Abstract
This article explores the relationship between economic development strategies and conflict outcomes in Africa, using a comparison of Côte d’Ivoire and Ghana. Findings suggest that development strategies that create contestation for resources along identity group lines, absent provisions for co-opting disenfranchised groups, increase the likelihood of communal violence, whereas development plans that transcend ascriptive social identities bear the greatest chance of generating long-term peace, irrespective of their short-term economic success. The author complements the comparative case study with evidence of intracase variation in Ghana and with original micro-level data collected in 2009.
Communal violence is often rooted in contests over scarce resources, but the underlying mechanisms remain debated. One perspective suggests that the instrumental interests of group members can make violence a worthwhile endeavor (Collier & Hoeffler, 2004). A second perspective holds that disputes over resources are a function of social and demographic patterns that create polarization (Montalvo & Reynal-Queral, 2005) or dominance (Cederman & Girardin, 2007). Undertheorized in that debate, however, is the role that political choice plays in shaping access to those resources and opportunities, thereby affecting patterns of contestation that can persist over a long time frame.
The goal of this article is to evaluate the impact of one important political choice—early economic development strategies—on long-term prospects for peace and conflict in Africa. How might the choices that African political leaders make to promote growth foster peace or increase the likelihood of conflict much later? The findings, in keeping with a literature on overlapping cleavages and horizontal inequalities (Stewart, 2008), suggest that economic development strategies that generate resource contestation along social identity lines can create a foundation for conflict and communal violence, whereas development plans that transcend ascriptive social identities bode well for long-term peace.
This article focuses on Africa for three reasons. First, most states in sub-Saharan Africa gained independence during approximately the same time period, around 1960, allowing for a neat comparison of cases with 50 years of postindependence hindsight. That time frame is long enough to evaluate persistent social patterns but also short enough and with a stark enough origin to evaluate the impact of economic development policies as distinct from significant historical differences. Second, national economic development strategies were central to the identity of independent African states. Reginald Green, writing in 1965, noted that “African states—even those with distinctly conservative governments—see national social and economic planning as a logical historical development from the national independence effort, now to be channeled into the rapid improvement of standards of life” (Green, 1965, p. 249). Thus, not only is the time frame suitable, but the implementation of clear economic development policies also makes African states an appropriate unit of analysis for this study. Finally, nearly half of African states have experienced communal conflict since 1990 (Marshall, 2005), suggesting both urgency in understanding the roots of that violence and also ample variation in outcomes.
Undoubtedly, myriad intervening factors can affect social and political relations over a 50-year period. The argument developed in this article is not that early economic development strategy represents the primary explanation for variation in peace and conflict since proximate political factors better explain the inspiration for partisans to actually pick up arms and fight. The research aims instead to highlight broader patterns of resource allocation and the emergence of group-level tensions that can make conflict more or less likely.
Nevertheless, the methodological challenges that arise in linking development strategies to future conflict outcomes are significant. For one thing, the causal process to be explained is a complex one: It involves economic policy choices that are also fundamentally political, groups whose interaction is shaped not only by policies but also by historical circumstances, exogenous shocks that alter long-term plans, changes in political leadership over time, and the creation of social tensions that may or may not turn violent. Furthermore, the independent variable of interest—economic development strategies set in motion in postindependence African countries—requires careful development. Some elements are components of national development plans, but others are shaped by broad principles, statutes, and relationships beyond those development plans. Given these difficulties, coding national-level variables for statistical analysis would not constitute an appropriate strategy.
Instead, the methodological strategy I employ is to rely on comparative case studies. I first construct an argument linking development strategies to conflict outcomes with reference to examples from across sub-Saharan Africa. I then test the argument with a comparative study of two cases—Côte d’Ivoire and Ghana—that holds several key factors constant but varies the explanatory variable of interest (economic development strategy). Neighbors on the Gulf Coast of West Africa, Côte d’Ivoire and Ghana were virtual matches in terms of climate, geography, agricultural crops, ethnic divisions, and north–south disparities. Yet whereas Côte d’Ivoire prioritized export-led growth reliant on foreign factors of production, particularly agricultural labor, Ghana adopted a path of state-led development and import substitution industrialization. Their experiences with peace and conflict have differed significantly since, which I evaluate over different periods in their postindependence history. I then conduct an additional test of the argument using evidence of intracase variation from Ghanaian traditional culture.
Case studies alone may be insufficient to establish strong causal claims, but, when coupled with findings from other studies, they add important inferential value to our explanations of complex processes. I complement the case comparisons with an analysis of original, micro-level data collected in the field in 2009. Those data provide contemporary evidence of cross-border and north–south differences regarding government discrimination and social discrimination that are consistent with the argument linking development strategy to conflict outcomes in Côte d’Ivoire and Ghana.
Apart from the empirical findings, this article makes four contributions. First, by evaluating an underlying source of group-level tensions, it suggests strategies that may help to foster peace, rather than explaining only those outcomes that result in conflict or offering strategies only for resolving conflicts already underway. Second, it leverages the period of African Independence to impose relatively reliable controls on an evaluation of long-term economic development consequences. A third contribution is the use of combined methodologies; the comparative case studies untangle complex national-level processes, whereas the statistical analysis of individual-level data helps to address observable implications of the broader argument using a different unit of analysis. Finally, this article treats economic development strategies as political-economic tools of instrumental elites, but it also identifies some systematic if unforeseen consequences of those strategies.
The Relationship Between Economic Development Strategies and Conflict
Most contemporary conflict occurs within nations, fought between communities divided along either regional or ascriptive identity lines, such as ethnicity, religion, or language (Heidelberg Institute for International Conflict Research, 2008). Literature describes civil conflict of this sort as a function of several competing factors, including “primordial” tensions (Horowitz, 1985), ethnic polarization (Esteban & Ray, 1999; Montalvo & Reynal-Queral, 2005) and dominance (Cederman & Girardin, 2007), the institutional weakness of the state (Fearon & Laitin, 2003), disputed control of lands (Moore, 1993; Straus, 2006), and economic opportunity (Collier & Hoeffler, 2004). Despite appeals to either participant “grievance” or “greed,” a common denominator among these explanations is contestation over the control and allocation of resources, such as oil, territory, or state jobs and revenues. Thus, insofar as resource allocation decisions are a component of economic development plans, a plausible link exists between development strategies and future conflict, irrespective of the proximate causal explanation.
One means to understand the mechanisms connecting development processes to conflict outcomes is to explore the long-term sociopolitical effects of early development strategies. Doing so provides two advantages. First, it facilitates a distinction between proximate tensions and the underlying policy decisions that gave rise to those tensions in the first place by favoring certain sectors, regions, or identity groups in ill-conceived ways. Second, it offers an opportunity not just to understand the contextual factors that drive conflict but also to categorize the choices political leaders make as being more or less favorable to peace, even 50 years later.
The logic is as follows. National economic development strategies shape access to resources, land, jobs, and opportunities, which are valued not just by elites who distribute them for political advantage, but most fundamentally by their constituents who desire improved standards of living. When those resources are distributed along ascriptive identity lines, however, one’s economic opportunities are perceived as a function of the social community to which one belongs. Disenfranchised or threatened groups thus perceive ascriptive differences or “ethnic hatred” as a reason for their unequal treatment. Since disenfranchised groups cannot easily remedy their predicament, tensions now defined along ascriptive identity lines linger and can easily erupt. Conversely, policies that avert the assignment of resources along identity lines disassociate economic inequalities from ascriptive identities, so the tendency to evoke “ethnic hatred” as an explanation for inequality decreases.
Strategies that avert the distribution of resources along ascriptive identity lines can take two forms. One is to ensure that economic divisions crosscut important social cleavages. Overlapping economic and social cleavages tend to conflate socioeconomic status with ascriptive traits, but policies designed to crosscut those social divisions allow members of dominant and minority social groups to share economic progress and blame (Stewart, 2008).
Genocidal violence in Rwanda was preceded by a failure to implement crosscutting economic policies. In a quest for organizational simplicity, Belgian colonial rulers applied strict categorization rules in an otherwise amorphous ethnic setting and selected the minority Tutsi for administrative favor (Hintjens, 1999). By independence in 1962, however, Belgian support for the Tutsis had waned, and the Hutu majority adopted a position as “sons of the soil” (Hintjens, 1999, p. 254). The early economic development strategy thus aimed to capitalize on preexisting economic strengths while redressing colonial-era favoritism: Tutsis were allowed to operate freely as merchants and in intellectual enterprises but otherwise faced strict quotas in public schools, civil service employment, and the army (Prunier, 1995). The result, unsurprisingly, was a socioeconomic fissure between the economic elite and the new Hutu political elite and their supporters, which broke along ethnic lines. An economic crisis in the 1980s was blamed on Tutsi business leaders, and increasingly scarce lands put added pressure on the “outsider” Tutsi population (Straus, 2006). By the early 1990s, socioeconomic tensions along ascriptive lines were high enough that Hutu supremacists successfully incited panic (Hintjens, 1999).
A second strategy that averts distribution of resources along identity lines prioritizes initiatives of national, rather than regional, scope. In Burkina Faso (formerly Upper Volta), leaders promoted the cultivation of cotton—long the country’s primary national export and only significant cash crop—throughout the country and especially in the relatively impoverished east and center–north, and an early 1980s revolution to reenergize development promoted national unity via an indigenized economy (World Bank, 2010). Despite swings in economic growth, Burkina has avoided animosities along ethno-regional lines. Nigeria, conversely, has experienced ethno-regional conflict throughout much of its postindependence history. Already saddled with ethnic divisions that overlapped with religious and regional ones, Nigeria’s government adopted a conservative plan with little income redistribution, a regional focus on directly productive economic sectors (particularly oil), and insufficient public spending (Green, 1965). According to Green (1965, p. 254), there was “no serious effort . . . to make the economic policies represent national goals in more than the vaguest sense.” Those policies exacerbated inequalities along ascriptive identity lines, fueling much of Nigeria’s sporadic conflict (Falola, 2009).
An issue that arises in the implementation of economic development strategies is the creation of new, class-based tensions. Those may be unavoidable in state-led as in free market economies, and in agricultural as well as industrial sectors. Provided that class divisions do not overlap with ascriptive cleavages, however, “ethnic hatred” cannot easily explain differences in economic opportunities, so negotiation and changes in policy remain possible.
Two factors can mitigate the challenges that arise from resource distribution along social identity lines. First is the use of patronage to co-opt groups slated as economic losers; simply put, economic development strategies can foster peace even in the context of ethno-regional divisions so long as some resources are dedicated to buying off disenfranchised groups. In Nigeria, the oil boom of the 1970s corresponded to the only sustained period of communal peace in the country. 1 Public expenditures increased, but more important, excessive patronage payoffs to both regional and ethnic leaders became a staple of the Gowon, Mohammed, and Obasanjo regimes in the 1970s (Frank, 1984). According to Frank (1984, p. 296), “[I]n the 1970s, the oil windfall made . . . dispersion of benefits much more rewarding.” Corruption in Nigeria has remained endemic, but the end of the oil boom undermined the use of patronage as a tool for fostering peace, and communal violence returned in the 1980s, 1990s, and 2000s.
Second, strategies that include investment in a robust civil society can temper the ill effects of resource distribution along identity lines. In Botswana’s political economy dominated by one party and one sector, education nevertheless received “rigorous attention” in national development strategies (Weeks, 1993, p. 50). A relatively open press since shortly after independence has also helped to strengthen economic interest groups despite underlying local–foreigner tensions (Nyamnjoh, 2002). By contrast, southerners in Sudan were long subject to exclusion from educational opportunities and from the right to organize peacefully (Deng, 1995), a strategy that helped to preserve oil revenues for the north but that also aggravated racial, ethnic, and religious differences.
In sum, economic development strategies that avert the distribution of resources along ascriptive identity lines, either by instituting economic policies that crosscut social cleavages or by promoting policies of national unity and scope, stand a better chance of fostering peace in the long-term. Even in the absence of such plans, development strategies can temper threats of communal violence by co-opting disenfranchised groups or fostering robust civic participation.
One alternative explanation for peace and conflict outcomes could be that poor economic conditions, rather than strategies that shape the distribution of resources, better explain the likelihood that an African country will be affected by communal violence. The evidence, however, fails to support that perspective: Some of Africa’s historically wealthier countries, such as Nigeria, the Democratic Republic of Congo, and South Africa, have been beset by major instances of communal conflict, whereas Benin, Burkina Faso, Malawi, and others among the poorest of the poor have avoided large-scale conflict. 2
Supporters of a second alternative explanation—that economic development strategies pale in comparison to the effects of social divisions predating postindependence policies—might point to Rwanda’s colonial history to support their claim. Certainly, the postindependence period in Africa cannot be viewed as a tabula rasa in terms of social group relations. The aim of this article is instead to suggest that economic development strategies can have important tempering or exacerbating effects on preexisting social divisions.
A third alternative explanation could be that communal violence is better explained by other postindependence political changes, such as the end of the cold war, differences in regime type, or contemporary political power grabs. However, civil violence in Africa followed a secular upward trend until the early 1990s and then began to decline (Fearon & Laitin, 2003), casting doubt on the cold war hypothesis. The states referenced in this study experienced either one-party systems or frequent military rule, and all have undertaken some form of democratic transition, suggesting that variation in conflict outcomes cannot be explained alone by regime type. Finally, power struggles help to explain the mobilization of ascriptive identities, but they do not provide insight into the inequalities that fuel those divisions in the first place.
In the complex process toward communal violence, other postindependence political choices may have some plausible effect on outcomes. In the section that follows, I provide evidence to suggest that one particular political choice, regarding economic development strategy, is an overlooked factor in shaping long-term trends in communal violence and peace in Africa.
Case Comparison: Côte d’Ivoire and Ghana
The methodological challenge to identifying a causal relationship between economic development strategies and conflict should by this point be clear. Notably, in examples I have provided to support the argument, the choices regarding development plans could be viewed as endogenous to preexisting factors on the ground: Rwanda’s independence-era government was influenced by colonial divisions, Nigeria’s by the discovery of oil in the 1950s, and Burkina Faso’s by the savannah conditions that made cash crops other than cotton impractical. A convincing test of the argument should therefore rely on cases where development strategy was instead a matter of ideological choice, largely exogenous to factors on the ground. If, under those conditions, different strategies implemented in otherwise similar environments lead to different conflict outcomes, the case for a causal relationship will be strengthened.
Côte d’Ivoire and Ghana fit that description. At the dawn of independence, Kwame Nkrumah, then prime minister of Ghana, and Côte d’Ivoire’s soon-to-be first president, Felix Houphouët-Boigny, whose ideologies had been shaped in distinct political milieu, engaged in a now-famous wager: The two countries would pursue fundamentally different development strategies, and a comparison of their economic standings a decade later would determine whose model was most effective. As Houphouët-Boigny stated shortly thereafter, “[Y]ou are now witnessing the start of two experiments . . . and in ten years we shall compare the results” (Woronoff, 1972, p. 13).
At the time of independence, both were sparsely populated countries: 1960 population density figures for Côte d’Ivoire and Ghana stood at 11 and 27 per square kilometer, respectively. 3 Coastal climates and environmental conditions in the southern regions favored agricultural production, and manufacturing was largely neglected. Both have remained among the world’s primary producers of cocoa and coffee throughout independence (Langer, 2008). Similarities also exist in terms of ethnic diversity: Ivoirian society comprises more than 60 distinct ethno-linguistic communities, and Ghana is home to at least 50. The northern regions of both countries are predominantly Muslim, are more sparsely populated, and have faced poverty gaps as high as 7 to 1, compared to the largely Christian, economically advantaged southern regions (Alpine & Pickett, 1993, p. 40). 4 Finally, Côte d’Ivoire and Ghana both began the independence era with one-party, weakly authoritarian governments (Foster & Zolberg, 1971).
The principal difference between the two aside from their development strategies comes in their colonial histories. Although the British policy of indirect rule created the sharpest practical distinction between Ghana and Côte d’Ivoire, the two were in fact more similar than different: Indirect rule in Ghana came only after a generation of direct rule that had already undermined local traditional elites, Christian missionaries were discouraged from confronting Muslim communities in the north of both countries, and dependence on the metropoles was tantamount for both (Woronoff, 1972). Thus, at the outset of independence (from England in 1957 for Ghana and from France in 1960 for Côte d’Ivoire), little distinguished these neighboring states in practical terms other than their pointed divergence in development strategy. The Ivoirian and Ghanaian cases have thus lent themselves to decades of comparative analysis, most of which has focused understandably on economic success (Alpine & Pickett, 1993; Crook, 1990; Foster & Zolberg, 1971; Langer, 2008; MacLean, 2010; Ryan, 1999).
In economic terms, although Ghana was slightly better off than Côte d’Ivoire in 1957, Côte d’Ivoire quickly established itself as the “miracle” in Africa, outpacing all regional counterparts (Daddieh, 2001). Ivoirian economic growth rates averaged over 7% annually from 1960 to 1980, whereas swings in the Ghanaian economy created anemic average growth over the same period (see Figure 1). Falling prices in agricultural commodities and rising debt sparked a 1980s crisis that the two economies weathered differently: Although Ghana reestablished annual growth rates in the 5% range that have persisted for 30 years, Côte d’Ivoire has endured a long period of moribund growth, particularly after the 1980 shock and from the late 1990s to mid-2000s.
Economic growth rates since independence
Economic strength per capita tells a more complex story. Figure 2 illustrates that whereas GDP per capita in Côte d’Ivoire rose to more than 5 times Ghanaian GDP per capita before declining sharply, it remains more than 50% higher than the relatively stable, slowly rising economic productivity per capita in Ghana. The patterns raise questions regarding change in well-being and conflict outcomes: Did Ghana’s development strategy hinder economic performance in the shorter term but also promote peaceful intergroup relations in the long term? Did Côte d’Ivoire’s government lose critical peacemaking advantages when per capita productivity declined?
GDP per capita since independence
Patterns in conflict offer preliminary evidence in support of those claims. Ghana faced minor coup-related conflicts in 1966, 1981, and 1983, during a two-decade period of instability (see Figure 3). The north–south divide never reached a level of violence, however, and the only notable communal conflict was a localized 1994 dispute between two small, northern ethnic groups (explained in greater detail below). In 2008, Ghana held its fourth consecutive peaceful, democratic election. In Côte d’Ivoire, Houphouët-Boigny’s rule (spanning 34 years, from 1960 to 1993) remained free from communal conflict, but the period of economic decline was followed by disputed elections, violence described in north–south, religious, and ethnic terms, 5 and a civil war from 2002 to 2007 that divided the country in half and left thousands dead. This article argues that the very strategy that led to early economic success in Côte d’Ivoire also created resource contestation along identity lines (and hence the potential for communal conflict), whereas the development strategy in Ghana deterred resource struggles among identity groups and contributed to intercommunal peace.
War-related deaths
Economic Development Strategy and Violence in Côte d’Ivoire
Strategy
The dominant economic development strategy in postindependence Côte d’Ivoire is inseparable from its long-ruling president, Houphouët-Boigny, who won 98% of the popular vote for president after August 7, 1960, independence (Woronoff, 1972, p. 91). Having learned pragmatism as a member of the French parliament (Foster & Zolberg, 1971), Houphouët-Boigny’s vision for independent Côte d’Ivoire demanded that economic development come first, prior to significant political change. He forsook the appeals of militant nationalism and instead favored economic sectors that offered the best path to efficient growth.
To specify the independent variable, the Ivoirian economic development strategy can be described in terms of five principal elements. The first was sectoral: Taxes, pricing, and subsidies favored a market-driven agricultural economy that mirrored the “plantation economy” model employed by the French since the 1930s. Although most newly independent African states sought to keep producer prices low to provide lower food costs to urban residents, the Ivoirian model guaranteed higher prices for producers, particularly on the major export crops of cocoa and coffee (Widner, 1993). Given environmental conditions in the north versus south of Côte d’Ivoire, the focus on commercial agriculture resulted in a concentration of opportunities in the southern, temperate regions.
Second, liberal immigration policies were an integral component of the Ivoirian economic strategy. Houphouët-Boigny took advantage of sparse lands and uncertainty over land rights in two ways. One was to increase migrant labor to meet the requirements of the commercial agricultural sector: Houphouët-Boigny ended the practice of forced labor but maintained a policy of attracting foreign workers with promises of land and opportunity in exchange for their labor in the commercial agriculture sector. The second was to solidify advantages for his closest constituents—commercial farmers and members of his Baoulé ethnic group (an Akan ethnicity). Thus, cultivators were favored over original land owners, Baoulé commercial farmers from the southeast gained control of cocoa and coffee land in the southwest, and local elites were pressured to welcome northern migrants (Woods, 2003).
Third, Côte d’Ivoire’s model for economic development depended critically on open markets and foreign interdependence, particularly vis-à-vis its former colonizer, France. Independence was, for Houphouët-Boigny, sufficient in “nominal” terms (Woronoff, 1972, p. 53). The French aided Côte d’Ivoire’s cash crop industry by guaranteeing an export market with sufficiently high price levels, and Côte d’Ivoire minimized its defense expenditures by virtue of a 1961 defense pact that surrendered external security to the French in exchange for economic policy considerations (N’Diaye, 2005). An investment code that guaranteed the repatriation of profits for foreign companies further established Côte d’Ivoire as an investment-friendly environment for foreign firms.
Fourth, despite a preference for open markets, the Ivoirian development strategy was less one of laissez-faire economics than “state capitalism” (Woronoff, 1972, p. 204): Government cooperatives and a large public sector complemented free enterprise to create and accumulate wealth in Côte d’Ivoire. At their peak, more than 200 state enterprises operated in Côte d’Ivoire, and early challenges from workers and students were met with increased salaries, renewed price guarantees for agricultural products, scholarships, and new schools (N’Diaye, 2005).
Finally, Côte d’Ivoire’s development strategy relied on the classification and co-optation of regional and social identity groups. “Ethnic arithmetic,” whereby the representation of social identity groups was formally balanced in government posts and civil service jobs, was an integral component of “Houphouëtism” (Boone, 2007). Potential political adversaries were co-opted with benefits: The armed forces were highly paid, and military and ethnic group leaders benefited from plush positions in ministries and parastatals (N’Diaye, 2005). Despite a policy that discouraged ethnic claims in favor of loyalty to the president, however, social identity groups were compartmentalized as foreign workers or national civil servants to define the distribution of patronage (Bakary, 1997). Maintaining stability in this environment of co-optation required rigid central control, which the Ivoirian government achieved using excessive secrecy and by undermining civil society organizations. Identity groups and their leaders thus enjoyed excessive spoils during the era of economic prosperity, but the trade-off came in the form of limited political participation and civic voice.
Patterns and Consequences
The Ivoirian development strategy, while generating impressive short-term growth, also created patterns of deficit spending that pushed debt levels as high as 231% of gross national product (Klaas, 2008, p. 113), as well as excessive reliance on volatile cash crops. Patterns of subnational division also emerged: Those accumulating lands from entrenched southern ethnic groups were perceived not just as foreigners but also, by virtue of overlapping identity types, as religious and ethnic associates of the north, and thereby as a social threat.
The consequences of those patterns can be broken into two periods. During the Houphouët-Boigny era, social tensions were managed successfully and peace prevailed for over 30 years. Agriculture, land tenure, and civil service policies had the social effect of categorizing citizens along identity lines, but they also had the short-term economic effect of stimulating rapid growth. Thus, despite ethno-regional competition for lands and jobs, development policies generated enough revenues (via either commercial agriculture or borrowing) to appease major interest groups. In this sense, the economic vision of Houphouët-Boigny can be viewed as a short- to medium-term success in terms of collective peace, as tensions were offset by astute cultivation of patronage relationships. During the first period of Ivoirian independence (from 1960 to the 1980s), Houphouët-Boigny’s development plan both invited potential threats to collective peace and tempered those threats through the co-optation of potential challenges from diverse identity groups.
The second period began with the downturn in commodity prices. When cocoa rents evaporated, the Ivoirian government was sapped of its ability to appease politically disenfranchised identity groups who represented the bulk of migrant labor. Fearing competition from those laborers, the Baoulé and Bété formed a southern, indigenous opposition to the foreign presence in Côte d’Ivoire (Woods, 2003). Scaled-back public spending forced the government to make hard choices, and protests emerged from communities in the north when the state abandoned plans to enable the northern economy to catch up to that of the south (Gyimah-Boadi, 1994). Having had little experience with political participation, however, those communities sought refuge behind social identity labels—primarily northerners, Muslims, northern ethnic groups, and foreigners—and the state’s political institutions were ill equipped to handle the power vacuum that developed after Houphouët-Boigny’s rule.
On Houphouët-Boigny’s death in December of 1993, power was transferred to the speaker of the National Assembly, Henri Konan Bedié, a Catholic Baoulé like Houphouët-Boigny, who then won a disputed election in 1995. 6 Lacking the resources to co-opt out-groups, Bedié introduced the ideology of Ivoirité, which created explicit advantages for those born in Côte d’Ivoire of southern Ivoirian heritage by distinguishing “pure” Ivoirians from those of foreign descent (Le Pape, 2003). That initiative prevented a northern Muslim, Alassane Ouattara, from running for president, but the social effects ran much deeper: At the communal level, northerners, immigrants, and Muslims all faced discrimination in access to public services, harassment at police checkpoints, and seizure of lands. 7 Disputed elections in October 2000 perpetuated the identity divide, and newly elected president Laurent Gbagbo used Ivoirité to buttress a civil war that lasted through much of the 2000s and whose effects remain a concern.
In sum, the Ivoirian development model fostered impressive short-term growth at the expense of social equality, by distributing resources and opportunities along identity group lines. That policy succeeded during times of largesse, but absent resources to co-opt politically disenfranchised and threatened identity groups, policies that categorized inhabitants by identity undermined collective peace.
Economic Development Strategy and Peace in Ghana
Strategy
Ghana obtained independence from Britain on March 6, 1957, with little animosity but in a clear culmination of the colonial relationship; Prime Minister Nkrumah’s own stated mission was to “prove to the world that Africa could conduct its own affairs” (Woronoff, 1972, p. 34). In 1960, confronted with a multiethnic, regionally divided country similar to Côte d’Ivoire, Nkrumah led a shift from multiparty democracy to a one-party republic guided by the Convention People’s Party, or CPP (Gyimah-Boadi, 1994). Nkrumah was appointed president for life in the same year that Houphouët-Boigny was elected president in Côte d’Ivoire.
Ghana’s economic development strategy comprised four main components. First and foremost was the structural shift to state-led import substitution industrialization, which entailed an expanded public service, heavy taxation of exports, and an overvalued exchange rate. An explicit aim of Ghana’s national development plan of 1959–1964 was the construction of “not less than 600 factories” in a time span of 5 years (Woronoff, 1972, p. 173); the following development plan (1964–1970) declared as a policy objective the advancement of industries in chemicals, electronics, and machine tools (Huq, 1989, p. 128). Unlike Houphouët-Boigny’s strategy of relying on foreign know-how, the Ghanaian dirigiste approach was to foster learning by doing and to avert subnational tensions with policies that influenced sectors of the economy on a national scale (Alpine & Pickett, 1993). State-centered development has shaped the Ghanaian economy over much of its postindependence history, particularly under Nkrumah (1960–1966), Acheampong (1972–1978), and parts of the Rawlings’ regimes (1979, 1981–2000).
Second, to finance the development of industry and infrastructure, the Ghanaian economic development strategy relied on domestic agricultural support. Commercial production for export came almost exclusively from the cocoa industry in Ghana, which, in the early independence era, accounted for 40% of the world market (Woronoff, 1972, p. 173). Like that of Côte d’Ivoire, Ghana’s strategy entailed a fixed price for producers with surpluses controlled by the Cocoa Marketing Board (CMB) that would finance development. Unlike in Côte d’Ivoire, however, prices paid to producers in Ghana were exceedingly low: The CMB kept more than 40% of profits, and by the 1980s farmers were receiving less than 20% of the price of cocoa sold on the world market (Alpine & Pickett, 1993, p. 71). Proceeds were explicitly directed toward industrialization projects rather than reinvestment in commercial agriculture.
Third, industrialization required a supply of human capital, so Ghana’s economic development strategy included an investment in civic strength. That investment took two forms. One was a focus on building the educational capacity of the state: The 1961 Education Act set universal primary education as a target well in advance of similar measures in other African states, and its earliest beneficiaries were from northern areas disadvantaged during colonial rule (Tsikata & Seini, 2004). In 1965, school enrollment reached 55%, far above Côte d’Ivoire’s 29% rate (Crook, 1990, p. 29). The second component of civic development was the encouragement of civil society groups (Tsikata & Seini, 2004). The rationale for developing a strong civil society was to promote diverse economic interests with national scope and to prevent ethno-regional attachments with little value to national development (Gyimah-Boadi, 1994, p. 126). Groups like the Trades Union Congress, the Farmers’ Council, the Ghana Youth Organization, and the Ghana Bar Association flourished; some were co-opted by the CPP, but others maintained meaningful independence (Gyimah-Boadi, 1994).
Fourth, Ghana’s development strategy depended on the “Ghanaianization” of economic production. Colonial bureaucrats were quickly replaced by Ghanaian administrators (Foster & Zolberg, 1971, p. 188). Land laws also favored Ghanaians, but particularly those in the commercial classes: the influence of traditional chiefs was constrained in the fertile south, and ownership rights were granted to large scale cooperatives and to smaller commercial farmers in the CPP, instead. Losers were local subsistence farmers and “stranger farmers” who migrated to Ghana for its ecological and natural advantages. The Aliens Compliance Act of 1969 further limited the role of foreigners in the Ghanaian economy: More than 200,000 foreign laborers without resident permits, many of whom migrated from Upper Volta and other neighboring countries to work on cocoa farms, were permanently expelled (Gyimah-Boadi, 1989).
Patterns and Consequences
Ghana’s strategy of import substitution industrialization, exploitation of the cocoa industry, investing in civic strength, and “Ghanaianization” fostered three sociopolitical patterns. First, deficit spending created a debt problem closely linked to ensuing political swings: At least seven regime changes, four of which resulted from coups, took place during the period corresponding to Houphouët-Boigny’s stable political rule in Côte d’Ivoire. A pattern of rising class interests also emerged, but competing wage labor, farmer, and bourgeoisie interests transcended any particular region or ascriptive identity group (Adu Boahen, 1997; Gyimah-Boadi, 1989; Tsikata & Seini, 2004). Tsikata and Seini (2004, p. 17) note that at independence, “[e]thnicity took a backseat. Instead, class and status became the cleavages along which society was divided.” That pattern continued through the Rawlings’ era, when ideological commitments relegated ethnic concerns to a secondary position (Adu Boahen, 1997). 8 Finally, experienced civil society groups acted as sources of pressure on policy makers (Adu Boahen, 1997).
The consequences of Ghana’s economic development strategy on peace and conflict (this study’s dependent variable) may also be evaluated in two periods. First, the strategy contributed to a period of political uncertainty. The economic hardship of the first two decades of independence fueled skepticism of the industry-driven model, and the political coups that followed not only were rooted in disagreements over economic policy but also were in some cases violent in nature (refer back to Figure 3). Nevertheless, no violent social divisions—whether north–south, Ewe–Ashanti, or otherwise—emerged during the period of economic stagnation. Co-optation of identity groups through patronage, however, is an unlikely explanation: The state in Ghana never enjoyed the degree of largesse needed to make that strategy a plausible one as it was in Côte d’Ivoire. Instead, development strategies that transcended identity group divisions, promoted national unity, and stimulated active political participation kept Ghana free of collective violence despite political unrest.
The period of political uncertainty is consistent with the argument set forth in this article. Ghana’s early economic development strategy entailed political costs, and a strong civil society responded with political demands. Critically, however, those challenges could have taken shape along social identity lines when the debt problem undermined prospects for growth, but they did not. Although the state in Côte d’Ivoire suppressed underlying identity group interests that ultimately erupted in the absence of patronage, Ghanaian interest groups—defined largely along ideological and class lines rather than ethno-regional ones—expressed grievances in terms of class, labor, and economic interests. As a result of power swings and alternation in economic policy, those interest groups experienced political life in and out of the favor of government.
Furthermore, the indigenization of the Ghanaian economy tempered identity fissures during the period of political uncertainty. Land allocation policies created opposition to the state from the farming class but that tension was not specific to any particular region or ethnic group. Restrictions on ownership of farming lands, along with the expulsion of foreign workers, also helped to avert potential conflict between those considered native Ghanaians and migrant workers of foreign descent (with overlapping ethnic and religious differences). Thus, restricting opportunities to foreigners in the 1960s, a policy that could be viewed not only as economically counterproductive but also as tinged with xenophobia, may have saved Ghana the challenge of managing one aspect of social group relations during an unstable political era.
The second, post-1980s period has been marked by economic interests distinct from intractable identity ones. Ghana’s interest groups negotiate their positions in a political sphere, and as a result, political setbacks are not viewed as permanent. Broad-based austerity measures adopted in the 1980s proved effective in remedying the debt burden, and democratic norms emerged (Langer, 2008). Following his 1992 transition to civilian rule, Rawlings served the maximum two terms and then saw his National Democratic Congress (NDC) party defeated by John Kufuor of the New Patriotic Party (NPP). Kufuor served two terms, and in 2008 the NPP lost the presidency to the NDC and John Atta Mills. Thus, despite the radical ideology of the Nkrumah era, Ghana’s economic development strategy softened the appeal of ethnic divisions and instead inspired a preference for pragmatic economic policy, negotiation, and democratic peace. Observers now call Ghana “an attractive investment destination” (Anaro, 2012) and “among Africa’s most peaceful and progressive” countries (World Bank, 2008).
Intracase Variation in Ghana
Evidence from Ghanaian traditional culture provides some intracase variation in support of the argument. An observable implication is that if a parallel system of governance in Ghana did not follow the development model that averted ascriptive identity tensions but instead distributed resources along identity lines, one would expect to find a greater likelihood of conflict in that setting. In some northern Ghanaian localities, a strong traditional or customary culture can be viewed as a parallel form of governance: Matters such as formal education and infrastructure are the purview of the state, but other concerns, such as neighborhood disputes, land use, and the informal economy, tend to follow traditional guidelines. In particular, the state officially relinquished control over the distribution of lands in northern Ghana in 1979, returning control to traditional leadership (Tonah, 2009). Thus, if the development model in the traditional context differs from the economic development strategy of the state, variation in peace and conflict outcomes could also be anticipated.
Ghana’s major episode of communal conflict, the 1994 Guinea Fowl War, can be explained according to this logic. The violence began with a market dispute over the price of a guinea fowl, between two men who happened to be of the Konkomba and Namumba ethnic groups. When the disagreement ended with the death of the Namumba man, violence quickly engulfed the neighboring communities, resulting in an estimated 1,000 casualties (Centre for Systemic Peace, 2010).
Traditionally, the Konkombas had little hierarchical structure and as a result lost land control to chieftaincy-based groups like the Namumbas, despite claims of indigenous rights (Assefa, 2001). As population pressures mounted, Konkombas began first to cultivate land within the Namumba territory and eventually to stake renewed claims to those lands. Traditional development policies in the region, however, assigned access to land (a critical development resource) along social identity lines: Customary chiefs of Namumba ethnicity granted land tenure rights only to fellow Namumbas. Those policies contributed to an environment of ascriptive identity tensions that ultimately erupted over a seemingly minor dispute.
The Guinea Fowl War was limited in scope by the fact that the Konkombas and Namumbas are relatively minor ethnic groups located in a confined area in northern Ghana. As a result, the potential for widespread civil conflict with “indigenous versus foreigner” labels was much weaker than in neighboring Côte d’Ivoire. Nevertheless, the conflict may be viewed as a microcosm of the challenges facing Côte d’Ivoire, only in a custom-based context. Traditional development policies that assigned opportunities along identity lines, absent alternative resources to appease the disenfranchised group, contributed to the onset of communal violence. The Ghanaian state, meanwhile, beginning with Kwame Nkrumah’s industry-driven approach that promoted national unity, took a different path and has averted widespread civil conflict.
The Sociopolitical Climate Fifty Years Later: Data and Analysis
Those descriptive accounts address the macro relationships between broad economic strategy and variation in conflict outcomes over a 50-year period. In this section, I provide a contemporary, micro-level analysis of how the management of economic and identity group interests affects perceptions among residents in the two countries. The data in this section address perceptions of government discrimination and social discrimination.
Research Design
To test individual-level perceptions in Côte d’Ivoire and Ghana, the study relies on survey data collected in 2009. 9 One shortcoming is that the study could not include all localities, primarily because of the challenge of multiple languages. Instead, a “most different systems design” is employed (see Collier, 1993): Two regions were selected, one in the north and one in the south, in each country (see Figure 4). In view of the close link among ethnicity, religion, region, and development, the design ensures stark variation in research settings: Savanes in northern Côte d’Ivoire is home predominantly to the Muslim northern Mandé and Senoufo and remains poorer than the south. In southern Côte d’Ivoire, Sud Bandema is primarily the territory of the Christian Krou and Akan, although immigration policy has created a demographic mix. In the northern region in Ghana, the predominantly Muslim Mole-Dagbani predominate, and the Christian Fante (an Akan group) predominate in the wealthier central region. In each region, participants were drawn from the provincial capital and from up to four rural villages.
Research regions in Côte d’Ivoire and Ghana
Randomly selected participants provided background information and were asked a battery of questions to gauge sociopolitical perceptions.
10
Respondents were asked,
How often are you treated unfairly regarding services from public agencies?
Responses were coded on a 5-point scale from never to very frequently, nearly every time I deal with them. The question was phrased as a matter regarding public agencies rather than “the government” to avoid associations with the political party in power and to focus instead on the state and its bureaucratic arms. Second, participants were asked a question to gauge perceptions of social discrimination:
How often do you face discrimination from other people or groups in the country?
Answers were again coded on a 5-point scale from never to very frequently.
Findings
With respect to treatment from public agencies, overall levels of discrimination are comparable in Côte d’Ivoire and Ghana (28.2% vs. 23.8%; in a difference-of-means test, p = .44). 11 The real story, however, appears in regional discrepancies. First, the moderate cross-national difference in perceptions of government discrimination comes entirely from the northern regions: As Figure 5 illustrates, respondents in southern Côte d’Ivoire actually report less government discrimination than their counterparts in southern Ghana, but northern Ivoirians appear about 30% more likely than northern Ghanaians (39.8% vs. 29.4%, p = .117) to perceive discrimination in public agencies. More important, respondents in the north of Côte d’Ivoire are more than twice as likely to feel regular discrimination from the government as are respondents in the south of Côte d’Ivoire (39.8% vs. 16.7%, p = .001).
Perceptions of government discrimination
The pattern of higher government discrimination in northern versus southern Côte d’Ivoire invites two possible explanations. First, respondents in the north may have perceived the short time horizons of the Forces Nouvelles rebel group (which took control of the north during the Civil War) to inspire the exploitation of its own constituency. More likely, respondents considered the question to be in reference to agencies associated with the official Ivoirian government, which lost formal control of the north in 2002 but whose bureaucrats are viewed as responsible for implementing the policy of Ivoirité that set the tone for the recent conflict and ongoing tensions.
With respect to social discrimination, overall figures are slightly higher in Côte d’Ivoire (34.3% vs. 26.6%, p = .198). Again, however, disaggregation by region reveals important distinctions between the two countries. Although approximately one third of northerners feel social discrimination in both Côte d’Ivoire and Ghana (35.2% and 32.9%, respectively), people living in southern Côte d’Ivoire are much more likely than those living in southern Ghana to report such discrimination (33.3% vs. 20.2%, p = .046). In fact, the level of social discrimination in the south of Côte d’Ivoire is nearly as high as in the north.
These findings raise an additional question: Who, precisely, feels socially discriminated against in southern Côte d’Ivoire? The answer, not surprisingly, is those of northern and foreign descent: More than half of respondents in the south of Côte d’Ivoire (52 out of 99) reported being born in a foreign country or in the north of Côte d’Ivoire, and those respondents were 4 times more likely than indigenous southerners to report social discrimination (22.0% vs. 5.2%, p = .001). Conversely, just 16 out of 88 respondents in the north reported being of southern heritage, and they were no more or less likely than their nonmigrant counterparts in the region to feel social discrimination. These figures are in keeping with the country-level analysis presented above: Development strategy in Côte d’Ivoire fostered heavy migration southward and divisions along identity group lines, and it suppressed opportunities for political expression. As a result, generations of northerners brought to the south by economic opportunity now feel socially disenfranchised according to their ethnicity, religion, or place of origin, and little trust exists between themselves and “pure” southern Ivoirians.
The bivariate relationships presented above could mask alternative explanatory factors that happen to be correlated with region and migrant status. To guard against the possibility of spurious relationships, Table 1 presents results from logistic regressions with dichotomous dependent variables that measure experience with discrimination. The results, reported as marginal effects coefficients, 12 support the bivariate findings: Column 1 shows that being Muslim makes respondents more likely to report government discrimination regardless of where they live in Côte d’Ivoire, but simply by virtue of living in the north, holding other socioeconomic and demographic characteristics fixed, respondents are approximately 10% more likely to report government discrimination. Column 2 confirms that migrant status is the most important explanation for reporting social discrimination. 13
Determinants of Perceived Discrimination in Côte d’Ivoire
Logit regressions with coefficients reported in marginal effects terms and standard errors in parentheses. Standard of living is measured on a 5-point scale, with points for access to necessities, ownership of assets, and occupation type. Migrant status is coded 1 if the respondent was born over 250 km away or in a different country, 0 otherwise. Ethnic group fixed effects includes dummy variables for membership in the Dioula/Malinké, Senoufo, Mossi, Bété, Baoulé, Dida, or Yacouba groups. The omitted category is other ethnic group.
p < .05. **p < .01.
I cannot rule out the possibility that the violence in Côte d’Ivoire caused these perceptions, rather than the perceptions of social division contributing to violence. The results from the survey data do not by themselves tell a causal story. Nevertheless, the individual-level results are consistent with the case comparisons linking economic strategy to conflict outcomes: In Côte d’Ivoire, policies that created contestation for land and resources along identity lines generated tensions that have left northerners skeptical of the government and foreigners and indigenous Ivoirians skeptical of one another. In Ghana, economic policy mitigated against ethno-territorial tensions, such that although localized disputes have affected the north and northerners continue to face greater discrimination, north–south conflict has been averted.
Conclusion
The aim of this study has been to suggest a link between economic development strategies and future experiences with peace and conflict. However, the dynamic and long-term path from cause to effect is fraught with evidentiary challenges, many of which could not be overcome simply by adding more national-level data. The strategy was thus to present a comparative case study suggesting a plausible link and to complement that analysis with individual-level data consistent with the case study findings. Any observational study such as this one will have its limits, but the exercise hopefully elucidates important mechanisms that underlie more proximate explanations for national-level variation in conflict.
Economic development strategies in Côte d’Ivoire and Ghana both entailed excessive reliance on the cocoa industry and on public spending that generated balance of payment problems. In other regards, their development strategies parted ways significantly: Côte d’Ivoire embraced an open-market economic model that relied heavily on migrant labor and then classified groups along ethnic and regional lines for the distribution of patronage. That strategy fostered a tenuous peace but also generated contestation for resources according to ascriptive traits. Conversely, Ghana’s economic development strategy fostered a class divide that transcended regions and supplanted ethno-territorial tensions. Coupled with the development of a robust civic culture, Ghana weathered poor economic growth and emerged as a stable, democratic, and peaceful state.
Poor economic conditions alone cannot explain communal conflict outcomes—places like Côte d’Ivoire and Nigeria are still much wealthier than their more peaceful neighbors. More proximate political factors certainly play a role in the decision for partisans to engage in communal violence, but the evidence presented here suggests that economic development decisions shape social group relations in important ways, setting the stage for either negotiated policy outcomes or entrenched identity tensions.
This comparative study highlights a number of important considerations for averting collective violence. First, redistributive economic policies will provide advantages to some at the expense of others, but states stand a better chance of cultivating peace if those advantages crosscut ascriptive identity lines. In Ghana, redistribution sparked urban–rural challenges, but those challenges did not overlap with identity group interests that could exacerbate social inequalities, as was the case in Côte d’Ivoire. Second, development strategies that foster perceptions of group-level discrimination can provoke tendencies toward collective violence even among the privileged. Côte d’Ivoire’s indigenous southern population enjoyed geographic advantages but perceived the immigrant labor community as a threat. Similarly, migrant laborers gained access to land ahead of nonfarming southerners, but they could not escape the label of “foreigner” that would eventually underpin a civil war. Finally, development strategies that prioritize civic strength and give voice to economic interest groups may effectively downplay ascriptive identity tensions. If the corresponding economic and class issues are managed adroitly, as they were in Ghana after a period of political uncertainty, reductions in the potential for communal violence may be realized. Thus, initiatives that include diverse social and economic stakeholders, such as the Vision 2030 plan currently in development in Liberia, bear the greatest chance of generating long-term peace, irrespective of their shorter-term economic success.
Footnotes
Acknowledgements
I would like to thank William Ascher, Emmanuel Gyimah-Boadi, Natalia Mirovitskaya, Tyson Roberts, and anonymous reviewers for helpful insights and suggestions. Halsey DiSerio, Karim Tondossama, Issahaku Al-Hassan, and Mildred Wryter contributed excellent research assistance.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The National Science Foundation and the Pacific Basin Research Center provided financial support.