Theme: This ARI addresses the analytical and empirical links between resource extraction, governance and development, with a focus on the resource-curse thesis. The rent curse is rooted in policy failure, which the theory of rent cycling attributes to the impact of rent on elite incentives and also on development trajectory. The paper provides some examples of conditions that have facilitated this process in the context of Sub-Saharan Africa.
Summary: The so-called resource curse is part of a broader rent curse that can be triggered by regulatory rent and foreign aid (geopolitical rent) as well as by natural resource rent. Although resource-driven growth challenges macro management due in part to commodity price volatility, the policies required to limit adverse impacts such as the Dutch disease effects are well known. Consequently, the rent curse is rooted in policy failure, which the theory of rent cycling attributes to the impact of rent on elite incentives and also the development trajectory. The theory argues that high rent incentivises the elite to deploy rent through patronage channels for immediate personal enrichment, but this represses markets and distorts the economy, which lowers investment efficiency and triggers a growth collapse that is protracted because rent recipients resist economic reform. The risk of the economy falling into this ‘staple trap’ development trajectory increases in the presence of point source resources (notably minerals), statist policies, ethnic tension and democracy that is youthful. This implies that initial conditions were unpropitious for most African countries (especially the mineral economies) at independence because they were typically ethnically-mixed, resource-rich, new democracies with a predilection for fashionable state control. Most African economies did experience protracted growth collapses from the 1970s but some, like the Ivory Coast and Kenya collapsed later than most. Just two economies, Botswana and Mauritius, evaded the curse. Botswana’s conditions appear unique but Mauritius’s pursuit of a dual-track growth strategy offers a useful model for other African governments.
Analysis
Introduction
This paper draws upon the literature to analyse, first, why most resource-rich African economies have under-performed since independence and, secondly, how to improve their development outcomes. Early explanations for the resource curse stressed the Dutch disease effects arising from the over-rapid domestic absorption of commodity rents, which leaves economies vulnerable to negative trade shocks because it shrinks the share of non-mining tradeables in GDP (Sachs & Warner, 1995). However, the policies required to manage commodity-rent soundly are well known, and they centre on establishing funds to sterilise the rent in order to smooth the revenue flow so that the rate of absorption matches domestic absorptive capacity (Gelb et al., 1988). Consequently, the explanation for the resource curse lies in policy failure and its causes.
This paper draws upon the emerging theory of rent cycling, which synthesises the resource-curse literature, to argue that high rent can adversely impact both the incentives of the elite and the development trajectory. First, high rent incentivises the elite to cycle rent through patronage channels for immediate personal gain rather than through markets to promote long-term economic growth. Secondly, the resulting misallocation of resources lowers investment efficiency and triggers a growth collapse. Growth collapses are protracted because once rent recipients are established they resist economic reform because it shrinks their scope for rent extraction. This implies that for it to be successful, economic reform requires an explicit political component to align the interests of the elite with those of the majority in providing public goods and efficiency incentives to make the economy grow through the long term. This paper analyses the development since independence of Kenya, the Ivory Coast, Botswana and Mauritius in search of a blueprint for reform.
The structure of the paper is as follows: it first explains African under-performance with reference to rent cycling theory, then it analyses the four case-study economies in order to identify policies to limit the adverse effects of high rent and, finally, it summarises the conclusions.
Rent Cycling Theory and Growth Collapses
The emerging theory of rent cycling suggests that the scale of rent relative to GDP impacts both the incentives of the elite and the development trajectory (Auty, 2007). Critically, low rent incentivises elites to grow the economy because that increases the level of taxation, which in the absence of sizeable rents is the principal source of discretionary expenditure, and one that the elite frequently benefit from disproportionately. In contrast, high rent deflects elite incentives towards cycling rent in order to boost elite patronage from which the elite derive more immediate (and assured) personal rewards than from the long haul of wealth creation. As a source of revenue that can be separated from the activity that generates it, rent is up for grabs and therefore triggers political contests for its capture.
The low-rent economy pursues a development trajectory of competitive industrialisation that provides a useful counterfactual with which to understand the staple track trajectory of the high-rent economy that has been typical of sub-Saharan Africa. Low rent aligns the incentives of the elite with those of the majority in growing the economy by providing public goods and efficiency incentives for the reasons noted earlier. This strategy relies on markets rather than patronage channels so that the economy adheres to its comparative advantage, which in low-rent economies initially lies in early competitive labour-intensive manufactured exports, a development trajectory that launches three mutually reinforcing economic, social and political virtuous circles. The economic circle sees export production rapidly absorb surplus rural labour, which raises wages[1] and automatically drives diversification into skill-intensive and capital-intensive sectors. The social circle reflects the early urbanisation caused by early industrialisation that accelerates the demographic cycle to reduce the dependent/worker ratio and thereby raise the share of investment in GDP, which accelerates per capita GDP growth (Bloom & Williamson, 1998). Finally, the associated rapid structural change drives a political circle as social groups proliferate, which constrains policy capture by any one group. It also strengthens three sanctions against anti-social governance as: (1) firms protect their investment by lobbying for property rights and the rule of law (Li et al., 2001); (2) unsubsidised urbanisation strengthens civic voice (Isham et al., 2005); and (3) the government relies for revenue on taxing income, profits and expenditure in the absence of commodity rent, which spurs demand for accountable public finances (Ross, 2001). Overall, incomes rise rapidly and equitably, and democratisation occurs incrementally.
In contrast, high rent incentivises the elite to pursue immediate self-enrichment by channelling rent through patronage networks at the expense of markets, which distorts the economy. Commodity exports drive the economy and over-rapid domestic rent absorption sustains an exchange rate that impedes competitive industrialisation so surplus labour persists and governments deploy rent to create employment that markets would not support in protected industry and an over-expanded bureaucracy. The accelerating demand of the subsidised sector for transfers eventually outstrips the rent (due to structural change or falling commodity prices) and absorbs returns to capital as well as rent so the economy becomes locked into a staple trap of reliance on a weakening primary sector. Declining investment efficiency slows the GDP growth rate, triggering a growth collapse that is protracted because rent recipients resist reform. The growth collapse arrests the demographic transition and exacerbates unemployment, which boosts income inequality and social tension. Finally, the three main sanctions against anti-social governance languish as: (1) businesses find it more profitable to lobby for political favours than for the rule of law; (2) urban dwellers are heavily dependent on government expenditure; and (3) government reliance on rent for revenue dampens pressure for fiscal accountability.
Rent cycling theory recognises that the adverse impacts of high rent are exacerbated (and therefore more intractable) in the presence of: statist policies, high ethnicity, concentrated commodity linkages and young parliamentary democracies. This implies that conditions for effective rent deployment were unpropitious in most African economies at independence since most were statist-leaning, ethnically-mixed resource-rich (and sometimes mineral-dependent) democracies. First, statist ideology amplifies the adverse effects of high rent by boosting the scope for governments to override markets and extract rent (Van der Walle, 1999). Secondly, ethnic tension encourages the use of rent to win political support so that ethnicity is negatively associated with the rate of investment, the rate of economic growth and the quality of government (Montalvo & Reynal-Querol, 2005, p. 294). Third, when domestic commodity linkages concentrate rent on governments and elites, as with mining especially, rent is deployed less effectively than when it is dispersed across many economic agents, as with peasant cash cropping (Baldwin, 1956; Bevan et al., 1999).
Finally, with respect to the political state, Collier & Hoeffler (2009) find that in the presence of high rent democracies underperform autocracies in terms of economic growth whereas the reverse is true in the presence of low rent. They suggest that high rent makes it politically more profitable for democracies to channel public revenue through patronage networks to secure the support of swing votes, rather than into public goods, which confer no electoral edge since they benefit supporters and opponents alike. In addition, Keefer (2007) identifies the age of a democracy as significant: young democracies cannot make credible pre-election promises to voters so they are less trusted than both autocracies and established democracies. Political parties in young democracies therefore under-provide non-targeted goods (like universal education, property rights or access to information), which benefit all while they over-provide targeted goods (like employment and public work projects) that clearly deliver favours to key voting blocs. They are also more corrupt than mature polities. None of the other explanations that Keefer evaluates (political institutions, ethnicity, voter information and civil conflict) cause under-provision of non-targeted public goods. Many, if not all five, of the conditions that are unpropitious for effective rent deployment characterised the sub-Saharan African economies at independence so it is unsurprising that most experienced growth collapses. The next section examines two initially successfully economies, Kenya and the Ivory Coast, and two that avoided sustained growth collapses in order to explain their relative success.
Policies to Improve Resource-Driven Outcomes
Almost all the economies of sub-Saharan African traced a staple trap trajectory and experienced growth collapses through the 1970s and early 1980s. Maladroit resource-rent deployment played a central role, but the resource curse is part of a broader rent curse that can be triggered by regulatory rent in resource-poor economies (as in the Sahel) and by foreign aid (geopolitical rent) in economies seeking to recover from growth collapses. One implication of this is that case studies may provide more reliable insights than multi-country regressions, which have failed to take account of non-resource rent. A second implication is that resource-constrained African economies such as those of the Sahel could experience growth collapses due to misallocation of regulatory rent and aid.
Kenya and the Ivory Coast initially averaged useful rates of per capita GDP growth until the late 1970s when they were destabilised by mismanaged commodity booms. The elite in both countries was initially drawn from wealthy peasant farmers and their post-independence governments espoused open trade policies, cautious macro policy and limited direct state involvement in production. In 1968, however, Kenya enacted legislation that encouraged elite investment in urban activity, which triggered a shift to an inward trade policy that eroded Kenya’s erstwhile impressive export manufacturing competitiveness (Sharpley & Lewis, 1990). A coffee boom conferred a rent windfall of 8% of GDP annually in 1975-79 on Kenya (Cuddington, 1989) and a cocoa boom of 16% of GDP annually in 1976-80 on the Ivory Coast (Ghanem, 1999). The governments of both countries absorbed the windfall revenue too rapidly: the Ivory Coast by expanding state industry to compensate for the absence of industrialists who were nationals and Kenya by stimulating public consumption. In doing so they created entitlements that proved difficult to trim when the booms ended and so they accumulated debt they could not service. Each experienced a growth collapse that destabilised the polity and proved protracted, with evidence from Kenya that an increase in foreign aid intended to ameliorate the growth collapse merely intensified corruption and reduced pressure for reform.
Botswana and Mauritius avoided growth collapses and their experience reinforces the early post-independence lessons from Kenya and the Ivory Coast regarding the benefits of pursuing cautious macro management and open trade policy to facilitate private-sector wealth creation. Botswana inherited from the UK at independence in 1966 a market economy and parliamentary democracy that was strengthened by a tradition of consensual politics. The elite initially relied on cattle exports, which disposed them towards open trade policies. However, somewhat uniquely, the associated elite bias towards caution was reinforced when the economy grew to rely on diamonds and little else. The diamond rent depended upon the survival of a controversial monopoly and the government reacted to the risk of an abrupt collapse in diamond prices as if it managed a low-rent economy. The government targeted long-term wealth creation, counting on prudent rent management to sustain its authority. It established funds in 1972 to stabilise the mineral rent (Maipose & Matsheka, 2008, p. 523), which increased four-fold from 1973-93 and offset falling aid to yield a net total rent flow of one-fifth to one-quarter of GDP. The government typically allocated two-fifths of the revenue to offshore investments and financial reserves reached 125% of GDP by 1998 (IMF, 1999).
Although mining revenue lifted public expenditure to two-thirds of GDP (Harvey & Jefferis, 1996) the Botswana government resisted fashionable policies to create unsustainable entitlements by protecting ‘infant’ activity, expanding state enterprises or subsidising prices. Instead it consciously converted diamond rent into human capital and economic infrastructure, eliminating the backlog from colonial neglect. The government also deployed rent to nurture the private sector by sub-contracting goods and service purchases and encouraging foreign investment. It stimulated the rural economy, which until the late-1990s employed the majority of Tswanans, by supporting not only cattle-rearing but also until the 1990s small-holder agriculture through a cautious policy of food self-sufficiency.
Diamond expansion drove rapid economic growth in 1970-90 with an efficient ICOR of 2.3. Sustained PCGDP growth raised all boats and, although income distribution was skewed, the gini coefficient remained stable at around 0.51 (Sarraf & Jiwanji, 2003, p. 15). High PCGDP growth helped the ruling party to retain a parliamentary majority whereas its largest opponent, a left-leaning urban party, typically held one-sixth of the seats. The ruling party offset the slow contraction of its bedrock rural constituency by accommodating social pressure rather than buying it off with opaque rent cycling.
But Botswana enjoyed unique advantages not only for its ethnic homogeneity, consensual tradition and suspicion of state enterprises, but also from quirks of its rent. Diamond rent not only engendered caution because it was precarious, but it also proved unusually stable, so Botswana’s institutional resilience remains untested. The absence of rent surges has avoided the sudden ignition of political pressure for government spending, while also evading the need for abrupt and draconian rent rationing as was required in Kenya and the Ivory Coast in the early-1980s. Moreover, Botswana’s economy has struggled when mine expansion has slowed because it remains mineral-driven.
Although Mauritius was already land-scarce by the 1960s, rent was still 7% or more of GDP annually and was derived from foreign aid and privileged market access for sugar and textiles. Mauritius’s rent deployment provides a policy blueprint for injecting the required political component into African economic reform. In the 1960s Mauritius was a mono-crop economy that depended on sugar plantations for 98% of exports, 35% of GNP (compared with 10% for other agriculture) and 35% of employment when its land frontier closed. Per capita cropland was a mere 0.13 hectares, similar to land-scarce north-east Asia and double that of Bangladesh at the time. A 1961 Commission of Enquiry recommended economic diversification through import substitution industry but the ‘infant’ industry failed to mature and merely expanded rent seeking: even in 1980 sugar rent subsidised rates of effective protection averaging 185% (Findlay & Wellisz, 1993). The protected industry also tended to be capital-intensive, which led the government to expand public sector employment to one-sixth of the workforce by 1967.
Nevertheless, amid the threat of ethnic conflict at independence in 1968 the elite brokered a coalition government that prioritised wealth creation and cycled a high fraction of the rent via markets. It did so by deploying rent as part of a dual-track strategy that expanded a dynamic market sector but weakened opposition to economic reform by shrinking the rent-seeking sector only slowly. In contrast to Kenya, foreign aid was steadily phased out after an initially modest flow. But Mauritius still drew rent from favoured access to markets: the Commonwealth Sugar Agreement conferred rent averaging 4.5% of GDP annually during 1977-2000 and the Mutli-Fibre Agreement yielded 0.5% of GDP in 1984, rising to 2.9% of GDP in 1996 as clothing exports grew (Subramanian & Rodrik, 2003, p. 223). But unlike its counterparts in most African countries, including Kenya and the Ivory Coast, Mauritius’s government responded to an unexpected windfall (in sugar revenue) equivalent to an extra 7.4% of GDP annually during 1972-75 (Greenaway & Lamusse, 1999, p. 214) by limiting windfall taxation to 5%-12% so that most of the 1970s’ sugar windfall flowed to private producers.
The planters responded prudently: gross saving jumped from 15.2% by 1970-72 to 34.2% of GDP during 1974-75 (Findlay & Wellisz, 1993) and although investment lagged, it increased by one-half to average 23% of GDP on a rising trend. The planters invested some windfall revenue in an export processing zone (EPZ) that the government established in 1970 as Track 1 of its dual-track strategy to promote a dynamic market sector of labour-intensive export manufacturing and tourism. Track 2 was the rent-supported state-dominated sector, where social spending increased during the sugar boom from 6% of GDP to 10% in order to undercut the political opposition. Social tension eased through the 1970s as the dual-track rent deployment drove per capita GDP at 6% annually. However, spending outstripped revenue in the late 1970s, forcing the government to turn to the IMF, which advised curbing social expenditure and dismantling all trade controls.
Thereafter, as part of its dual-track strategy the government still channelled one-third of its geopolitical rent (some 2.5%-3% of GDP annually) into a social safety-net, while private producers cycled much of the rest (5%-6.5% of GDP) into wealth-generating activity. As population growth fell to 1% EPZ expansion drove per capita GDP at 5.7% annually through the 1980s. Consistent with the competitive industrialisation model, manufactured exports increased from one-quarter of the total in 1980 to two-thirds in 1990, ending sugar’s dominance. EPZ employment trebled and unemployment fell from 21% to 4%, creating labour shortages by 1990. By the mid-1990s, Mauritius textile wages had reached four times those of China and Vietnam, prompting diversification into IT in the EPZ to sustain productivity growth (Chernoff & Warner, 2002). Services became the main driver of the economy in the 1990s, mainly by tourism and financial services.
Mauritius’s dual-track rent deployment successfully diversified its economy from an unsustainable resource-dependence into competitive industrialisation, despite initial conditions that also included acute ethnic tension and a young parliamentary democracy. Its achievements confirm the benefits of cautious macro management and cycling rent through markets and the private sector. The threat of a growth collapse initially motivated the Mauritius elite and its Tswanan counterpart to promote growth over rent seeking. Such a threat holds lessons for the elite of African economies struggling to re-establish growth: policies to sustain growth promise greater security for all, elite and majority alike.
Conclusion: The resource curse is part of a broader rent curse that can be triggered by foreign aid (geopolitical rent) and regulatory rent as well as by natural resource rent. Since the policies to avoid the problems of commodity-driven growth such as the Dutch disease effects are well established, the challenge created by rent is to neutralise the incentive it gives to the elite to circumvent markets and deploy rent through patronage channels for immediate and personal gain at the expense of sustained long-term economic growth. The literature shows that this incentive is especially strong in the presence of point source resources (notably mining), ethnic tension, statist policies and new democracies. This implies that initial conditions were unpropitious for African countries because most embarked on independence as ethnically-mixed resource-rich democracies with a predilection for state control.
Almost all African economies did experience protracted growth collapses, but a handful delayed the curse and two avoided it by espousing cautious macro policies and cycling much rent through competitive private firms. After promising starts, Kenya and the Ivory Coast slipped into unsustainable public expenditure in deploying their rent windfalls in the late 1970s but Botswana and Mauritius escaped the resource curse. Mauritius’s experience is especially instructive, whereas Botswana singularly benefited from the diamond rent’s manifest precariousness, which incentivised the elite to promote growth and in the event delivered a uniquely stable rent stream. Mauritius deployed its more modest rent through a dual-track strategy in which Track 1 grew a dynamic market economy by facilitating diversification of the private sector into first competitive manufacturing and then services. Track 2 maintained social programmes in the rent-dependent sector to undercut political opposition until the scale of the dynamic sector and its pro-growth political coalition reduced rent recipients to a minority.
Scope to pursue the dual-track strategy increased in sub-Saharan Africa when the growth collapses heightened dependence on the IFIs through the 1980s and 1990s. However, IFI influence has since weakened because incentives for elite rent seeking expanded due to the 2003-08 commodity boom and also increased Asian investment in African resources. Nevertheless, the consequences of previous growth collapses should prompt elites in Kenya, the Ivory Coast and elsewhere to reflect that they can prosper more securely by adapting the dual-track strategy of Mauritius (as well as Botswana, Chile, China, Indonesia and Malaysia) than by prioritising rent-seeking activity.
Richard M. Auty
Professor Emeritus of Economic Geography at Lancaster University
References
Auty, R.M. (2007), ‘Patterns of Rent-extraction and Deployment in Developing countries: Implications for Governance, Economic Policy and Performance’, in G. Mavrotas & A. Shorrocks (Eds.), Advancing Development: Core Themes in Global Economics, Palgrave, London, p. 555-77.
Auty, R.M. (2009), ‘Elites, Rent Cycling and Development: Adjustment to Land Scarcity in Mauritius, Kenya and Cote d’Ivoire’, paper presented at the UNU/WIDER Conference on Elites in Economic Development, Helsinki, 12-13/VI/2009.
Auty, R.M., & A.H. Gelb (1986), ‘Oil Windfalls in a Small Parliamentary Democracy: Their Impact on Trinidad and Tobago’, World Development, nr 14, p. 1161-1175.
Baldwin, R.E. (1956), ‘Patterns of Development in Newly Settled Regions’, Manchester School of Social and Economic Studies, nr 24, p. 161-179.
Bevan, D., P. Collier & J.W. Gunning (1987), ‘Consequences of a Commodity Boom in a Controlled Economy: Accumulation and Redistribution in Kenya’, World Bank Economic Review, nr 1, p. 489-513.
Bloom, D.E., & J.G. Williamson (1998), ‘Demographic Transitions and Economic Miracles in Emerging Asia’, The World Bank Economic Review, nr 12, p. 419-55.
Collier, P., & A. Hoeffler (2009), ‘Testing the Neo-con Agenda: Democracy in Resource-rich Societies’, European Economic Review, nr 53, p. 293-308.
Cuddington, J. (1989), ‘Commodity Price Booms in Developing Countries’, World Bank Research Observer, nr 4, p. 143-65.
Findlay, R., & S. Wellisz (1993), Five Small Open Economies: The Political Economy of Poverty, Equity and Growth, Oxford University Press, New York.
Gelb, A.H., & Associates (1988), Oil Windfalls: Blessing or Curse?, Oxford University Press, New York.
Ghanem, H. (1999), ‘The Ivorian Cocoa and Coffee Boom of 1976-70: The End of a Miracle?’, in P. Collier, J.W. Gunning & Associates, Trade Shocks in Developing Countries. Volume I Africa, Oxford University Press, Oxford, p. 142-162.
Greenaway, D., & R. Lamusse (1999), ‘Private and Public Sector Responses to the 1972-75 Sugar Boom in Mauritius’, in P. Collier, J.W. Gunning & Associates, Trade Shocks in Developing Countries. Volume I Africa, Oxford University Press, Oxford, p. 207-225.
Harvey, C., & K. Jefferis (1995), ‘Botswana’s Exchange Controls: Abolition or Liberalization?’, IDS Discussion Paper, nr 348, IDS, Brighton.
IMF (1999), IMF Staff Country Report 99/132: Botswana: Selected Issues, International Monetary Fund, Washington DC.
Isham, J., L. Pritchett, M. Woolcock & G. Busby (2005), ‘The Varieties of Resource Experience: How Natural Resource Export Structures Affect the Political Economy of Economic Growth’, World Bank Economic Review, nr 19 (1), p. 141-164.
Li, S., S. Li & W. Zhang (2000), ‘The Road to Capitalism: Competition and Institutional Change in China’, Journal of Comparative Economics, nr 28, p. 269-292.
Keefer, P. (2007). ‘Clientelism, Credibility and Policy Choices of Young Democracies’, American Journal of Political Science, nr 51(4), p. 804-821.
Maipose, G.S., & T.S. Matsheka (2008), ‘The Indigenous Developmental State and Growth in Botswana’, in B.J. Ndulu et al. (Eds.), The Political Economy of Economic Growth in Africa 1960-2000: Volume 2, Cambridge University Press, Cambridge, p. 511-46.
Montalvo, J.G., & M. Reynal-Querol (2005), ‘Ethnic Diversity and Economic Development, Journal of Development Economics, nr 76, p. 293-323.
Ross, M. (2001), ‘Does Oil Hinder Democracy?’, World Politics, nr 53/3, p. 325-61.
Sachs, J.D., A. Warner (1995), ‘Natural Resources and Economic Growth’, mimeo, HIID, Cambridge MA.
Sarraf, M., & M. Jiwanji (2003), ‘Beating the Resource Curse: The Case of Botswana’, Environment Department Papers, nr1, World Bank, Washington DC.
Sharpley, J., & S. Lewis (1990), ‘The Manufacturing Sector in the Mid-1980s’, in R.C. Riddell (Ed.), Manufacturing Africa, James Currey, London, p. 206-41.
Singh, R.J., M. Haacher & K. Lee (2009), ‘Determinants and Macroeconomic Impacts of Remittances in Sub-Saharan Africa’, IMF Working Paper, nr 09/216, IMF, Washington DC.
Subramanian, A., & D. Roy (2003), ‘Who Can Explain the Mauritian Miracle? Meade, Romer, Sachs or Rodrik?’, in D. Rodrik (Ed.), In Search of Prosperity: Analytic Narratives of Economic Growth, Princeton: Princeton University Press, p. 205-243.
Van de Walle, N. (2001), African Economies and the Politics of Permanent Crisis, 1979-1999, Cambridge University Press, Cambridge.
[1] It also constrains income inequality by putting a floor under the wages of the poor even as rapid skill diffusion caps the skill premium.