Varieties of Common Sense
Varieties of Common Sense
Abstract and Keywords
Policy officials are often influenced by two broad varieties of conventional wisdom: the set of ideas broadly associated with neoclassical economics; and those ideas flowing from third worldist, anti-imperialist, and structuralist development economics. We show how these apparently opposing perspectives often have a surprising amount in common. Reflexes of ‘impossibilism’ and ‘naive optimism’ are often shared across an ideological divide. Thus, pessimism in orthodox trade theory suggests no African economy can hope to accelerate structural change by defying the signals of comparative advantage; and pessimism in structuralist trade arguments claims limited gains from exporting, especially from exporting primary commodities while the terms of trade are declining. Both forms of pessimism can easily switch to naive optimism when they imagine the ease of rapid and ‘inclusive’ development. But the switch requires that unrealistic conditions are put in place: perfectly competitive markets or idealized South–South cooperation.
3.1 Introduction: The Fog of Development Economics
The air that policymakers breathe is thick with the smog of conventional wisdom: with particles of ideas and assumptions that stick loosely together to form a miasma of apparently self-evident truth—of common sense. The effect is that assumptions and evidence are not always challenged. Common-sense works like the mental short-cuts that people make, relying on habits of thought and bias: the field studied by behavioural psychologists that has begun to influence economists and policy officials.1
In this chapter, we outline two archetypal forms of conventional wisdom that have come to hang over the atmosphere in which policy decisions are taken in Africa. First is the set of ideas associated with orthodox, neoclassical economics. Second is the bundle of arguments linked to anti-imperialist ‘third worldist’ development theory, as well as some forms of structuralist development economics. At first sight, these are very different, contradictory approaches. However, rather than tracing their origins and complex trajectories, which has been done many times elsewhere, our objective is to draw attention to what they have in common.2 The ideas of these two supposedly contending approaches can often be seen to overlap in the views and statements of policy officials, advisors, and university economists. One example, discussed in Section 3.4, is reflected in widespread support for the African Continental Free Trade Area (AfCFTA).
If there are two different schools of thought influencing policymakers, neither is wholly self-contained. Rather, each has a range of variants, some more narrowly intolerant than others. Additionally, approaches from both schools often share one of two distinct attitudes: pessimism (and ‘impossibilism’), or naive optimism. We will show how an attitude of ‘impossibilism’ is shared by development economists who otherwise would regard themselves as sharply at odds. This ‘impossibilism’ unites, for example, those who are pessimistic about gains from processing primary commodities in low-income countries with those who insist that such countries can never create firms capable of climbing higher than the (p.47) lowest rungs on the ladder of global value chains. On both sides of the rhetorical divide, impossibilists believe that rapid capitalist industrialization in poor economies will fail to create a sufficient number of jobs in factories.
We will then go on to look at a very different, but still commonly shared, attitude: naive optimism. In particular, we are sceptical about widespread programmes promoting bottom-up, micro-enterprise, and cooperative development. We argue that an analytical attitude of possibilism should not be confused with naive optimism, and, as elsewhere in the book, warn against prettifying capitalist development (or, indeed, African long-run history). The contradictions discussed in Chapter 2 are not a ‘bolt-on’ for economic analysis, but are front and centre of our analytical framework, and, we argue, ought to inform all analysis and policy design (and evaluation).
A glance at some of the evidence on economic growth over the past sixty-odd years, for example, suggests that common-sense stories about the preconditions of growth are misleading. The country that had the least exposure through colonial settlement to Western European ‘inclusive institutions’ (that Acemoglu, Johnson, and Robinson see as critical to long-run development) achieved more cumulative growth between 1960 and 2016 than almost every other African country (Figure 3.1).3 What brief colonial settlement there was in Ethiopia took the form of fascist Italian occupation, during which some infrastructure was built and there (p.48) was a spectacular massacre in Addis Ababa.4 Ethiopia was not landlocked for all this period but became so after 1991, and was cut off even more from sea ports in the war with Eritrea in the late 1990s. The country is also clearly in what some economists classify as a ‘bad neighbourhood’. It has been characterized for much of this period by what well-known orthodox economists call ‘policy syndromes’ (i.e. non-orthodox policies). And it has for most of the period clearly not been a liberal multi-party democracy. Some of those African countries with the weakest cumulative growth in this period—including South Africa and Zimbabwe—had, by contrast, far greater experience of Western settler colonialism and—especially in the case of South Africa—have adopted ‘market-friendly’ economic policies more resolutely than other countries. Not only does the evidence in Figure 3.1 highlight the diversity of economic experience in sub-Saharan Africa, it also challenges a number of common-sense assumptions about the ‘drivers’ of economic development.
3.2 Common Sense and the Reproduction of Powerful Ideas
Exercising power involves both force and consent.5 Hegemony relies on acceptance. Getting people to regard a set of ideas and policies as ‘natural’ involves the deployment of ‘rhetorical commonplaces’: key phrases and ideas that are circulated, repeated, and insisted upon, until they come to shape the contours of legitimate debate.6 We will highlight a number of rhetorical commonplaces in this book, including ‘reforming the labour market’ and ‘getting prices right’, which distil complex bodies of thought and argument into oversimplified forms.
Rhetorical commonplaces reinforce conventional wisdom: the ‘popular conception of the world’. While common sense can mean ‘good sense’—the type of things people say and do that are generally considered obvious and sensible—it can have more complex meanings and ambiguous political implications, as Antonio Gramsci has shown. For Gramsci, common sense involved the ‘givenness’ of a set of beliefs and ideas—their apparently self-evident quality. Common sense can be a barrier to political change as it helps cement the idea among poor and oppressed people that their circumstances are inevitable, or even for the best. However, if popular pressure can be creatively combined with a disciplined effort by intellectuals, it may be possible to reshape common sense to support political change.7 In this chapter, we challenge conventional wisdom with a view to developing more convincing accounts of previous experiences, as well as more effective policies for the future.
(p.49) In the battleground of common sense and how it relates to processes of socio-economic transformation in Africa, ideas are weapons. They are used to disarm opposing views, and reflect, serve, and combine with material interests. As Keynes famously put it: ‘Practical men who believe themselves to be quite exempt from any intellectual influence, are usually the slaves of some defunct economist. Madmen in authority, who hear voices in the air, are distilling their frenzy from some academic scribbler a few years back.’8 Donald Trump’s presidency draws on a number of such scribblers, not least Ayn Rand and Samuel Huntington. Meanwhile, economic policy in Julius Nyerere’s Tanzania was heavily influenced by the ideas of the Fabian Society in the UK.9
There is never a straightforward causal relationship between ideas and policy implementation. It is facile to explain Pinochet’s policies in Chile as merely the outcome of ideas put forward by the ‘Chicago boys’, or to account for deflationary policies introduced by the Central Bank of Uganda with reference only to the intellectual influence of the Washington financial institutions.10 Even so, the market for ideas is rigged: a huge amount of resources continue to be invested in legitimizing a specific range of ideas about economic development in Africa and elsewhere. The World Bank and donor governments have invested heavily in ‘capacity building’—a rhetorical commonplace that suggests obvious technical needs while neutralizing the politics of such support.11 The International Food Policy Research Institute (IFPRI) represents a specific example of this. IFPRI’s ideological and political priorities were clearly described by Washington aid insiders in a discussion of the sustained, influential, and well-funded ‘support’ programme in Ethiopia:
[T]he great bulk of IFPRI research (and the training of its research staff) is rooted in neo-classical economics. The upshot is that a significant share of IFPRI’s research in Ethiopia may be fairly characterized as seeking (second-best) solutions to problems created by market failures, while another large share is directed towards the creation and operation of institutions as a possible prelude to first-best solutions.12
One IFPRI-recommended solution was the strengthening of the Ethiopian Commodity Exchange, the results of which were disastrous for Ethiopia’s most important export commodity—coffee.13
(p.50) Another example of international support for ‘capacity building’ is the African Economic Research Consortium (AERC), funded by a variety of major donors, including the United States Agency for International Development (USAID), the World Bank, the Bill and Melinda Gates Foundation, and the UK Department for International Development (DFID). Evaluations of the AERC have noted that its training has a ‘narrow’ focus on a technical and mainstream approach to economics and that a restricted set of macroeconomic issues linked the international financial institutions’ views of stabilization policy. The emphasis is on ‘rigour’—a persuasive rhetorical commonplace in itself, with Thandika Mkandawire describing the AERC as ‘perhaps the most spectacular attempt to transform a discipline ever carried out in any continent’.14 There have been close links between the AERC and the Harvard Institute for International Development, which has promoted a narrow neoclassical version of political analysis, as well as strongly influencing African policy debates by embedding staff in the Mozambican and Kenyan Ministries of Planning.15 The AERC is just one part of a network of economists that helped to disseminate a single and exclusive neoclassical vision of economics within Africa from the 1980s onwards.16 The broader phenomenon is not confined to Africa, as former foreign advisors to Eastern Europe ‘seem to have acquired the same “educated incapacity” … in the same schools and universities, and to be applying in later life the same narrow, abstract and irrelevant methodology’.17
When the AERC did start engaging with ‘political economy’, it did so largely by appealing to public choice theory. The turn to public choice theory as a response to criticism that AERC research lacked political economy is a good example of the instinctive reaction of those working in the economics mainstream, which has:
a curious self-sealing capacity. Every breach that is made in it by criticism trying to let in some air from reality is somehow filled up by admitting the point but refusing to draw any consequences from it, so that the old doctrines can be repeated as before.18
It is not only orthodox economists, though, who are immune to the accumulation of contradictory evidence. Many of those firmly on the other side of the intellectual divide, with their emphasis on structural barriers to development and the dangers posed to poor countries by the capitalist world market, are also remarkably stubborn in the face of inconvenient evidence. One example, discussed in more detail in Chapter 5, involves their shifting theoretical explanations for the impossibility of achieving growth in export revenues, leading to the standard (p.51) policy conclusion—focus on production of manufactures for the domestic market, or on South–South trade.
3.3 Sources of Pessimism in Development Economics
Mission Impossible: Constrained by a Factor Endowment Straitjacket
Economists trained in neoclassical economics and many of those trained in more structuralist development economics, or in neo-Marxist political economy, often share the view that low-income countries have little realistic prospect of developing into industrial economies. This view taps into two streams of common sense, each of which distils complex economic thinking into simple home truths. One draws on the idea of comparative advantage, while the other draws on structuralist (and sometimes dependency theory) ideas about the inevitable consequences of trade between richer and poorer countries.
If ever there were a rhetorical commonplace in economics, it is the idea of comparative advantage, with the authoritative New Palgrave Dictionary of Economics proclaiming it ‘the deepest and most beautiful result in all of economics’.19 Comparative advantage is frequently deployed to deter governments from intervening to accelerate structural change, and even when economists enamoured of the idea recognize its limits, it still leads them to be extremely pessimistic about the prospects of government intervention successfully flouting the principle.20
The principle, deriving from David Ricardo’s eighteenth-century insights,21 suggests that a country should specialize in producing those things in which its production costs, relative to other countries, are cheapest, irrespective of whether it has an absolute cost and efficiency advantage in producing other things. This involves producing and trading in products generated by locally abundant (relatively cheap) inputs, with the least possible use of those inputs that are scarce (relatively more expensive than in other trading partner countries). As suggested by Chandra, Lin, and Wang:
An industry is an economy’s latent comparative advantage if, based on the factor costs of production which are determined by the economy’s endowment structure, the economy could be competitive in this industry …22
(p.52) One influential version of the factor endowment approach to understanding comparative advantage analyses differences between countries in the ratio of skills to cultivable land. According to this analysis, African countries—which are endowed with great expanses of cultivable land but relatively few people with high-level technical skills—ought to specialize in unprocessed primary commodities for international trade. Gradually, productivity in these African countries may rise and, eventually, the stock of skills may increase relative to the stock of cultivable land. Such an analysis is pessimistic about the scope for immediate government intervention to support, for example, primary commodity processing. The Chief Economist of DFID between 2000 and 2005 has underwritten this analysis, arguing that:
The export share of primary products is consistently larger in the land-abundant regions than in the land-scarce regions, as [Heckscher–Ohlin] theory and common sense predict.23
It is easy to point to poor, labour-abundant countries that invested in capital- and skills-intensive production only for their factories to run aground, failing to come close to internationally competitive standards. Using comparative advantage as a cautionary tale when advocating a reduction in barriers to trade, however, is not the same as drawing policy conclusions from, say, a modified Heckscher–Ohlin model built on bizarre assumptions about full employment, capital, and skilled labour immobility. For some economists the non-realism of assumptions is a badge of scientific honour; but it does actually matter that the way the model is constructed does not in any way represent the world it is supposed to capture. ‘When policy conclusions are drawn from such models, it is time to reach for one’s gun’, argued one mathematical economic theorist.24 The trouble is, people do draw policy conclusions from such models, again and again.
Such policy conclusions should also be rejected because of the evidence on trade and economic growth. Even a generous interpretation of the regression literature concludes that ‘the nature of the relationship between trade policy and economic growth remains very much an open question’.25 Unfortunately, a great many economists were unwilling to await answers to this question, and ‘The tendency to greatly overstate the systematic evidence in favor of trade openness has had a substantial influence on policy around the world.’26
It was this attitude of blind faith that, in the 1990s, prompted the World Bank to pressure Mozambique into implementing a radical trade liberalization policy. World Bank economists called for the removal of barriers to exporting (p.53) unprocessed cashew kernels, as well as the lifting of state protection for local factories processing raw cashew nuts. Mozambique had been the largest cashew nut producer in the world and had a substantial processing industry. However, after years of warfare and the exodus of most of the business class, the cashew sector was, by 1992, in crisis. The government’s continued protection of the processing industry, through a ban and then a high tariff on raw nut exports, had failed to generate a regular supply of high-quality processed cashew exports. To World Bank officials advocating trade liberalization, this was a classic case of the dangers of state support for factories in Africa, as well as of the costs of breaking the comparative advantage commandment.
If the government were to get rid of tariffs rapidly, what would happen next? While World Bank economists expected the domestic processing industry to suffer, they did not seem worried by the consequences for large numbers of female factory workers. Any short-run shock would, they believed, be outweighed by the large gains that were expected to flow later through ‘getting prices right’.27 The common-sense logic was that a tariff (or a physical ban on exports) distorts prices, sending the wrong signals about relative abundance and scarcity: in this case, processors in Mozambique could buy raw cashew kernels at an artificially low price.
Removing tariff protection would allow cashew traders acting for international buyers to ‘find’ the ‘right’ price for unprocessed, raw cashews. In a liberalized market, traders would compete to offer poor farmers a competitive price; farmers would respond with an increased supply; and about a million rural households would earn a substantial income by redirecting their unprocessed cashew output away from inefficient local processing factories to meet buoyant demand (mainly from India). This would tilt the economy back against the ‘urban bias’ that Paul Krugman argued had previously prevailed: the export tax ‘almost certainly subtracts from, rather than adds to, the country’s miserably low income’.28 The suggestion that the 10,000 female employees in Mozambique’s processing factories who lost their jobs had been the beneficiaries of an urban bias in policymaking is absurd. These women were not a privileged urban worker elite, isolated from rural households. Starting in the late colonial period, thousands of women had migrated from rural areas to the cities, out of desperation, in search of jobs sorting, shelling, and packing cashews. Once employed, they were commonly stigmatized, labelled as loose women, and they were paid below the minimum wage:
They were suspect in the urban African community as a whole, in part because they were disproportionately divorced, widowed, separated or single. Factory work … was neither sought after nor prestigious.29
(p.54) But reforms invoking the common-sense commandments to ‘get prices right’, liberalize trade, and follow comparative advantage achieved very little. The Bank had not fully analysed the global cashew market or the political economy of a rapidly changing post-war Mozambique. The market was not a perfectly competitive level playing field: other, much more powerful countries were intervening in their own economies to support cashew processing and production; the market was dominated by Indian demand patterns rather than being spread globally; and simple price reforms cannot address the fundamental causes of low output and productivity in agriculture.
According to one assessment of the reforms, the benefits to cashew growing households amounted to an average $5. Margaret McMillan and Dani Rodrik argued that efficiency gains were ‘puny’. If prices did rise and some rural incomes did grow a tiny bit, the overall gains were offset by losses: factory closures and protracted unemployment. They concluded that the ‘aggregate static gains produced by the liberalization were a wash’, that is, they came to nothing.30 A more recent World Bank research paper reviewing the Mozambican cashew drama found that real producer prices by 2008 were about the same as before the reforms. Output per cashew tree remained generally very low, though higher among the larger-scale producers. The larger-scale capital-intensive processing factories went out of business. A number of smaller, more labour-intensive processing firms did slowly emerge in the 2000s, though they depend heavily on public and external support. The reforms also seem to be promoting a ‘re-cottagization’ of cashew processing and there are claims that workers are paid below minimum wages.31 The fundamental problem with the sector remains the low level of output, the poor quality of the cashews produced, and low productivity.32 Unshelled cashew export volumes have never recovered from a post-war peak in 2001(Figure 3.2) and revenues have been very volatile.33
Mission Impossible: Paralysed by the Prebisch–Singer Hypothesis
If comparative advantage incites pessimism among neoclassical economists, the pessimism of many other development economists springs from critiques of comparative advantage. The source for much of this pessimism lies in the work of structuralist economists, who nail a version of the Prebisch–Singer thesis to the door of neoclassical orthodoxy. The thesis, which claims there is a secular decline in the net barter terms of trade between primary commodities and manufactures, (p.55) is based on empirical claims about patterns of income, demand, and production globally. In 1949, Hans Singer (at the UN Department for Economic Affairs) and then Raúl Prebisch (at the UN Economic Commission for Latin America and the Caribbean) highlighted important trends in the world economy: poorer countries tended to specialize in the production of primary commodities; richer countries in manufacturing. As incomes rise, people typically spend a smaller share of their income on basic (food) goods. In a growing global economy, this means demand for manufactures tends to be strong, while the demand for primary commodities remains relatively weak. Put differently, there is a relatively high income elasticity of demand for manufactures, and a relatively low income-elasticity of demand for primary commodities. Consequently, there is upward pressure on the price of manufactures relative to commodity prices.
An excess supply of labour in poor countries, combined with the greater bargaining power of trade unions in richer societies, adds to this upward pressure. So, in another pessimistic account of the export prospects of developing countries, the power of trade unions in advanced capitalist countries means that most gains from productivity increases and trade will be distributed unevenly: the advanced capitalist country will exploit primary commodity-exporting countries by trading manufactured commodities (embodying small quantities of high-paid labour) for primary products (embodying much greater quantities (p.56) of low-paid labour).34 Additional explanations for declining primary commodity prices point to historic patterns of technological change—including the introduction of synthetic substitutes (for example, Madagascan and Ugandan vanilla beans, substituted in many confectionary factories by synthetic vanilla essence, or Tanzanian sisal, which was hit by the introduction of polypropylene substitutes in the 1970s)—as well as to declining prices received by primary commodity exporters caused by the growing power of oligopsonistic purchasers in these commodity markets.
Perhaps the most dramatic argument about the disadvantages of primary commodity exports claims that they are associated with declining domestic food grain availability. Distinguished Marxist scholars have argued that:
The classic case in more recent years of production for the metropolitan market undermining food security relates to Africa. The integration of Africa as an agricultural exporter to the advanced capitalist countries was greatly strengthened under the diktat of the Bretton Woods institutions at the beginning of the 1980s … because of the adverse shift in the terms of trade for their primary commodity exports vis-à-vis manufactured goods … This thrust on exports of agriproducts meant a shift away from food crop production.35
Similar arguments have been used by scholars, political leaders, and non-governmental organizations (NGOs) to support a common-sense policy promoting food self-sufficiency. However, the African evidence on recent and historical trends in agricultural exports, food production for the domestic market, and nutritional status, provides little support for such policies (see Chapter 9).
Those pessimistic about the export prospects of developing countries also argue that
in their eagerness to industrialize, for which foreign exchange was needed for importing machinery and other capital goods, each of the newly decolonized economies of the periphery competed with the others to push out as much of its traditional primary commodity exports as it could. The suppression of nominal prices of tropical commodities (and the adverse movement in the terms of trade) … continued even after decolonization for this reason.36
A similar argument goes that, since an increasing number of low-income economies have tried to emulate East Asian success in exporting labour-intensive or low-tech manufactured exports, the market prices of these manufactured goods have fallen relative to the high-tech manufactures produced in advanced capitalist (p.57) economies.37 The United Nations Conference on Trade and Development (UNCTAD) has for many years repeated the view that ‘It is well-known that export-led growth strategies must sooner or later reach their limits when many countries pursue them simultaneously.’38
The policy implication of this dismal view of the external environment is that a greater proportion of production in developing countries should be geared towards domestic, rather than world, markets. UNCTAD theorists, like the dependency theorists in 1970s Tanzania, hope that this ‘rebalancing’ of production will also reduce imports and the balance of payments constraint on growth (discussed further in Chapters 4 and 5).39
Although ‘few hypotheses in development economics have been studied more intensively’ than the Prebisch–Singer suggestion that there is a long-run negative trend in the prices of internationally traded primary commodities relative to those of manufactured products, ‘findings have varied to such an extent that there is still no consensus in the literature’.40 There is also no clear consensus on the empirical case for greater diversification among primary commodity producers (this being a key policy response to anxieties about the terms of trade). International Monetary Fund (IMF) (and other) analysis of the evidence concludes that ‘higher incomes are not typically associated with greater diversification for commodity exporters’.41
Mission Impossible: The Global Business Revolution
Some dependency theorists introduce yet another set of pessimistic arguments, highlighting the domination of research and development (R&D) by developed economies. As a consequence, developing economies in sub-Saharan Africa cannot produce the most globally competitive or innovative manufactures, as the leading corporations in developed countries have created unbreachable barriers to entry and ‘captured rents’.42 A less pessimistic view of the prospects for some developing countries describes ‘the global business revolution’.43
This revolution, beginning in the 1980s, unleashed powerful forces of concentration, leading to cascading layers of suppliers being controlled by ‘system integrator’ firms. Supply relationships are structured in what are known as global value chains or global production networks, with the relevant literature often emphasizing how very difficult it is for firms in developing countries to climb the ladder of global production due to the barriers imposed by firms higher up. (p.58) During an earlier period, a few national economies were able to achieve industrialization, using protectionism and state intervention; they then tried to ‘kick away the ladder’ of industrialization they had previously climbed. Now, first-mover ‘system integrator’ firms grease the rungs so that those lower down can rarely, if ever, get a purchase on the ladder’s higher reaches.
The massive supermarkets at the head of ‘buyer-driven’ value chains, or the firms at the summit of producer-driven chains, have enormous power, often blurring the lines differentiating one firm from another. For example, it is not always clear to what extent a Kenyan farm enterprise growing ornamental plants for export to the Netherlands is independent from the Dutch corporation that controls much of the genetic R&D and input supply, and that buys almost all of the farm’s output. There is, though, evidence of successful national firms within developing countries ‘climbing the ladder’,44 including Thai firms like Thai Union Frozen Products (TUF) (tuna and other seafood) and Charoen Pokphand Foods (CPF) (poultry) that have become global leaders, or the Brazilian JBS (the largest meat-processing firm in the world).45
Dammed If You Do
Impossibilism appears in many guises in development economics. Two further forms it takes can be seen in the resource curse literature and in the injunction to avoid large-scale infrastructural ‘mega-projects’. In Chapter 6 we address the gloomy argument that mega-projects—such as the proposed Grand Inga Dam on the Congo River—are ‘over budget, over time, over and over again’.46 As in the cases already mentioned, we argue that this stance creates a cast-iron law out of evidence that may not be universally relevant, and involves a partial view of the history of such projects. As regards the resource curse, much has been written on the subject, so we will make only brief comments here.
The resource curse argument, a version of primary commodity pessimism, itself has variants. One—the original—is a straightforwardly economic argument, which proposes that an abundance of natural resources leads to ‘Dutch disease’. A resource boom leads to a foreign exchange bonanza, which pushes up the value of the exchange rate in the country in question. An appreciating currency then prices the country’s other exports out of international markets and, because they are no longer competitive, these other exports collapse, together with their associated jobs, fiscal revenue, and potential for domestic linkages. Dutch disease (p.59) is exacerbated if we acknowledge that many natural resources—oil being the classic example—may be ‘niche’ or ‘enclave’ activities. These are geographically concentrated (and may even be offshore); are hugely capital intensive; rely on highly specialized expatriate expertise; and generate little by way of domestic employment or linkages. Oil production in Nigeria is often regarded as a classic case of Dutch disease in Africa.
The other variant of the resource curse argument involves a slightly more complex political discussion. Here, natural resources offer ‘easy money’ to greedy governments, generating a large flow of royalties into the state budget from highly concentrated sources. A government in this position need not bother with the politically difficult and administratively challenging business of taxing the population and providing public goods in return. With less connection to the population and given the scope for resource rents to become something of a ‘honey pot’ attracting all manner of rival claimants, this is a potential recipe for corruption, extreme inequalities, and violent conflict.47
There is nothing particularly African about the resource curse: after all, the disease was first identified in studies of the booming price of wool in Australia in the first half of the nineteenth century and is named for the decline in manufacturing in the Netherlands after the discovery of natural gas in 1959. It is also worth noting that neither country’s experience of the Dutch disease destroyed the longer-run capacity of their economies.48 Routes to economic collapse take many strange turns and are not inevitable, even in Venezuela, which managed to combine oil production with diverse industrial output and gross domestic product (GDP) growth for rather a long time.49
Within Africa, South Africa’s very rich resources of minerals and energy have been associated with an inadequate growth rate of an unusually import-intensive manufacturing sector; but manufacturing output did grow (and diversify) to some extent and in some periods, when encouraged by state intervention. The complex political economy of exchange rate and macroeconomic policy formulation in South Africa—and of investment in its state-owned enterprises (SOEs)—cannot be understood within a simple resource curse framework.50
Finally, Norway’s history should give pause for thought to the geographical determinists who have written about Africa’s poor and cursed prospects. Norway industrialized ‘through the development of geographic features usually considered adverse’ and on the basis of ‘the country’s greatest natural resource’ (at the time): waterfalls. Hydropower fuelled the rapid development of a multitude of industries, from electrochemical and electro-metallurgical factories to pulp and cellulose plants.51 Much later, Norway managed to develop a successful oil sector—albeit with strong upward pressure on the exchange rate—while preserving the country’s (p.60) remarkably egalitarian governance of natural resources and supporting a wide array of technically sophisticated industries.52
3.4 Naive Optimism and Capitalism with a Human Face
Impossibilism often involves insisting that efforts to accelerate structural change will fail because of the binding constraints afflicting all late-late industrializers. A very different approach is the ‘possibilism’ of Albert Hirschman, who thought that the field of development was ‘an exceptionally good hunting ground for exaggerated notions of absolute obstacles, imaginary dilemmas, and one-way sequences’. In Chapter 2, we showed how many econometricians have claimed to explain Africa’s development ‘failure’ by referring to a miscellaneous mix of obstacles, encapsulated in the ‘Africa dummy’ variable.53 The essence of the possibilist approach, by contrast, ‘consists in figuring out avenues of escape from such straitjacketing constructs in any individual case that comes up’.54
Hirschman’s possibilism is not, however, to be taken as a licence for naive optimism. Widening the limits of what may be thought possible comes at the cost, Hirschman accepted, of ‘lowering our ability, real or imaginary, to discern the probable’.55 There are clear indicators of when possibilism is in danger of spiralling into naive optimism. When the list of ‘ifs’—of all the conditions that would need to be in place for a scheme to work—stretches too far, then the project or policy may be unrealistic.
This section highlights two strains of common sense that exemplify naive optimism. One has already been alluded to: the belief that reading the runes of current comparative advantage will unleash future economic development and structural change. The other is the belief that the best way to stimulate growth, expand employment, and reduce poverty is by promoting small farms, as well as small, or even micro, enterprises. Both are misleading forms of fantasy.
Although absolute advantage underpins most growth in world trade, African policy officials are misguidedly encouraged to commit to the principle of comparative advantage. Naive optimism characterizes the insistence of neoclassical economists that merely following the pointers of comparative advantage or ‘latent’ comparative advantage will unleash a smooth process of ‘convergence’, ‘catching up’, and development. Such naive optimism also characterizes the faith of other economists that policies such as additional South–South cooperation or a new development bank will magically allow poorer economies to catch up by trading among themselves.
(p.61) Additionally, at a time when global capitalism is increasingly dominated by massive corporations acting as ‘systems integrators’ for internationalized production chains, a great deal of policy advice (and resources) continues to push African policymakers towards strategies built around tiny enterprises. Advisors and economists who usually claim to be fiercely opposed are again united, this time in supporting ‘bottom-up’ approaches to structural transformation. One version promotes faith in decentralization (in the hope it will empower local people and increase responsiveness to the needs of local communities), alongside the belief that dynamic smallholder farmers, micro-finance, and micro-credit will create a breeding ground for female and other entrepreneurs. Another version is influenced by the analytical presumption in neoclassical economics that everything can and should be rooted in individual choices—the so-called micro-foundations of macroeconomics. One of the most pervasive nuggets of conventional wisdom in political discourse around economics is the idea that an economy is similar to, and therefore should be run like, a household.
Bottom-up policy prescriptions are consistent with the widespread but naive notion that capitalism can be given a makeover by enlightened leaders committed to the liberal values of equality, solidarity, and citizenship—‘capitalism with a human face’. In the late 1980s, influential development economists scoured the world in search of a country successfully pursuing macroeconomic policy that ‘fully protects the human dimension’.56 For these individuals, capitalism’s ‘human face’ represented only good things; a social-democratic visage unscarred by conflict, crisis, or exploitation. In effect, it is ‘photo-shopped’ or Barbie doll capitalism: filtered, prettified, with the contradictions smoothed and airbrushed away.
To imagine a fully inclusive, appealing ‘capitalism with a human face’ is to be fooled by Oscar Wilde’s protagonist Dorian Gray, whose features remain exquisitely beautiful, young, and pure. A painting of Dorian, hidden away in a locked attic, does change and age, unlike his human face; it bears the imprint of his betrayals and crimes. When Dorian finally stabs the portrait with a knife, his servants, hearing a scream, go up to the attic and find a painting of a beautiful young man, as well as their master—with a knife through his heart.
He was withered, wrinkled, and loathsome of visage. It was not until they had examined the rings that they recognized who it was.57
We believe that understanding economic performance in Africa will be improved by visits to the attic, by studying both sitter and portrait.
Strains of gloom and optimism run parallel through much of the history of economic thought. What separates reflex pessimism and naive optimism is the mix of assumptions made. Assumptions are a set of conditions—‘ifs’—that allow a model or theory to hold together logically and generate predictions. Despite often being counter-intuitive, these ifs are what underlie much of what is accepted as common sense. They require closer scrutiny by policy officials.
The formal models that take Ricardo’s original insight about comparative advantage and insist it should inform contemporary policy are built on a string of ifs. Comparative advantage produces optimal outcomes and a ‘theoretically harmonious world’58 if there is no involuntary unemployment; if capital is immobile between countries; if prices adjust instantly, automatically, and painlessly following trade liberalization; if the only relevant gains from trade liberalization are static gains; if the flow of ideas and investment and economies of scale are irrelevant and there are constant returns to knowledge accumulation. Some models go further. One, for example, predicts universal gains from trade liberalization if knowledge is a non-rivalrous public good (that is, all technical knowledge is freely available to everyone), and if trade flows facilitate full diffusion of knowledge.
The ifs are like advertising small print—few people bother to examine them closely. These, however, are not minor details, but are at the very core of the models built around them. A substantial body of critical literature addresses these conditions, pointing out that, in reality, capital flows in ways that differ radically from the assumptions made by Ricardo.59 The critical literature also argues that African policy officials should not be expected to sign up to a deal based on the fanciful supposition that knowledge is a non-rivalrous public good, or that payments adjustments are quick, automatic, and without costs. In a global economy in which many sectors are dominated by huge corporations with integrated systems of suppliers, knowledge is often jealously guarded within supply chains. A single Illy coffee capsule, for example, has seven patents in its design.
In a world economy where the ifs of comparative advantage theory are unfulfilled wishes, absolute advantage dominates. Relative prices are not enough to propel changes in the structure of output. The relative price of land, labour, and capital—the keys to the comparative advantage doctrine—matter less where rents from ‘rivalrous’ knowledge are protected, and where competitiveness owes as (p.63) much to state interventions creating the institutions determining productivity as to relative ‘factor’ costs.60
Meanwhile, many people nervous about global economic integration instead pin their hopes on South–South cooperation and regional economic communities (RECs). They expect trade among ‘Southern’ economies to cut the enduring bonds of colonial history, thereby allowing for greater self-reliance and policy autonomy. The swell of support for South–South cooperation and initiatives such as the AfCFTA is partly based on a pessimistic view of possible integration into the wider world economy. Such pessimism, however, quickly turns to optimism once the prospect of escape from the clutches of the West (or North) is clear. Strains of economic nationalism, pessimism about the prospects of wider global economic integration, pan-African and broader ‘South–South’ solidarity, and orthodox economics have been woven together to produce a blanket optimism about prioritizing African economic integration.
The AfCFTA is expected to ‘boost intra-African trade, stimulate investment and innovation, foster structural transformation, improve food security, enhance economic growth and export diversification’.61 Accelerating intra-African trade is regarded as especially beneficial because, while ‘Africa exports mainly commodities to the rest of the world, intra-African trade displays high concentrations of value-added products (and services)’.62 The AfCFTA is expected to ‘leverage Africa’s progress towards attaining several Sustainable Development Goals’, such as poverty elimination and gender equality.63 Specifically, the hope is that regional integration will allow for the exploitation of economies of scale, will encourage investment, and will stimulate productivity-enhancing competition. Regional value chains will emerge as a result. All these developments combined will, in turn, lead to poverty reduction.64
These visions of African economic integration rest on the same naive and unrealistic ifs underpinning comparative advantage doctrine. For example, while the case for integration may be made by spinning out predictions from a static computable general equilibrium model, these models often—as the authors of one such exercise note in a footnote—ignore ‘missing’ or ‘inefficient’ markets. (In other words, these models ‘work’ if there are complete and perfectly competitive markets.) Further ifs include: the AfCFTA will achieve all its goals if measures are taken to address regional inequalities, if governments implement the agreements they sign up for, if entrenched interests that have made it difficult for African RECs to succeed are set aside. The list goes on.
(p.64) The push for greater regional integration in Africa attempts to make a virtue out of a weak performance in wider global international markets. Figure 3.3 illustrates this failure, showing that despite an increase in the value of exports from African countries since the early 2000s, their share in global exports has remained below 6 per cent. In fact, their share is now smaller than in the late 1990s. Even Africa’s primary commodity exports (regardless of whether oil exports are included) have shrunk to less than 5 per cent of global primary commodity exports (see Figure 3.3b), meaning that other parts of the world became successful exporters of precisely the commodities regarded as unpromising by some advocates of intra-African trade. Such advocates optimistically believe that, unlike trade with the rest of the world, trade with other African (p.65) countries will avoid this gloomy prognosis, as new opportunities for trade in other commodities—including more processed exports—will emerge.
One reason to query high expectations for intra-African trade is the ‘dismally poor implementation record’ of regional integration agreements.65 Despite decades of negotiations and agreements within subregions and RECs in Africa, intra-African trade remains a tiny proportion of the continent’s overall trade. As Figure 3.4 shows, even taking into account some recent intra-trade growth, it still only accounts for a fifth of total trade.
In short, while African regional economic integration and greater intra-African trade may be rhetorically appealing on grounds of economic nationalism or South–South solidarity, as a blueprint for accelerated development it is a fantasy—based on exactly the same theories and conditional assumptions as a push for global free trade on comparative advantage grounds.
Small and Bottom Up is Beautiful: Micro-Finance, Entrepreneurship, and Cooperative Development
Another variety of common sense is the belief in the individual and the small-scale as wellsprings of development, technological dynamism, and structural change. (p.66) Supposedly competing views converge here too, for example in arguments for financial inclusion and micro-credit, for entrepreneurialism, and for prioritizing small farmers.
Many politicians have made use of the idea that ‘an economy is just like a household’ in order to justify the economic policies they favour. Electorates are routinely told that just as households need to be careful not to spend beyond their means, so countries need to adopt the same attitude. This means national debt should be reduced, and public and private borrowing (and spending) reined in. This was at the heart of Margaret Thatcher’s rhetoric about fiscal policy in the UK in the 1980s. Meanwhile, in the wake of the Global Financial Crash of 2008, German Chancellor Angela Merkel used the common-sense maxim that ‘One should simply have asked the Swabian housewife.’ This was because every housewife ‘knows that we cannot live beyond our means’.66
This is perhaps the most striking case of a ‘common-sense’ idea in economics: apparently obvious and sensible, easy to understand, powerfully influencing policy debates, while at the same time deeply ideological (in Marx’s sense of ideology as an inversion of reality). An economy is not like a household. It is not even like a business firm. There are two main ways in which this Swabian housewife logic fails. First, compared to private households, governments have many more and effective tools at their disposal to manage debt. Furthermore, public sector debt can be key to sustaining and expanding economic activity (the returns to which can be used to make repayments). Second, the Swabian housewife logic runs into a fallacy of composition. As Keynes pointed out, calling it the ‘paradox of thrift’, an individual may increase saving by withholding spending, but if all individuals in an economy raise their saving simultaneously then there will be a decline in output and income, meaning aggregate saving will be unchanged (at best).
Methodologically, this logic is linked to the idea that macroeconomics is simply the aggregation of individual micro-economic choices. Together, these form the ‘micro-foundations’ of macroeconomics, with causation running up from these micro-foundations to the overall macro-economy. What is missing from this vision is the possibility of ‘downward causation’: the idea that the structural features of a society, including macroeconomic change, can exert a huge influence on individual behaviour (it also leaves out the possibility that ‘macro’ and ‘micro’ can be jointly affected by other factors).
In development economics, this emphasis on the individual as the foundation of the economy has wider implications, forming the backdrop for strongly held beliefs about how economic policies should be designed to support plucky individual development efforts in the smallest production units. African economies are, in fact, defined by the small scale of most of their productive units, being (p.67) dominated (numerically) by tiny firms and very small farms. By one count, approximately 80 per cent of farms in Africa are small, with the conventional wisdom—discussed at length in Chapters 7 and 9 – being that ‘with few exceptions, land resources are distributed in a relatively equitable manner’.67 When we began fieldwork in Ethiopia in 2000, the internationally celebrated general manager of the country’s most important coffee cooperative union spun us the same story he had offered to other useful idiots, that all farmers producing both Fairtrade and other coffee had one hectare.68 It took 15 minutes of interviews with members of a showcase primary cooperative to establish how deeply misleading this claim was: there was very clearly a wide size distribution of farms among cooperative members—from less than 0.5 hectares to more than 35 hectares. Some members only had a few coffee bushes while others were hiring wage labour to cultivate their large coffee farms.
African businesses are predominantly very small, with small and medium enterprises (SMEs) said to account for 95 per cent of all African firms. In one sample of countries, the most common size category (the modal group) of formally registered enterprises comprised those employing between zero and nine people.69 This understates the reality, however, as it ignores the proliferation of tiny micro-enterprises, most of which are not formally recorded in enterprise surveys and national accounts. The Rwandan economy, for example, ‘remains dominated by a mass of single employee micro-enterprises’.70 Many people interpret this as showing that these micro- and small farms have ‘the greatest development potential’. African heads of state have pledged to introduce ‘specific policies and strategies to assist traditional small farms in rural areas’.71 Similarly, many people argue that ‘SMEs are central to wealth creation by stimulating demand for goods, investment and trade’ in Africa, through their contribution to job creation, discovery of new markets, and capacity for innovation.72
Such a ‘pervasive belief in the importance of small businesses for growth and economic development’ has encouraged wanton distribution of resources—by governments, donor agencies, and philanthrocapitalists—to SMEs or micro-, small, and medium-sized enterprises (MSMEs), including generous funding for training, financial access, and business development schemes.73 This belief also lies behind the commitment to entrepreneurship programmes by some African leaders. For instance, Rwandan President Paul Kagame places responsibility for economic development squarely on the country’s individuals (specifically, individual entrepreneurs), arguing that ‘entrepreneurship is the surest way for a nation … to develop prosperity for the greatest number of people’.74 The (p.68) Hanga Umurimo strategy has encouraged the National University of Rwanda, the Private Sector Federation, the International Finance Corporation (IFC), and a selection of NGOs to back Kagame’s enthusiasm.75 The strategy aims to inculcate an entrepreneurial spirit among Rwandans, and exemplifies the assumption among many policy officials that changing attitudes and customs is a prerequisite to structural and economic change. It was precisely this view that Albert Hirschman challenged, arguing that often the direction of causality ran the other way round.76
Where it is clear that there are many obstacles preventing small farms and enterprises from succeeding, the next-best solution advocated by many is supporting the clubbing together of small producers to form associations, credit schemes, or producer cooperatives (or SME clusters). When economists argue for cooperatives and group credit, they often emphasize that this can help with pooling transaction costs and enabling economies of scale. Furthermore, economists in the ‘information theoretics’ tradition argue that peer monitoring in small lending groups can address the information problems (chiefly concerning creditworthiness) that beset rural finance and that often explain the high administrative costs of loan disbursement and steep interest rates charged to poor rural borrowers.77 In such groups, borrowers collectively bear the burden of selection, monitoring, and enforcement that would otherwise be faced by the lender.
Other, often sweeping, claims are sometimes made in favour for cooperatives: that they are a bulwark against communist uprisings,78 that they protect small producers against venal middlemen, that they are ‘pro-poor’, and that they are ‘empowering’, especially for women.79 It has even been argued that rural cooperatives
contribute in significant ways to … ensuring environmental sustainability, tackling the HIV/AIDS and malaria pandemic, and mainstreaming gender.80
For some naive optimists, rural cooperatives are a haven for women living in a ‘post-wage’ society and signpost possibilities of a life ‘outside of wage labour and production circuits’.81 Bina Agarwal is perhaps the most prominent advocate of the potential benefits of group farming for women, though she does acknowledge that ‘there has been no systematic study of the impact of group farming on women in developing countries, based on carefully collected quantitative and qualitative data’.82
(p.69) Our own research into the political economy of rural cooperatives in Africa suggests such grand hopes are ludicrously unrealistic. Ledgers from coffee cooperatives in Ethiopia and Uganda explode rural egalitarian myths.83 These are producer organizations that serve the interests of a very small minority of members, namely the farmers with the largest plots of land, who can obtain the greatest benefits from access to cheaper fertilizer and higher priced market opportunities. Cooperatives of Ethiopian coffee producers and Ugandan tea and coffee producers—including those certified by Fairtrade—probably increase rural inequality. Because they draw additional resources (e.g. in the form of ‘ethical’ trade price premiums), these disproportionately swell the income of the minority of member-producers who sell the bulk of the cooperative’s ‘certified’ output. These cooperatives also often fail to serve the interests of their poorest members, let alone those excluded from joining in the first place.
Powerful ideas about the virtues of cooperative and small-scale farming enterprises are sustained not on the basis of evidence, but on the breeze of modern economic theory and the idealism of many working in the NGO sector. Nowhere is this clearer than in the arguments promoting small family farms, which claim there is an inverse relationship between farm size and productivity: small farms are deemed more efficient (crucially, on a measure of productivity per hectare rather than labour productivity) than larger farms. One explanation for this concerns the difficulties larger farmers have in monitoring labour on great tracts of land.
We spend more time examining these arguments in Chapter 9, arguing that African policy officials ought not to design agricultural policies on the basis of such an empirically weak claim. Here, we stress how the faith placed in the efficiency of small farmers by conventional economists—many occupying senior positions in Washington institutions and the United Nations (UN)—fits surprisingly well with the claims and preferences of a much wider range of intellectuals. These include those in pro-peasant agrarian social movements and NGOs who lobby against ‘big ag’ from bases in university-based think tanks. The Oakland Institute, for example, has published not very rigorous research on large farms in the Democratic Republic of the Congo (DRC) (and Ethiopia), lamenting the devastating economic and social costs of ‘land grabs’ by agribusiness, while dispensing patronizing policy advice: ‘Congolese farmers should be considered innovative and hard-working entrepreneurs.’84
For many very small-scale farmers in Africa, the hard work and ‘efficiency’ lauded by so many economists is that of eking out a limited existence while failing (p.70) to avoid chronic undernutrition. It is a form of ‘efficiency’ that is irrelevant for structural change, long-run development, or reducing the exploitation of women. For example, a large proportion of Ethiopian coffee farmers cultivate probably less than a quarter of a hectare.85 A rapid and complete overhaul of Ethiopia’s ageing coffee plant stock is urgently required, but the cost and logistical demands of replanting on millions of dispersed, minuscule plots is prohibitive. Very poor, very small-scale farmers cannot even afford to introduce low-cost changes—for example, in planting density, mulching, weeding, and in how beans are harvested and cared for after picking—that are known to generate relatively quick gains in yield. Newly planted coffee takes a long time (about four years) to produce significant increase in yields—too long for those living hand to mouth.
Such tiny plots cannot be the basis for ‘catching up’ with the increasingly complex, high-productivity, capital-intensive production that dominates global agriculture (and the world coffee market). Colombia, for example, has been much more successful in coffee production than Ethiopia, achieving substantial yield increases and engineering a radical restocking of the national stock of coffee trees. But Colombia’s success has not been achieved by ‘micro’ coffee farmers, whose prospects are regarded as ‘bleak’. Farms growing less than one hectare accounted for only about 6 per cent of the total area producing coffee in Colombia in 2007.86
Small and micro-enterprises, as well as the urban self-employed, fare no better than very small farms, a situation not confined to African or other developing economies. In the UK, for example, self-employed people have lower median earnings than employees do, while employees in new small firms have jobs that are more volatile, less productive, and less well paid; huge state subsidies encouraging new firms have not succeeded in creating sustainable new jobs.87 The UK incidence of self-employment is relatively high, but research finds no evidence that changes in self-employment rates over time have any bearing on shifts in real GDP.
Failure rates for small businesses are high internationally, with about half of new small businesses in the USA and the UK failing within five years of starting up. In Africa, about half of new small businesses fail within three years, with one study finding that five out of seven new businesses fail within a year. South Africa appears to have particularly high rates of failure for new small start-ups, with estimates ranging between 50 and 95 per cent, depending on the sample and sector.88 Even if small firms in Africa do manage to survive (at least initially), they tend not to grow.
(p.71) SMEs that do survive in Africa usually make only a modest contribution to the growth process. The value added of firms employing less than thirty workers in Kenya (where very few start-ups last even a year) is lower than in larger firms, with smaller firms found to be less effective in absorbing new technology.89 Small firms in Africa export less than larger firms too, so it can be argued they are less useful for macroeconomic balance and sustained growth (see Chapter 4).90 Furthermore, evidence from developing countries more broadly suggests that workers in small enterprises have less access to decent on-the-job training than employees of larger firms.91 We will discuss other research on the contribution of small firms to sub-Saharan African employment in Chapter 8.
The policy response following the rapid demise of SMEs is often not to query this policy fad or make more strenuous efforts to resist the lobbyists for entrepreneurs. Instead, the same tired reform proposals are wheeled out, from improving the ‘doing business’ climate to throwing even more resources into business support services and training. This is despite substantial resources having already been poured into programmes supporting SMEs in low- and middle-income countries, with underwhelming results.92
The results of the Hanga Umurimo programme in Rwanda are typical. The idea was to identify people with the ‘right aptitude’ and expose them to IFC-backed entrepreneurship training, thereby nurturing ‘good business ideas’. The reality was that hardly any wage employment was created. Many beneficiaries thought the loans were actually grants, resulting in low repayment rates. Also, the vast majority of micro-credit loans (84 per cent) intended to support entrepreneurial activity in Rwanda are actually taken to smooth consumption, which is in line with wider evidence internationally.93
What goes unnoticed are the ‘top-down’ policies that have made a difference to the success and wider role of SMEs in some countries. Successful state promotion of SMEs has not been merely facilitative, but has introduced red tape and violated market theory. SMEs (variously defined) in South Korea, Taiwan, and Japan have advanced fastest where they are in concert with larger firms through various forms of subcontracting. Between 1972 and 1981, the South Korean government prioritized the promotion of heavy and chemical industries, which involved support for very large firms. At the same time, the government did support the SMEs supplying these large industries; it increased this support through the liberalization of the 1990s and into the 2000s.94
Our conclusion is not that a new mix of top-down state interventions combined with the embrace of global markets and deals with large firms will promote development in all contexts. Nor is our main aim the exposure of sworn ideological enemies appearing to advocate shared ‘common-sense’ policies. Rather, we have sought to identify a problem: the fog of development economics.
In subsequent chapters we aim to cut through the murk and, above all, encourage policymakers to query the theory and evidence underpinning common-sense views, whether these are imposed by creditors, donors, and financial institutions, popularized by well-meaning (bien pensant) NGOs, or advocated by home-grown nationalists. Such views are particularly unhelpful to policymakers when underpinned by impossibilist assumptions and/or tinged with naive optimism.
(8) Keynes (1936: 383).
(15) Bates (2017).
(31) Penvenne (2015: 229).
(33) Food and Agriculture Organization Statistical Database (FAOSTAT).
(34) Smith and Sender (1983) criticize this form of unequal exchange theory.
(36) ibid. (2017: 107).
(40) Cashin and Pattillo (2006: 845).
(42) Kaplinsky and Morris (2008).
(46) The best-known and most persistent academic critic of mega-projects is Flyvbjerg (2011). NGOs such as the Oakland Institute and International Rivers (https://www.internationalrivers.org/resources/the-new-great-walls-a-guide-to-china%E2%80%99s-overseas-dam-industry-3962) also publicize pessimistic views about the benefits of large-scale infrastructure projects in Africa.
(55) Hirschman (1971: 28).
(61) UNECA, African Union, and African Development Bank (2017: xi).
(63) UNECA (2017: 11).
(64) Mold and Mukwaya (2016).
(68) He has a starring role in an influential film promoting a prettier model of trade:
(72) Muriithi (2017: 9).
(79) Wedig and Wiegratz (2018: 349).
(86) García-Cardona (2016: table 5.1 and 183).
(89) Muriithi (2017).
(92) Among recent reviews see Bloom et al. (2014); White, Steel, and Larquemin (2017); Piza et al. (2016). On the consequences and risks of fiscal and political decentralization in Africa see: Green (2008) and Erk (2015).