict variable: 1400–1600. The main results remain robust.
27 Similarly, in Appendix Table A10, we employ the UCDP/PRIO Armed Conflict Database (2018) to compute the main civil conflict variable for three alternative periodizations: 1960–2014, 1960–90, and 1990–2014. The coefficient for Conflicti × Africa remains positive and highly significant for both 1960–2014 and 1960–90, but just misses statistical significance for 1990-2014. Thus, the Cold War period appears to be particularly important to the civil conflict result.
28 The results are similar in magnitude and significance if we use all conflicts fought in Sub-Saharan Africa between 1850 and 1899 rather than only intra-African conflicts (not shown to save space).