AFRICAN POVERTY
Duncan Kennedy*
Abstract: African extreme poverty is probably a function (although not solely) of the balkanized post-colonial geopolitics of Africa. It is also probably a function (although not solely) of the income distribution generated by a typically perverse African political economy, through its effect on the allocation of resources to development. As between these two causes, the second is probably much the more important. This reinterpretation puts considerably more of the blame for African poverty on the Western great powers than does the “poverty trap” analytic that is a common contemporary way of thinking about the African economic situation.

INTRODUCTION This essay, which really is an essay rather than a sustained scholarly encounter with the problem, proposes an alternative to the “poverty trap” analytic for understanding extreme poverty in sub-Saharan Africa. The poverty-trap idea is well instantiated by the following quotation from Jeffrey and Lisa Sachs, and it is common among liberal Western commentators on African economy. For the world’s poorest people, daily life is a struggle for survival, with millions of impoverished people each year losing that struggle to famine, disease, environmental catastrophes, and violent conflicts that arise in conditions of extreme deprivation. . . . One basic point, not always remembered, is that impoverished countries lack their own budgetary resources needed to supply vital—indeed life-saving—services such as primary healthcare or support for smallholder famers. The poor are thereby trapped. The lack of public services leads to hunger, poverty, and disease, while the poverty means that the tax base of government is too small to support public policies to alleviate hunger, poverty and disease. Foreign assistance is then needed
* Carter Professor of General Jurisprudence, Harvard Law School. Thanks to Alice Amsden, Raymond Atuguba, Mekonnen Firew Ayano, Abdul Baasit Abdul Aziz, Emmanuel Bagenda, Trudy Dako, James Hackney, David Kennedy, Jeremy Perelman, Holger Spamann, David Trubek, and Lucie White. Errors are mine alone. The author retains the copyright in this article and authorizes royalty-free reproduction for non-profit purposes, provided any such reproduction contains a customary legal citation to the Washington Law Review.

205

206

WASHINGTON LAW REVIEW

[Vol. 87:205

to break the vicious circle.1 My argument has four parts: In the first, I discuss the impact of the “balkanization” of sub-Saharan Africa on its economic development. I argue that although precise answers are impossible, it seems likely that balkanization has been a significant factor in the poor economic performance that has relegated about half the population to “extreme poverty” sixty years after the Western powers reluctantly acceded to African demands for independence from colonial rule. The second part of the essay argues that in the great majority of African states, the postindependence elites, under the strong influence if not outright control of the coalition of anti-communist Western great powers, put in place a form of political economy that precluded the massive allocation of surpluses to development that would have been necessary if Africa was to develop in a way that more rapidly reduced extreme poverty. The third part suggests that although economic theories of self-sustaining economic development are in disarray, the apparent take-off of the Chinese, Indian, and Brazilian economies pushes us to look again at the “poverty trap” as an explanation of African poverty. The fourth part uses two counterfactual hypotheticals to compare balkanization and perverse political economy as possible alternative explanations, and suggests that as between the two, the second was more important than the first. I. A. BALKANIZATION AS A CAUSE OF AFRICAN POVERTY Definition and Causes of Balkanization

By “balkanization” I mean to denote the division of Africa into dozens of independent states, corresponding to the borders consolidated by the colonial powers at the end of the nineteenth century. It is an important aspect of this order that the boundaries arbitrarily combined and divided major ethnic groupings, and that the newly independent countries varied enormously in size, population, natural resources, and natural advantages such as access to deep-water ports or navigable rivers. The combination of resources, population, and access features for any given state was largely random. There is no single answer to the question of why balkanization occurred. But a key factor to consider is that the colonizing states, allied
1. Jeffrey D. Sachs & Lisa E. Sachs, Foreword to STONES OF HOPE: HOW AFRICAN ACTIVISTS RECLAIM HUMAN RIGHTS TO CHALLENGE GLOBAL POVERTY, at xi–xiii (Lucie E. White & Jeremy Perelman eds., 2011); see also Jeffrey D. Sachs et al., Ending Africa’s Poverty Trap, 2004 BROOKINGS PAPERS ON ECON. ACTIVITY 117.

2012]

AFRICAN POVERTY

207

with the United States and with local African elites, chose to exercise their sovereign legal powers under public international law to configure independent Africa as the Balkans. The post-colonial regime of public international law explicitly declares the continuing legitimacy of the boundaries set at independence. With very few exceptions, the French have maintained a significant military presence across their ex-empire, supporting dictatorial regimes under an implicit deal that discourages wars of expansion, but provides defense against aggression by their neighbors (often ex-British colonies), in return for continued French economic dominance. The ex-British colonies have had considerably more autonomy, but their dependence on Western aid and expertise has militated against “adventures.” The very fragility generated by the random boundaries and internal heterogeneity of the balkanized states suggested to African leaders that wars of expansion might well “open Pandora’s box,” or “upset the applecart,” with highly unpredictable results. In their sub-continental institutions they have repeatedly affirmed the importance of respecting the boundaries, however open to critique they may be as colonial impositions. Of course, Africa was hardly unified before it was colonized, and the colonial powers did not set out to fragment it—that was just the unintended consequence of their relentless pursuit of their understanding of their own interests. B. Consequences of Balkanization Balkanization seems likely to have been important on many different levels. First, in the period between independence and 1980, orthodox development theory, whether communist, socialist, liberal or merely nationalist, called for import substitution industrialization (ISI), with accompanying infrastructure requirements. Across the Global South (e.g., India, Brazil, Egypt, and Indonesia), the state was supposed to take a very active if not a completely dominant role in executing this strategy. The rationale was that underdeveloped markets, co-existing with subsistence production in an only partially commodified economy with a limited entrepreneurial class, under-rewarded the investment that would have been necessary to spark privately driven development. The ISI strategy aimed to counter the tendency towards stagnation through a host of policies, including: tariff protection for infant industries; quantitative restrictions on imports; manipulation of exchange rates and various kinds of control of capital movements with the same objective; marketing-board monopsonies permitting the taxation of agricultural exports to finance the new industries;

208

WASHINGTON LAW REVIEW

[Vol. 87:205

development banking intended to make capital available to new enterprises at below market rates; the large-scale state provision of public utilities; the creation of mixed public–private or “parastatal” enterprises; and from time to time the regulation of prices, particularly food prices and wages in the “modern” sector.2 (The Asian Tigers alternative, pursuing both ISI and export promotion right from the beginning, also required a dominant state role, although this aspect of their strategy has been underplayed in the neoliberal account.3) Balkanization meant that the size of the internal market in most of the four dozen sub-Saharan African states was much too small to permit the economies of scale necessary to make industrial production competitive with imports, even with heavy tariff protection. And even in the larger countries, it was impracticable to undertake more than a small sample of industrial projects. (Of course, a continental free trade zone could have cured that, but it is hardly surprising in retrospect that none was destined to emerge.) Balkanization occurred at the end of a colonial order that had invested very little in training native cadres with the capacity to manage a state, let alone to manage a developmental state pursuing ISI. The duplication of state structures meant that there were dozens of central banks and dozens of finance ministries and dozens of development planning agencies, but in 1960 I doubt that there were fifty African PhDs in economics in all of sub-Saharan Africa. (Of course, foreign experts were available if the price and living conditions were right, but it is hardly surprising in retrospect that they were more often flying in and out as very short-term consultants or negotiating from the side of Western commercial interests than serving permanently or negotiating from the African side.) From the point of view of state-initiated development, material resources located in one state might have been profitably combined with populations and locations in other states, but balkanization made such combinations subject to prohibitive transaction costs. Balkanization also meant that the surpluses available for taxation or other modes of appropriation were dispersed across jurisdictions, and so could not be aggregated and redeployed according to a large-scale plan. An interesting categorization of African states as “landlocked,
2. The intellectual and political genealogy of ISI is complex and interesting. For a first attempt to situate it in its global context, see Duncan Kennedy, Three Globalizations of Law and Legal Thought: 1850–2000, in THE NEW LAW AND ECONOMIC DEVELOPMENT: A CRITICAL APPRAISAL 19 (David M. Trubek & Alvaro Santos eds., 2006). 3. See generally ALICE H. AMSDEN, THE RISE OF “THE REST” (2001).

2012]

AFRICAN POVERTY

209

resource-scarce,” “coastal, resource-scarce,” and “resource-rich,” suggests that each faces specific challenges and opportunities.4 This kind of analytic is useful, but it is important, it seems to me, that it presupposes the balkanized order of states: combining landlocked and coastal resource-poor states with resource-rich ones would eliminate most of the “challenges,” while multiplying the opportunities. Even the “Dutch disease” challenge to the resource rich states, i.e., the tendency to replace development policy with struggles over the distribution of the rents from the resource, would look different if the riches characterized a region of a country with only moderate overall resource endowment, rather than dominating a small national scene. I think we can blame balkanization as a factor (no more than a factor, yet still worth assessing) in the political instability of the region, with accompanying tendencies of groups with access to state power to act in brutally self-interested and exploitative ways, as indicated by the figures for deaths from war and famine cited below. I don’t think there’s much disagreement that these tendencies have been destructive of the chances for development across the sub-continent (not least by discouraging, not just foreign investment, but investment of all kinds).5 The links between balkanization and instability are of course speculative, but still plausible to my eye. Small African states are vulnerable to influence by much larger developed states (obviously including the ex-colonial power, supposing it is a “great power”), and their whole economies are much smaller in straight dollar terms than the balance sheets of the financial, commercial, and industrial multinationals on which they are dependent in many different ways. The rewards of alliance with these rich actors (and the punishments for defying them) may have often seemed more plausibly available than those that might have been derived from even a successful economic nationalism. Within small states, polarized ethnic or religious divisions can lead to winnertake-all situations, and the anticipation that power is likely to shift decisively can make it seem rational for temporary winners to indeed take all while the taking is good. Of course, none of this is at all decisive. The multiplicity of African states might conceivably have functioned as so many “laboratories of
4. 1 THE POLITICAL ECONOMY OF ECONOMIC GROWTH IN AFRICA, 1960–2000, 3–75 (Benno J. Ndulu et al. eds., 2008); see also John Luke Gallup, Jeffrey D. Sachs & Andrew D. Mellinger, Geography and Economic Development, in ANNUAL WORLD BANK CONFERENCE ON DEVELOPMENT ECONOMICS 1998, at 127 (Boris Pleskovic & Joseph E. Stiglitz eds., 1999). 5. See generally ROBERT H. BATES, WHEN THINGS FELL APART: STATE FAILURE IN LATECENTURY AFRICA (2008).

210

WASHINGTON LAW REVIEW

[Vol. 87:205

democracy,” to borrow Justice Louis Brandeis’s famous hopeful phrase for describing the states in the American federal union, or in this case as “laboratories of development policy,” analogous to the Asian congeries of Singapore, Hong Kong, South Korea, Taiwan, and Thailand. Two (out of the three dozen or so) very small African states—Botswana (neighboring South Africa) and the island nation of Mauritius—have managed rapid economic growth with poverty reduction, so small size is obviously not categorically fatal. And the largest African states, both in terms of area and population—Nigeria, the Democratic Republic of Congo, Ethiopia, and Sudan—have not managed to turn size into development. It is nonetheless hard to see the fragmentation of Africa as neutral from the point of view of development. If it seems more likely than not, in spite of the uncertainties, that balkanization reduced and continues to reduce the possibilities for effective development policy across subSaharan Africa, then it seems fair to place some of the blame for extreme poverty on balkanization. Indeed, given the obstacles, it is striking that in the ISI period, up to 1980, sub-Saharan Africa registered significant growth in per capita income, although nothing like a take-off into selfsustaining growth occurred. Recognizing balkanization as a factor does not negate the existence of a poverty trap for today’s African states. The trap may be a crucial fact about that world, precisely because it is a balkanized world. Given balkanization, the existence of a trap provides a strong argument for increasing Western aid. But it is important to remember that the trap is not a “natural” or inevitable aspect of the post-colonial situation, but rather the product of a particular sequence of events, unfolding within a particular normative order, enforced through a combination of global law and power politics. The next section proposes a similar critique of the trap idea from the point of view of political economy. II. A PERVERSE POLITICAL ECONOMY STIFLES DEVELOPMENT

In critical legal studies, we devoted a good deal of attention to ways in which legal rules, justified on the ground that they lead to efficient resource allocation, may produce distributive outcomes so unjust as to make the efficiency advantages seem unimportant or even perverse.6
6. Summarized in Duncan Kennedy, Law and Economics from the Perspective of Critical Legal Studies, in 2 THE NEW PALGRAVE DICTIONARY OF ECONOMICS AND THE LAW 465 (Peter Newman

2012]

AFRICAN POVERTY

211

(We also emphasized the likelihood of multiple efficient equilibria with sharply different distributive outcomes, and the bogus quality of many arguments for the efficiency of particular rich-friendly rules.7) In this section, I am going to pursue a reverse kind of argument, along the lines that: the distributional outcome produced by a given institutional order (legally enforced) can preclude the investment needed if there is to be economic development that reduces poverty. Of course, it is questionable whether we know enough about what causes povertyreducing growth to be able to evaluate the effects of any particular economic regime, and I will address this well-justified skepticism in the next part of the essay. A key idea here is that across the sub-continent, and North Africa and the Middle East as well, the post-independence configuration produced a particular political economy, obviously not everywhere identical, but with many common features.8 In African countries at the time of independence, and still today, there has been both very significant foreign ownership, direct and indirect, of what exists in the way of commercial, agricultural, mining, and industrial assets, and significant foreign debt owed to private, governmental and international financial institutional (IFI) creditors. In every African country there are more-or-less Westernized dominant groups or classes. The local elites are partly pre-colonial and partly postcolonial. They use some combination of “traditional” tribal authority, private ownership under a Western legal regime, political party organization, control of the military and police, access to international (Western, once Russian, now also Chinese) aid and investment, and control of the distribution of state revenues from taxation and ownership of natural resources to preserve their position vis-à-vis the rest of the population.9 The poor in Africa have been, since the 1920s, more and more thoroughly economically integrated both with the foreign economic forces and with the economically dominant domestic groups. The selfsufficient village subsistence economy has been gone for several generations. Because of their weak bargaining power vis-à-vis all of these actors, their incomes are so small as to preclude significant saving ed., 1998). 7. Id. at 469–70. 8. For an example of how this pattern played out in Egypt, see Lama Abu-Odeh, On Law and the Transition to Market: The Case of Egypt, 23 EMORY INT’L L. REV. 351 (2009). 9. See generally MAHMOUD MAMDANI, CITIZEN AND SUBJECT: CONTEMPORARY AFRICA AND THE LEGACY OF LATE COLONIALISM 3–27 (1996).

212

WASHINGTON LAW REVIEW

[Vol. 87:205

for investment in development on their own behalf. They are also far too weak politically to force the elites to invest with their interests specifically in mind. At the same time, the different groups that appropriate most of the surplus the economy generates have very limited incentive to invest their part in the kinds of enterprise that could over time generate povertyreducing development. Investments likely to redound to the long-run benefit of the poor comprise a minuscule part of the total value generated in the economies of which the poor are a part and on which they are completely dependent. In other words, the extremely poor do not have the resources but those who do have other things to do with their money. First, there is the collection of interest by foreign creditors and the repatriation of profits from foreign owned enterprises, whether in manufacturing or in resource extraction, often operating according to the World Bank recommendation that resource rich countries grant ultra-favorable concessions to the multinationals to induce their investment. Second, governments, money-lenders, and international trading companies appropriate a big share of the surplus from agricultural exports (the famous “exploitation of the countryside”).10 Third, elites appropriate a large share of the surplus from extractive industries, after the foreign share. Fourth, governments allocate a significant share of surplus to buying the support of people in the gigantic cities, mainly along the coast, refugees from deteriorating agricultural life in the “interiors,” surviving with food subsidies in the urban informal economy and/or on government jobs that are nominally white collar (rather than jobs in industry or extraction). The way this works has changed over time. In the phase of import substitution industrialization (roughly, for any given state, from independence up to 1980), national elites controlled the new, often parastatal banks and industries, or worked for foreign multinationals that did that job. In the absence of pre-War manufacturing experience (except in South Africa), there was no basis for the enrichment of an established entrepreneurial bourgeoisie, and it turned out that the size of the internal market was in most cases much too small (balkanization) to permit the economies of scale that would have been necessary for local industries to sustain themselves without large subsidies continued indefinitely.
10. For a concise summary, see BATES, supra note 5, at 58–62. See generally ROBERT H. BATES, MARKETS AND STATES IN TROPICAL AFRICA: THE POLITICAL BASIS OF AGRICULTURAL POLICIES (Brian Barry & Samuel L. Popkin eds., 1981).

2012]

AFRICAN POVERTY

213

The surpluses generated by ISI went to many different subcategories of the population strategically placed as rent-seekers, ranging from recipients and corrupt issuers of import licenses, to unionized workers in the formal sector, to the directors and managers of the many kinds of mixed public/private enterprise, to civil servants, to politicians who siphoned off large chunks of aid money. Domestic elites were appropriators rather than entrepreneurs. After 1980, this regime underwent quite drastic change. The elections of Reagan, Thatcher, and Kohl, the consequent change in leadership of the IFIs, and the effects of devaluation and general global south financial crisis ushered in the new world of the Washington Consensus. The Consensus was in favor of dismantling the ISI institutions through deregulation and privatization, and opening the third-world economies to the global commodity and capital markets. The idea was in essence to shrink the role of the state, with the idea that its place would be more or less automatically taken by “the market.”11 The decline or demise of much of the subsidized industry, the privatization of public services and the turn to export-led growth in an economy opened to imports brought new forces into existence. But it was a serious error, both of theory and of observation, to think that “getting prices right” would eliminate or even diminish “rent-seeking.” Instead the sources of rents shifted from regulatory restrictions to privately generated rents in uncompetitive markets characterized by all kinds of informational and power asymmetries. Having “the” cell phone franchise might be worth more than access to import licenses ever was. In both phases, the surpluses not recycled to secure the support of the urban masses were either repatriated by foreign creditors and multinationals, invested abroad by members of the local dominant class, spent on the employment of service workers and servants, spent on imported consumption goods like automobiles, or invested locally in activities unlikely to increase productivity (for example, luxury housing construction). Of course, in the neoliberal fantasy of perfect capital markets, none of this would make any difference. The neoliberal argument is that the surplus extracted becomes part of the supply side of a now globalized capital market. The flow of capital from that supply depends on the
11. An important document in the African version of this turn was the Berg Report, published as THE WORLD BANK, ACCELERATED DEVELOPMENT IN SUB-SAHARAN AFRICA: AN AGENDA FOR ACTION (1981). It was a very odd idea that deregulated economies would default to “the market,” because there was no defined institution of this kind covering more than a small part of the economy in any African state other than South Africa.

214

WASHINGTON LAW REVIEW

[Vol. 87:205

decisions of the managers of foreign multinationals about what to do with repatriated profits, the decisions of creditor financial institutions about what to do with interest and amortization paid by African countries on private and public debt, the investment decisions of banks receiving deposits from them, and from the African appropriators of economic surplus (whether honestly or corruptly obtained). If there are opportunities for investment in growth-producing activities, then the managers of global capital should be able to spot them (indeed, according to the fantasy, much more able than even the rare well-intentioned government). In short, the expatriated savings generated by economic activity in Africa should be recycled right back to Africa to the extent it makes “economic sense” to invest them there. I doubt that many of my readers will find the neoliberal fantasy plausible. It seems much more likely, at least to me, that the political economy I have been describing has directed only a small proportion of domestically generated surplus into development-inducing investments that would have “made economic sense” from the point of view of the African poor. But it might be useful to explain this skepticism in a little detail. First, the quantity of savings that becomes part of the global supply of capital is, in this scenario, the result of the various extractive strategies of the various strong actors (creditors, multi-nationals, elites, etc.). The size of the surplus is arbitrary from the point of view of development policy, i.e., unrelated to the quantities that it would make “economic sense” for a developmental state to force as savings and then invest. Second, the surplus extracted will be allocated by the managers of global capital according to their assessment of economic return. These managers operate under conditions of imperfect information, different risk preference for different investor classes, and different time horizons for assessing future returns. They have no legal, and in their own view, no moral obligation to channel African surplus back to Africa. The postcolonial balkanization of Africa means not only that internal markets are usually small, but also that political risk is often severe. African investment opportunities compete for global capital against the opportunities in every other capital importing country. Even if we assume highly rational investment decisions, there is no reason to believe a priori that they direct back to Africa any significant proportion of the surplus extracted. It may well have been the case, if we suppose a well-functioning global capital market, that African surplus has financed development in other parts of the world, indeed everywhere but in Africa itself. In a world system characterized by circular causation and extreme path-dependency, this kind of pattern may well have

2012]

AFRICAN POVERTY

215

reinforced African disadvantage over time.12 From a Kaldor–Hicks economic efficiency point of view, this is a preferred outcome, so long as the return on investment of African surplus in non-African markets is higher than African return. The analogy here is to the investment of the savings of the AfricanAmerican inhabitants of U.S. urban ghettos, collected by the ghetto branches of national banks, in projects located everywhere but in the ghetto.13 This phenomenon led in the 1970s to the Community Reinvestment Act, which made at least a gesture toward requiring some reinvestment of savings in the low-income communities that generated them.14 Let’s imagine counterfactually that reinvestment of African surplus in Africa is somehow required. The comparative claim—that the market allocations that have actually occurred over the last fifty years have been better than the alternatives—is based on a conservative critique of a progressive argument in favor of state-led development. The progressive argument was that what makes “economic sense” from the point of view of the private return to investment may—or most decidedly may not— correspond to what makes “economic sense” from the point of view of poverty-reducing economic growth. This is just a variant of the old, old argument of Pigou: private return may diverge dramatically from social return.15 In this case, private return to growth-neutral or even growthimpeding investments may be larger than the return to socially productive investments. Within the growth-neutral category, povertyreducing growth may be less profitable than poverty-exacerbating growth. In the context of African development, the argument has been that the social return to investment in local health, education, infrastructure, the mechanization of agriculture and, yes, import-substituting or exportgenerating industrial enterprise (for example, fostering “national champions”) was and is much higher than the private return. The neoliberal response has been that private actors going for private return are actually more likely to induce growth than state actors attempting the

12. The classic study of this type of phenomenon remains GUNNAR MYRDAL, THE DRIFT TOWARD REGIONAL INEQUALITY IN A COUNTRY, in ECONOMIC THEORY AND UNDERDEVELOPED REGIONS (1957). 13. See DANIEL R. FUSFELD & TIMOTHY BATES, THE POLITICAL ECONOMY OF THE URBAN GHETTO 225–31 (1984). 14. For an introduction to this topic, see generally A. Brooke Overby, The Community Reinvestment Act Reconsidered, 143 U. PA. L. REV. 1431 (1994–1995). 15. See A.C. PIGOU, THE ECONOMICS OF WELFARE (1920).

216

WASHINGTON LAW REVIEW

[Vol. 87:205

impossible tasks of identifying “true” social return (“picking winners”) and often without any loyalty to the task of pursuing it even if they could identify it.16 Their argument is that African leaders should, if anything, have tied themselves earlier and more firmly to the mast of the good ship Market, rather than trying to steer their own course. My argument has been that post-colonial balkanization combined with the specific institutional and policy arrangements, put in place first by socially oriented and then by neoliberal, globally-incentivized regimes, are the cause of African poverty. This combination led to governments that had neither the will nor the ability to plan and then effectively allocate for development in the directions and on the scale necessary to initiate poverty-reducing growth. III. CHINA AND INDIA SHOW THAT TAKE-OFF IS POSSIBLE BUT WE DON’T KNOW WHY IT HAPPENS The argument of the last Part depends, of course, on the plausibility of the counterfactual claim that some conceivable geo-political and political economic alternative might have done much better. In Part IV, I take up this kind of counterfactual argument, hoping that it will also be useful in comparing the effects of balkanization and policy failure. Before pursuing this thought experiment, it seems necessary to deal with the objection that, in its current state, development economics offers us no hope either of explaining African poverty, or of assessing counterfactual alternatives. A. The Weakness of Development Theory17

The initial approach of Western non-communist development economists was based explicitly or implicitly on the idea that proper policies could trigger “take-off into self-sustaining growth,” thought to lead more or less automatically, after an initial phase of growing inequality, to poverty-reduction on the model of what happened in the developed North Atlantic countries. The proper policies were explicitly

16. One—but not the only—virtue that neoliberals have claimed for private market allocation of investment has been put in technical economic terms: private capital markets are efficient, or at least reasonably efficient, compared to other allocative mechanisms. The major problems here are: first, that capital markets are notoriously inefficient and “second best problems” are rampant; and second, that the efficiency falsely claimed for them is in any case merely static. There is no reason to think that Kaldor–Hicks efficient outcomes correspond in any ways to growth-inducing outcomes. 17. For useful treatments of the issues raised in this section, see generally JAMES M. CYPHER & JAMES L. DIETZ, THE PROCESS OF ECONOMIC DEVELOPMENT (3d ed. 2008).

2012]

AFRICAN POVERTY

217

non- or anti-communist, but also initially quite interventionist, by contrast with the neoliberal phase that began around 1980. If we had a strong theory of what causes take-off into self-sustaining growth, we could assert with more confidence what the consequences of balkanization and “perverse political economy” have been for poverty in Africa. But theories of development have undergone what strikes me, as a person who was first interested in the topic as an undergraduate economics major in 1961, as a continuous and unrelenting fifty-year process of invalidation. This list of phrases is meant to evoke the proliferation of descriptive and prescriptive categories for understanding and inducing development: capital/output ratio, balanced growth, big push, take-off, unbalanced growth, ISI, size of the internal market, savings rate, unlimited supply of labor, infrastructure, electrification, irrigation, green revolution, human capital, community development, export-led growth, poverty reduction, rent-seeking, getting prices right, comparative advantage, foreign direct investment, rule of law, property rights formalization, developmental state, embedded autonomy, new institutionalism, national champions, new developmental state, Washington Consensus, post-Washington Consensus, . . . with the upshot being something like: the expert intones that “my expert opinion is that there is no one size that fits all,” which turns out to mean no theory at all. The economies of Latin America, the Middle East, and Africa grew substantially between 1945 and 1980, but they did not take-off into selfsustaining growth. Their failure to close the gap with the developed Western and communist economies, and then their disarray in the face of the oil shocks and world financial crises of the late 1970s and early 1980s, set the stage for the takeover of neoliberal “free market” theories, enforced by the international financial institutions and national aid agencies. The even more disappointing results of the neoliberal turn for Latin America, the Middle East and Africa after 1980, have led to the reconfigurations of development policy discourse in our time. The first new formula was “basic needs,” then “poverty reduction,” as opposed to “growth,” succeeded in interesting ways by “human development” or “development as freedom,” and then “extreme poverty reduction” (as against growth in GDP per capita), and economic and social rights, or “development as rights,” in place of statistical aggregates. The good thing about each of these formulae is that they have a distinct “distributive” focus. As opposed to GDP per capita, they state their goal in terms of impacts on the people at the bottom of the heap, rather than on national aggregates. The virtually exclusive attention to

218

WASHINGTON LAW REVIEW

[Vol. 87:205

aggregates had had the effect of validating development strategies that dramatically enrich the top while impoverishing or just neglecting everyone else. Another good thing about them is that they go beyond monetary measurement, explicitly defining development in concrete terms of health, education, life expectancy, and so forth. The down side of the new categories, to my mind, is that they deliberately remove the focus from the idea of a governmental project to dramatically increase productivity and initiate take-off. As Amsden has noted,18 their focus is instead on welfare, on conditions of life, rather than on production seen as something to be strategized about from a national center. The notable exceptions are the rise of micro-lending and of the “titling” idea. The promise of each of these is that a microreconfiguration of currently distorted incentives could, by increasing poor people’s access to capital, dramatically increase economic activity (however implausible this may seem). This is consistent with the neoliberal Washington Consensus idea that once developing countries got rid of their failed state enterprises and subsidies, their welfare and regulatory regimes, and impediments to foreign trade and capital mobility, prices “gotten right” would be enough not just to increase static Kaldor–Hicks efficiency but to take care of long term development as well. It is interesting that poverty reduction, human development, and economic and social rights, as goals or values, are quite well-suited to this neoliberal policy framework, not because of what they ask for but because of what they do not ask for, namely a developmental state driving for take-off. The appeal of these ideas was fed, toward the end of the period of import substitution industrialization, by something more than just disappointment (except for the East Asian Tigers), with state-led development. Another factor was that the states that failed to deliver take-off did deliver massive human rights abuses along with dictatorships and wars. If there was an upside to the loss of confidence that take-off was around the corner, it was that after 1980 it was hard to argue that these unpleasantnesses were the price of dramatic economic transformation, because that had not occurred. B. Take-Off Is Back

The shift in the emphasis of development theory from the grandiose to the pragmatic, eclectic, and welfare-oriented, has been once again
18. Alice Amsden, Say’s Law, Poverty Persistence, and Employment Neglect, 11 J. HUMAN DEVELOPMENT & CAPABILITIES 57 (2010).

2012]

AFRICAN POVERTY

219

destabilized, this time by the more or less simultaneous apparent take-off of the Chinese, Indian, and Brazilian economies. It is obvious that they are reducing poverty on a scale and at a speed that dwarfs and marginalizes not just the “accomplishments” of the Washington Consensus, but also the modest progressivism of the post-Consensus Millennium Development Goals project. It seems to have turned out that take-off is possible after all. Possible, but no more comprehensible than it was in the 1960s heyday of the Rostow theory,19 whose emptiness was well critiqued and put in institutional context by Gunnar Myrdal in The Asian Drama, published in 1968.20 China, India, and Brazil, from the point of view of the older theories, have surprisingly little in common as models. It seems that very different institutional orders can each generate a high savings rate, with the savings channeled consistently over time into the hands of economic actors who will invest, and then reinvest, in ways that lead to a dramatic shift from low-efficiency agricultural production to much more productive industrial, agricultural and extractive pursuits. In each case there was a sustained period of state-led development followed by a dramatic, but also highly regulated, turn to the market. (None of these states did anything like what the IFI’s were demanding and getting from the more or less bankrupt African states.) Foreign direct investment (FDI) played anywhere from a small to a large role depending on the case. There are many possible ways to fit these facts together into a theory: crediting or condemning the state-led phase, crediting or condemning the role of FDI, crediting or condemning the post-liberalization management of the economy, and so forth. Investment is obviously not the only driver of growth, and it is even possible to imagine scenarios in which a “take-off” is provoked not by new investment but by a dramatic increase in the efficiency of existing capital occurring without significant changes in savings/investment quantities and practices. Nonetheless, the intuition that balkanization and “perverse political economy” are causes of African poverty is based on the idea that they made it impossible, or at least unlikely, that there would be the mobilization and direction of savings in an initial phase laying the foundation for later growth. Likewise, that they precluded the aggressive management of the post-liberalization market so as to sustain not just savings, but the investment and reinvestment of savings in high19. See W.W. ROSTOW, THE STAGES OF ECONOMIC GROWTH: A NON-COMMUNIST MANIFESTO (1960). 20. 3 GUNNAR MYRDAL, ASIAN DRAMA: AN INQUIRY INTO THE POVERTY OF NATIONS 1850–55 (1968).

220

WASHINGTON LAW REVIEW

[Vol. 87:205

productivity activities, drawing labor out of agriculture.21 The idea that the mobilization and direction of resources for investment is the key to development is shared by the poverty-trap analytic, and explains the centrality of foreign aid within it. Jeffrey and Lisa Sachs, in the short introductory piece I quoted above, put it this way, focusing not on the Millennium Development Goals but on the denial of economic and social rights (ESR): [A]dequate financing is needed to alleviate the poverty-related burdens of weak governance, inadequate infrastructure, excessive population growth, poor health, low literacy, and resource depletion, which all contribute to extensive ESR violations. Yet economic development is precluded by precisely these poverty-related burdens, which decrease productivity, earning ability, and economic investments. The implication of poverty traps is that poor countries, on their own, are often unable to honor their basic economic and social rights obligations. Realizing these human rights requires increased public outlays and infrastructure that are beyond the financial means of poor countries. This is not to remove the responsibility of the home governments of low-income countries, but it is to emphasize that ESR is often a partnership affair—achievable through the joint efforts of rich and poor countries alike.22 My argument above was that while the Sachs are right to focus on investment, the cause of the poverty-trap is not, as they suggest, the poverty of countries in which the poor live, combined with the failure of the developed core countries to provide adequate aid, but that the poor are embedded in national and international political economies that predictably direct economic surpluses away from local reinvestment that might improve their situation through development. The next Part tests the plausibility of this argument through two counterfactual thought experiments. IV. COMPARING BALKANIZATION WITH POLICY FAILURE AS CAUSES OF AFRICAN POVERTY Is it at all plausible that either balkanization or “perverse political economy” is a cause of African poverty? If it is plausible, how much is due to the balkanized situation, and how much to the sequence of relatively ineffectual ISI followed by an equally ineffectual turn to
21. For a defense of this theory of “take off,” see CYPHER & DIETZ, supra note 17, at 308–40. 22. Sachs & Sachs, supra note 1, at xvi.

2012]

AFRICAN POVERTY

221

markets? With these questions in mind, I suggest two counterfactual thought experiments that are supposed to test, in a very loose and imprecise way, the plausibility of attributing African poverty to the combination of balkanization and perverse political economy. The first counterfactual imagines the same stages of political economy that occurred in post-independence sub-Saharan Africa, but managed by a sub-continental federal mega-state. The point is the contrast with the actual balkanized and unstable context. The second counterfactual imagines that a unitary sub-continental mega-state has been governed since its origin by a revolutionary communist party emulating the Chinese model of development through its post-Cultural Revolution stages. The first thought experiment suggests that a subcontinental state pursuing the policies followed in real life in a balkanized context would have done only a little better than the balkanized states did in fact. The second suggests that a revolutionary continental state might have done much better. It would seem to follow, in the roughest possible way, that while balkanization counts, policy counts more. A. ISI and Liberalization in a Sub-Continental Federal State

Let us suppose that the newly independent African countries had managed early on to form a single federal state, and that this state had managed to maintain a much higher level of internal political stability than has the balkanized continent. The main terms of federal union, let us imagine, have to do with the economy. There is a customs union, and rights of free movement of goods, workers, and capital within the federation. There is a central authority empowered to enforce the terms of union, to manage trade and financial relations with the outside world, to levy some sub-continent wide taxes and to carry out the financial, industrial and infrastructure policies characteristic of the moderate version of ISI. The moderate version has the features I described in Part II, above, including extensive foreign ownership, a mixed colonial and post-colonial domestic elite relying on a combination of “traditional” and modern property rights, along with military and police, party political and raw economic power to keep control of the state. Of course, it is hard to imagine how a pan-African federal state could possibly have come into existence, and even conceding that balkanization was a cause of political instability, it is hard to imagine that the state could have avoided all kinds of severe internal and external stresses. But we don’t have to imagine that every single square mile of the sub-continent is included, and we can allow for significant resources devoted to maintaining order, through military rule, civilian strongman

222

WASHINGTON LAW REVIEW

[Vol. 87:205

authoritarianism, or some system of power sharing through political parties. At its creation, in let’s say 1965, the state would have had a maximum possible area of approximately 23,650,000 square kilometers, compared to 9,327,450 for China and 2,973,190 for India.23 It would have had a relatively sparse population of approximately 260 million, compared to 715 million for China and 487 million for India, on their much smaller territories, and GDP (constant 2000 U.S. dollars) of approximately $488 per capita, compared to $100 for China, and $193 for India in that year.24 The (comparatively) high African GDP per capita reflects the settlerand foreign-owned mining economies of the subcontinent at independence, and also the relative prosperity of small farmers producing cash crops for export. Of course these statistics are only suggestive, given their radical imprecision, both conceptually and empirically. 1. The Federal Mega-State Might Well Have Outperformed Balkanized Africa During the ISI Period

The government of the federal state would have been a large entity, but because it would not have had to waste a very large proportion of the post-independence educated cadres on the duplication of state structures, it might well have been more effective than any of the actual successor states. It would not have faced the “challenges” implicit in division into “coastal resource-poor,” “landlocked resource-poor,” and “resource rich” (oil cursed) states. The federal government would have been able to raise revenue in one place and spend it in another, and to encourage the movement of complementary resources into conjunction across subnational boundaries. The state would have had a diverse set of agricultural and mineral exports, and therefore a more stable source of foreign exchange, than the four dozen balkanized states, often practicing export monoculture and with a single natural resource, have been able to attain.

23. WORLD BANK, WORLD DATABANK: WORLD DEVELOPMENT INDICATORS (WDI) & GLOBAL DEVELOPMENT FINANCE (GDF), http://databank.worldbank.org/ddp/home.do (follow the “Select Variables” tab; select “China” and “India”; follow the “Next>>” button; select “Land area (sq. km)”; follow the “Next>>” button; select “1965”; follow the “Next>>” button; then choose reporting format to view data for each country) (last visited Feb. 24, 2012). 24. Id. (follow “Select Variables” tab; select “China” and “India”; follow the “Next>>” button; select “GDP per capita (constant 2000 US$)” and “Population, total”; follow “Next>>” button; select “1965”; follow “Next>>” button; then choose reporting format to view data for each country) (last visited Feb. 24, 2012).

2012]

AFRICAN POVERTY

223

It would no longer be the case that a large part of the African population lived in states far too small to provide viable internal markets for an import substitution strategy. The combination of a continental internal market with political stability might have made the federal state more attractive to world capital markets, both for credit and for FDI, than was Africa under actual conditions. This would apply both to investment in the protected ISI infant industries and in the extraction of natural resources, to the extent the federal state followed say, Brazil and the Ivory Coast, friendly to FDI, rather than the more nationalist variants. Africa would have probably gotten back a considerably larger share of its surplus than under the actual regime. It is possible, though not certain, that political stability would have induced political power holders to pay more attention to long-term projects for growth, rather than to extracting as much as they could, by any means available, in the expectation that their chances for enrichment would be short-lived.25 2. The Mega-State’s Perverse Political Economy Would Likely Have Had Many Disastrous Effects

However, many of the factors that have precluded take-off oriented investment in real life Africa would have existed in the imaginary federal state. A vast array of assets would still have been owned by multinationals, trading companies, and ethnically Arab or Indian merchants and moneylenders, and foreign creditors would still have been in a position to strongly influence policy because of the dependence of the government on a continuous flow of credit. Surpluses would still have been controlled by foreign creditors and multinationals, the domestic elites would still have appropriated a massive share, shipped it abroad or spent it on servants and villas. They would have had the same need to exploit the countryside and buy off the urban masses and so forth. The tendency of ISI regimes to run trade and budget deficits, leading to periodic currency crises, vulnerability to speculative attacks and dependence on the IFIs, was aggravated by the instability of prices in world commodity markets for their primary product exports. But the tendency to balance of payments crises was also caused by the willingness of ruling groups to overvalue the currency, raising the price of exports while making imports nominally cheap, and then to impose massive quantitative restrictions on imports but with widespread exceptions and evasions.
25. See BATES, supra note 5, at 26–27.

224

WASHINGTON LAW REVIEW

[Vol. 87:205

Budgets deficits like trade deficits had appeal in part because of the always-threatening instability of the politics of post-independence Africa. Whoever was in power had to thread their way between the “traditional” tribal elites, the new African property holders, the military and police forces, the urban masses, the multinationals, banks and IFI’s, and Western governments always conditioning the aid (essential to grease these wheels) on avoiding any economic policy that looked even vaguely “red.” These problems would have existed in the federal state as well. And as in balkanized Africa, the federal state might have solved them through authoritarian one-man rule. But stability of this kind (think of Mobutu) would not likely have staunched the drain of social surplus in every direction except into large-scale effective growth-inducing and povertyreducing investment. For me, a key point here is that in dealing with a continental federal state pursuing ISI, the dominant ex-colonial powers, very much including the United States, would in all likelihood have worked even harder to reward cooperation with Western economic interests and to punish ruthlessly any kind of experiment in the direction of “de-linking,” expropriation, or “redistribution.” As happened in the real world, the Western powers and the multinationals would have allied with the domestic elites, offering them some guarantees against an uprising from below and access to personal riches, at the price of policy subservience, and a wide variety of public and private corruptions. I think a continental state might have done a good bit better than the balkanized states in the ISI stage up to 1980, but unlikely that there would have been anything comparable to the (costly) accomplishments of China, or for that matter of India. As for the stage of liberalization, the federal state pursuing the policies of the neoliberal IFIs would have opened its economy by freeing the currency, eliminating controls on capital movement, reducing or eliminating both tariffs and export subsidies, dismantling its public and parastatal enterprises, and allowing the ISI protected industries to fail if they could not compete with imports from say, China, Iran, Brazil or Bangladesh. It would have shrunk the state sector to a remnant of its former self, and made provision of water, education and health conditional on user fees. Rule-of-law programs financed by aid would have built new courtrooms and computerized them, in a probably vain effort to increase “access to justice” defined as access to the rule structure left over from the colonial period. Whereas it seems quite possible that ISI in a continental federal noncommunist state would have performed better than did African states in

2012]

AFRICAN POVERTY

225

balkanized reality, it does not seem plausible, at least to me, that the imaginary federal neoliberal program would have outperformed the real world African version. True, the oil and mining companies negotiating with a continental state would have had less bargaining power, but they would still have had full Western support for their outrageous contracts and enforcement through bilateral investment treaties. All the leakages abroad would have increased exponentially as the structure of ISI controls was relaxed, just as happened in real life Africa. B. Economic Development Under an Imaginary African Communist Mega-State

Given this assessment of the effects of balkanization, the remaining question is whether there was an alternative political economy that might have produced better results for Africa than did the one actually adopted under the conditions of balkanization. The main extant model of an alternative political economy that produced take-off is that of China, with the Indian and Brazilian models being currently contenders for a similar status. For reasons of space and competence, I am going to construct my second thought experiment using China, with some Indian comparisons, leaving Brazil out altogether. The exercise suggests that a sub-Saharan African communist megastate might have performed far better than the actual African states did, and also far better than did the hypothetical federal state pursuing the moderate non-communist version of ISI. And this in turn suggests that policy failure was a much more important cause of slow growth and the persistence of extreme poverty than was balkanization. 1. A Sub-Continental African Communist Mega-State as a Thought Experiment

To begin with the outcomes in real life, for the period between 1965 and 2010, sub-Saharan Africa compares as follows to China and India (with massive reservations as to accuracy):

226 TABLE 126

WASHINGTON LAW REVIEW

[Vol. 87:205

Africa GDP per capita (constant 2000 US$) India GDP per capita (constant 2000 US$) China GDP per capita (constant 2000 US$) Africa population India population China Population

1965 488 193 100 260,000,000 487,324,000 715,185,000

2010 640 823 2425 854,951,000 1,170,938,000 1,338,299,512

Imagine that in the early 1960s, at the time of African independence, a communist political movement had managed to consolidate all the excolonies of sub-Saharan Africa into a single state (including South Africa, Southern Rhodesia, Angola, Mozambique, and Portuguese Guinea), and then to maintain power up to today. Imagine that such a state had modeled itself after China, perhaps omitting the stage of the Cultural Revolution (over by 1971). It would have begun with the ruthless suppression of political opposition, including breaking of the power of tribal political institutions. The mega-state would have maintained domestic order over the whole sub-continent by a combination of secret police and brutal military interventions. That state, while very poor by any standard, would have had, by virtue of its sheer size, its control of relevant assets, and its political ability to force savings, very large resources in absolute terms, and would not have devoted money and very scarce Western educated manpower to duplicative state structures. By virtue of its centralized dictatorial governmental power it could have deployed those resources, for good or ill, in aggressive and far-reaching development programs, without reliance on the foreign aid that is the sole key to development in the poverty-trap model. These policies might, just possibly, have eventually produced the kind of self-sustaining transformative development process that has occurred in China. The consolidated communist state would have pursued a strategy of centrally planned industrialization, collectivization and mechanization of agriculture (including electrification and irrigation projects), and state development of oil and mineral resources. The Western powers might well have boycotted African mega-state exports of agricultural products

26. WORLD BANK, supra note 23 (follow “Select Variables” tab; select “China” and “India”; follow “Next>>” button; select “GDP per capita (constant 2000 US$)” and “Population, total”; follow “Next>>” button; select “1965” and “2010”; follow “Next>>” button; then choose reporting format to view data for each country) (last visited Feb. 24, 2012).

2012]

AFRICAN POVERTY

227

and natural resources,27 in which case they would have gone to the Soviet bloc and China, from which the African communist mega-state would have imported industrial equipment and low-quality consumer goods. Population control would have been a major policy, through something like the Chinese “one child” policy, adapted to the African circumstances. All of this would have involved a lot of deadly violence and oppression of all kinds. If we imagine this counterfactual in 1990, the year after the fall of the Berlin Wall and twelve years after the beginning of the Chinese turn to markets in 1978, I think most interested observers asked to speculate would have said that the vast human sacrifices that a post-1960 panAfrican communist state would have exacted would have produced little or no economic progress, with little or no prospects for the future. A first reason would be that as of 1990 it was by no means clear that China was approaching take-off. It was considerably more plausible that the state-led, centrally planned and collectivist phase of Chinese policy had been an unmitigated disaster from the point of view of development, setting the country back perhaps by decades. It would have seemed unlikely to say the least that my imaginary African version of the Chinese revolution could have a positive outcome. The Indian liberalizing turn was still a year in the future. This analysis of “what would have happened” under the counterfactual would have been reinforced by the experiences of Tanzania under Nyerere; Ghana under Nkrumah; Guinea under Sékou Touré; and by Ethiopia, the Sudan, and Congo-Brazzaville under their hard Marxist versions of ISI. As of 1990, these African versions of “radical” ISI, whether or not they were significantly influenced by the Chinese model, were clearly closer to it than the moderate versions. There may have been many and diverse reasons for their failure, including that they drew the intense hostility of the Western Cold War powers, but the fact remained that each seemed to have led to spiraling economic

27. Why is there no communist African state? Because the Western powers, as an important part of their Cold War strategy, did everything they could to prevent the emergence of anything that looked even a little like a communist regime in the Global South, and to overthrow or strangle the few that managed to come to power nonetheless. The United States has played a role: the assassination of Lumumba and installation of Mobutu, backing the pro-Western tribal groupings in Angola; backing first one side then the other in the Ethiopia/Somalia conflict; tolerating apartheid; working to isolate Nkrumah and Nyerere. At the same time, the British and French, and their multinationals, have been continuously active—as much by buying out potential radical left leaders as by suppressing them directly. A second reason, no less important, is that the African states that directly emulated either a Russian or a Chinese model did extraordinarily badly in the period leading up to the neoliberal turn of 1980 followed by the collapse of the Soviet Union in 1989.

228

WASHINGTON LAW REVIEW

[Vol. 87:205

dysfunction, and eventually to negative growth. In that atmosphere, and a year after the fall of the Berlin Wall, looking for a thought experiment to assess the causes of slow African growth, it might have seemed plausible to ask, “What would have happened to African growth if there had never been an ISI and the subcontinent had pursued a strong free market policy, with a small state, no regulation, and open economies?”28 In 1990, the neoliberal radical turn to the market and away from the state had just passed its peak, as the disastrous results for Africa in the 1980s were becoming obvious. It might have been conceded in 1990 that the results of the radical neoliberal turn of the International Monetary Fund and the World Bank (the IFIs) after the crises and political changes of 1980 had been particularly disappointing in sub-Saharan Africa. But the conclusion would have likely been that this merely made a case for a partial return to the pre-1980 policies, or perhaps for a turn to institutional development focused on the rule of law and human capital, both feasible only with substantial foreign aid and aid-funded foreign expertise. Is that reading of “what might have been” still plausible today? Let us continue with our thought experiment. Imagine that after massive forced saving (derived mainly but not only from agricultural and mining exports) invested in industry, agriculture, resource extraction and infrastructure, much of it wasted, inefficient, or bureaucratically mismanaged, the mega-state made a turn to the market, perhaps somewhat later than in China itself. In this turn, export-led growth, perhaps with more emphasis on oil and mineral and agricultural exports and (just a little) less on manufactured goods, supported by strict control of consumer goods imports, the exchange rate, and foreign direct investment, and under strong “steering” by state economists, might or
28. Even today, an important study of African development can assert that a major cause of the lag in African growth by comparison with the rest of the developing world is the “state control syndrome,” supposedly inspired by “socialism” and defined to include all the characteristic policies of moderate ISI, from marketing boards to development banks to restrictions on capital flows. For this view, the interesting counterfactual is not a Chinese style communist regime, but an equally imaginary “syndrome free” or completely “free market” Africa. See THE POLITICAL ECONOMY OF ECONOMIC GROWTH IN AFRICA, 1960–2000, supra note 4, at 137–201. It is interesting that chapter nine of this study, by Benno Ndulu, The evolution of global development paradigms and their influence on African economic growth, seems to concede that it cannot be simply “controls” in the abstract, contrasted with an equally abstract idea of an idealized “free market,” that explains slow growth. Id. at 315–47. But by setting up their study in that way, the authors seem to be saying that “controls” per se, regardless of their content or the mode of their administration, are the problem. That contention seems to me to have been decisively refuted well before the publication of the study by take-off in China, India, and Brazil.

2012]

AFRICAN POVERTY

229

might not have led to rapid growth in GNP and GNP per capita, and mal-distributed but dramatic reduction of extreme poverty. Sometime around now, in the more favorable scenario, the entrenched political/economic/dynastic elite governing the mega-state would be reckoning with the various environmental disasters, from desertification to industrial pollution, generated by its development policies, while toying with preemptive cosmetic democratization moves. The Western core economic powers would be demanding trade opening as the price of belated membership in the WTO. In the less favorable scenario—but still imagining that the mega-state neither disintegrated nor became an exploited Chinese or Russian “vassal state”—there would have been slow growth, little poverty reduction, a boom for Swiss banks, and several million victims of political violence. There might also have been significant communiststyle educational and public health advances, and infrastructure construction, supposing that the less successful version of the regime could still do some things right. Of course, the real word outcomes for African countries were less than uniformly positive as well. When thinking about the genuine horrors that communist regimes perpetrated on their peoples, it is good to keep in mind that much less organized African violence has also been horrific. From a database of casualties in political conflicts around the world, I added up the figures for sub-Saharan African wars, civil wars and internal repressions, between 1965 and 2010, that caused more than 100,000 estimated deaths (not counting wars of independence). The total came to 6,550,000.29 The accepted but very possibly low estimated

29. The data below are from Monty G. Marshall, Major Episodes of Political Violence: 1946– 2011, CTR. FOR SYSTEMIC PEACE, http://www.systemicpeace.org/warlist.htm (last updated Dec. 6, 2011). I have included all conflicts with more than 100,000 casualties between 1965 and 2010, along with statistics for the civil wars in Sierra Leone, Liberia, and Congo–Brazzaville. There were another dozen conflicts with tens of thousands of estimated deaths. Just the conflicts with more than 100,000 casualties add to a total of 7,450,000 dead. Of course these statistics, as the source insists, are highly conjectural, and give no more than an order of magnitude.

230

WASHINGTON LAW REVIEW

[Vol. 87:205

violent death toll in the Chinese Cultural Revolution, 1966–71, is 500,000. Famine deaths between 1965 and 2000 for which mortality estimates are available look something like this30 (with the usual strong caveats). The deaths from famine during the Great Leap Forward, 1958–62, are estimated at eighteen to thirty-three million, but after that, the 2000 study finds no famines in China. In 1972–73, 130,000 died of famine in the state of Maharashta, India. That seems to be the full story for those two countries. I added up the numbers from thirteen African famines between 1965 and 2000, in Ethiopia, Sudan, Somalia, Uganda, Nigeria, and Mozambique, and got a figure of between 2,900,000 and 3,900,000 deaths. Of course, there have been many famine deaths in Africa since 2000.
TABLE 2 Year(s) 1966 1971–78 1974–91 1975– 2002 1981–86 1981–92 1983– 2002 1988– 2011 1990–97 1991– 2001 1993– 2005 1994 1996– 2011 1997–99 1998– 2000 2001–04 2003–11 State(s) Involved Nigeria Uganda Ethiopia Angola Uganda Mozambique Sudan Somalia Liberia Sierra Leone Burundi Rwanda Dem. Rep. Congo Congo–Brazzaville Ethiopia/Eritrea Nigeria Sudan Brief Description of Conflict Ethnic war (Biafra) Ethnic war (Idi Amin regime) Ethnic war (Eritreans and others) Civil war Repression Civil war North/South Civil war Civil war Civil war Ethnic war (Tutsi/Hutu) Ethnic war (Tutsi/Hutu) Civil wars Civil wars Interstate war Communal Civil war Deaths 200,000 250,000 750,000 1,000,000 100,000 500,000 1,000,000 100,000 40,000 25,000 100,000 500,000 2,500,000 10,000 100,000 55,000 350,000

30. Stephen Devereux, Famine in the Twentieth Century 6 (Inst. of Dev. Studies, Working Paper No. 105, 2000), available at http://www.ids.ac.uk/go/idspublication/famine-in-the-twentiethcentury.

2012]

AFRICAN POVERTY

231

As for extreme poverty, here are the approximations for the period 1981–2005 (keep in mind that estimated GDP per capital for India for the year 1965 was $193): TABLE 3: PERCENTAGE OF PEOPLE LIVING BELOW $1.25 A DAY31 Region 1981 1984 1987 1990 1993 1996 1999 2002 2005 China India SubSaharan Africa 84.0 59.8 53.7 69.4 55.5 56.2 54.0 53.6 54.8 60.2 51.3 57.9 53.7 49.4 57.1 36.4 46.6 58.7 35.6 44.8 58.2 28.4 43.9 55.1 15.9 41.6 50.9

These statistics do not in any way prove that Africa would have suffered fewer violent deaths and famines, or have fewer people living in extreme poverty, if something like a Chinese style revolution had occurred. But they do indicate that the cost of the actual course of events was high, higher in these very crude terms than the post-1970 costs to China and India of pursuing their models. And they give a sense that even if a revolutionary course had been followed and failed, the failed result might not have been much worse than what actually occurred. 2. The Political Economy of an African Communist Mega-State

The imaginary African mega-state began with significant assets. South Africa, Northern Rhodesia, the Congo, soon Nigeria and Portuguese Guinea, all had massive natural resources, and, for the two southern states, colonial era direct foreign investment. Today, subSaharan Africa counts as comparatively natural-resource rich compared to China (or India). The mega-state would have had sufficient resources to sustain numerous and ambitious development projects, even in the likely event of a boycott by the Western great powers. China (and the Asian Tigers as well), achieved striking growth and equally striking poverty reduction without either a lot of aid or a lot of direct foreign investment. The Chinese policy path began with a commitment to state-controlled development, within the general parameters of ISI, with of course the dramatic episodes of the Great Leap Forward and the Cultural
31. Shaohua Chen & Martin Ravallion, The Developing World Is Poorer than We Thought, But No Less Successful in the Fight Against Poverty 42 (The World Bank Dev. Research Grp., Policy Research Working Paper No. 4703, 2008), available at http://www.worldbank.org/reference/ (follow “Documents and Reports” hyperlink; then search “Documents & Reports” for “WPS4703”).

232

WASHINGTON LAW REVIEW

[Vol. 87:205

Revolution. The mega-state, I imagine, gets started after these have played themselves out, and tries nothing similar. The second stage, after 1978, combined liberalization—the competition between capitalist entrepreneurs and farmers driving development—with very strong residual state powers of intervention. The state managed the overall strategy of export-led growth by manipulating incentives, particularly credit, trade and currency policy (somewhat like the model described for the Tigers by Amsden32). The communist mega-state’s development strategy would have been based on a specific set of normative/institutional/legal premises, not adequately summarized by the typical Western cliché that “under communism everything belongs to the state and the economy is centrally planned.” First, the mega-state would have expropriated all foreign property, including all natural resources, agricultural plantations, and real estate and commercial assets, and expelled the banks and trading companies that in real life stayed on after the British and French officials had left. It would have run its new properties through a system of administrative delegation, within which the managers of enterprises operated with significant discretion under supervision by the Communist Party’s shadow network, within the rigid larger framework imposed from the center. Second, it would have adopted a policy of tight control vis-à-vis the world market, perhaps something close to closure in the event of a Western economic boycott (except for state controlled barter within the communist economic bloc). Third, it would not have respected the private property of local prerevolutionary groups, whether modernized or traditional, either in designing programs or in deciding what to do with the surpluses they generate, except so far as that would have seemed necessary, in the second phase, to keep the post-liberalization bourgeoisie on board. Finally, after the market turn, the mega-state, imitating China (and not India), would have kept control of the liberalized economy, by refusing the political/economic moves that would make the newly enriched capitalist class sufficiently autonomous and sufficiently powerful to be able to control the course of policy in its own interests. This is a crucial part of the thought experiment, guaranteeing that the elite continues to see its class interests identified with the national whole, rather than with the profitability of business, regardless of social consequences. To put it bluntly, the imaginary communist mega-state would neither
32. AMSDEN, supra note 3.

2012]

AFRICAN POVERTY

233

have permitted the massive expatriation of surplus nor have spent more than a relatively small part of what it controlled non-productively. Surplus would have gone (still in this best case scenario) to human capital, infrastructure, industrialization, the mechanization of agriculture, natural resource extraction and technological innovation, all under intense supervision, even after the supervisors had taken a giant step back to empower the multitude of partially liberalized new actors. As I’ve already insisted, they might very well have wasted what they controlled, or succumbed to the no less serious risks of communist corruption and stupidity, but if all went well, we might have a completely different idea than the one we have of what it means to be African. 3. Of Course a Thought Experiment Proves Nothing

Because our development theory is so weak, when we imagine an African mega-state that has the means for and the objective of rapid development, we still have no confidence in predicting what would happen. We know the model worked for China. But because we have no theory, it is possible or even likely that aspects of the imagined African situation that differ from the Chinese situation would mean a completely different outcome. The Chinese communist political and economic development model may seem implausible, even as a thought experiment, because, for example, China has a relatively culturally homogeneous population, has been politically unified for millenia, has a long tradition of centralized bureaucratic control, and was colonized only partially and briefly by Japan (plus the treaty ports), all by contrast with sub-Saharan Africa. Perhaps we can do the experiment more plausibly with India. India was thoroughly colonized for a century and a half, has a culturally and ethnically heterogeneous population that was largely illiterate at independence, and has nothing like the Chinese traditions of continuous centralization and bureaucracy. The African communist mega-state might have adopted, without relinquishing its dictatorial powers, some aspects of the more decentralized, Indian, mixed public/private version of ISI. The flight of British capital allowed the national bourgeoisie to acquire the foreign assets without expropriation. The Indian state created and favored state enterprises but also supported domestic entrepreneurship and capital formation within a heavily regulated, although often ineffective, planned economy (often inaccurately characterized as “socialist”), up to the dramatic moment of liberalization in 1991. In the Indian case, there was much more rule of law than in the Chinese case. The Indian

234

WASHINGTON LAW REVIEW

[Vol. 87:205

Constitutional Court gave some protection to industrialists but not landlords, consistent with the mixed economy model. Although the Indian elites pursued these policies in a democratic context, the hypothetical African mega-state would not have had to adopt that aspect in order to benefit (or not) from the economic paradigm. It is easy enough to find differences between Africa and India that might reduce the relevance of the Indian model: perhaps the late “opening” of Africa, by comparison with India (reversing the usual argument that a colonial past stifles development), or the role of Hindu and Islamic commercial civilizations as unifying as well as divisive factors, by contrast with the stunning plurality of non-commercial African cultures. Perhaps more important than any of the above for a hypothetical mega-state would have been the lack of pre-World War II manufacturing experience, by contrast with India (as well as with China, Japan, and Korea).33 All of these factors might have militated against African mega-state success. But they might not have, or might have been overcome, for example by massive expertise and technology transfer from China. If it is right as I argued that we have no plausible theory of take-off, uncertainty about the causes of development is now to be a permanent aspect of our situation, and that includes uncertainty about the causes of African extreme poverty. My belief that the communist mega-state would have done better than the non-communist federal state is no more than a hunch, obviously influenced by my background ideological commitments. I would say the same about the liberal/neoliberal contrary intuition.34 My larger point is a simple one: if China, India, and Brazil have managed to “take off” into self-sustaining, transformative economic growth, perhaps the mega-state could have done the same. If there is a poverty trap today for most African states, the choice of a weak version of ISI followed by a strong version of Washington Consensus liberalization, all in the context of balkanization, is an important part of the explanation. CONCLUSION This paper is a critique of the poverty trap understanding of African poverty. I think this understanding underestimates the significance of balkanization. But more importantly, it also underestimates the significance of anti-developmental resource allocation caused by the way typical African political economies distribute income among
33. Id. at 121. 34. See supra note 27.

2012]

AFRICAN POVERTY

235

competing economic interests. The Western powers are chiefly responsible for balkanization. Western public and private powers are chiefly responsible for the post-independence policies of non-communist African countries, and for the perverse political economy that has typically resulted. This reinterpretation puts considerably more of the blame for African poverty on them than does the poverty trap way of thinking about it, without in any way exonerating the African elites that benefited from choosing this path. Let me close by saying what should have been obvious: Western responsibility is not an argument that it would have been better for Africa to have had a continental Chinese revolution with all the suffering that would have entailed. If there is a normative cast to the argument, it is that the Western powers should have used their utterly disproportionate power not to support the compromised regimes that emerged post-independence, but a much more radical, popular mobilization-based version of ISI followed by tightly controlled, centrally guided liberalization. Of course nothing guarantees that this would have succeeded any better than did the actual policy.