The Journal of Development Studies

The paradigm of 'inclusive neoliberalism' that currently characterises international development places a particular emphasis on community-based responses to the often structural problems of poverty and exclusion. Such approaches have become increasingly controversial: celebrated by optimists as the most empowering way forward for marginal citizens on the one hand, and derided as an abrogation of responsibility by development trustees by sceptics on the other. Uganda provides a particularly interesting context to explore these debates, not least because it has become a standard bearer for inclusive neoliberalism at the same time that regional inequalities within it have become increasingly apparent. Our investigation of the flagship response to deep impoverishment in its northern region, the World Bank-funded Northern Uganda Social Action Fund, offers greater support to the sceptics, not least because of the ways in which the more pernicious tendencies within inclusive neoliberalism have converged with the contemporary politics of development in Uganda.

Inequality of agricultural labour productivity across the developing world has increased substantially over the past 40 years. This article asks: to what extent did the diffusion of Green Revolution seed varieties contribute to increasing agricultural labour productivity disparity across the developing countries? We find that 22 per cent of cross-country variation in agricultural labour productivity can be attributed to the diffusion of high-yielding seed varieties across countries, and that the impact of such diffusion differed significantly across regions. We discuss the implications of these findings for policy directed at increasing agricultural labour productivity in the developing world.

The mechanisms by which the poor benefit from economic growth remain a topic of debate in development literature. We address this issue in the context of rural Bangladesh, using a pooled dataset of three household panels between 1991-2001. Expansion of irrigation, paved roads, electricity, and access to formal and informal credit have (through different veins) led to higher rural farm and non-farm incomes, accounting for exogenous local agroclimatic endowments that explain a large part of the variation in the growth of infrastructure and credit programmes. However, this has not translated into substantial reductions in poverty for the poorest households.

This paper investigates how the South African state has understood the relationship between HIV and poverty and how individuals and community-based organisations have responded to these state interventions. It considers the ways in which liberal forms of government frame people living with AIDS as a particular category of ‘deserving’ and ‘entrepreneurial’ citizens, and then re-frames them through a package of health and welfare interventions. Based on ethnographic research with the members of Khululeka, a support group for HIV-positive men, the study pays particular attention to how masculinity has shaped the ways these men have experienced and transformed these state interventions.

The accommodation of livestock husbandry with crop agriculture is crucial for the future of the West African Sahel. Present trends are leading to greater restrictions on livestock husbandry and a growing convergence of livelihood practices among groups whose identities are tied to herding and farming. Using the cases of four rural communities in Niger, this study adopts an 'access to resources' framework to analyse the causal connections among: rural peoples' livelihood strategies, everyday social relations of production, perceptions of social groups' identities, and the potential for farmer-herder conflict. While the convergence of livelihoods arguably increases the frequency of conflict triggers, it has also, through the expansion of shared common interests and cross-group, production-related relationships, improved the ability of communities to effectively manage these incipient conflicts.

This paper critically reviews evidence from low and middle income countries that pensions are associated with better health outcomes for older people. It draws on new, nationally representative survey data from South Africa to provide a systematic analysis of pension effects on health and quality of life. It reports significant associations with the frequency of health service utilisation, as well as with awareness and treatment of hypertension. There is, however, no association with actual control of hypertension, self-reported health or quality of life. The paper calls for a more balanced and integrated approach to social protection for older people.

The increasing prevalence of private standards governing food safety, food quality and environmental and social impacts of agri-food systems has raised concerns about the effects on developing countries, as well as the governance of agri-food value chains more broadly. It is argued that current debates have been 'clouded' by a failure to recognise the diversity of private standards in terms of their institutional form, who develops and adopts these standards and why. In particular, there is a need to appreciate the close inter-relationships between public regulations and private standards and the continuing ways in which private standards evolve.

This article explores the implications of women's work in agriculture in Telangana, a region in the state of Andhra Pradesh, India. I suggest that higher capital costs for cultivators' post-liberalisation increased the pressure to contain wage costs in a region where women form the majority of the agricultural wage labour force. Under such conditions, when women perform both own-cultivation as well as agricultural wage work in the fields of others, they face pressure to restrict bargaining for higher wages, contributing to a widening gender wage gap. To the extent that wages shape intra-household bargaining power, the empowering effect of workforce participation for such women would thus be blunted. From available NSS data I provide some preliminary evidence in support of this argument.

This paper provides an historical survey of the evolution of rice technology in China, from the traditional farming system to genetically modified rice today. Using sociotechnological analytical framework, it analyses rice technology as a socio-technical ensemble - a complex interaction of material and social elements, and discusses the specificity of technology development and its socio-technical outcomes. It points to two imperatives in rice variety development: wholesale transporting agricultural technology and social mechanism to developing countries are likely lead to negative consequences; indigenous innovation including deploying GM technology for seed varietal development and capturing/cultivating local knowledge will provide better solutions.

This article explores the nexus between indigenous mobilisation, citizenship, and poverty in Argentina. A subnational comparison of land struggles among the Diaguita Calchaqu in Tucumn and the Mbya Guaran in Misiones shows that changing global and national opportunity structures, most prominently a new multicultural citizenship regime, set the stage for indigenous mobilisation. In turn, local transformations of capitalist development motivate indigenous mobilising efforts, whereas leadership patterns and state-movement relations shape the capacity to mobilise. Diaguita and Mbya mobilisation reveals that indigenous movements play a central role in the activation of formal citizenship rights and the contestation of dominant notions of poverty. At the same time, the current design of multicultural citizenship and the adverse socioeconomic incorporation of indigenous communities also counteract indigenous mobilising efforts in Argentina.

Bt cotton is accused of being responsible for an increase of farmer suicides in India. In this article, we provide a comprehensive review of evidence on Bt cotton and farmer suicides. Available data show no evidence of a 'resurgence' of farmer suicides. Moreover, Bt cotton technology has been very effective overall in India. Nevertheless, in specific districts and years, Bt cotton may have indirectly contributed to farmer indebtedness, leading to suicides, but its failure was mainly the result of the context or environment in which it was planted.

This paper analyses vulnerability to poverty of rural small-scale fishing communities using cross-section data from 295 households in Cameroon and 267 in Nigeria. We propose a vulnerability measure that incorporates the idea of asset poverty into the concept of expected poverty, which allows decomposing expected poverty into expected structural-chronic, structural-transient, and stochastic-transient poverty. The findings show that most households in our study areas are expected to be structurally-chronic and structurally-transient poor. This underlines the importance of asset formation for long-term poverty reduction strategies. Further refinements are possible with longitudinal data and information about future states of nature.

Evidence indicates that a much-feted conditional cash transfer programme designed to widen access to basic education in Bangladesh has failed in its aims. The programme is analysed here as an instance of the effort to govern chronic poverty. For the state, education fits within a national project of poverty reduction and creating governable citizens. For the poor, education signals social inclusion and access to the state. Yet class and social distinctions through which state actors 'see' poor children result in beneficiary selection practices that routinely exclude the poorest from school, with longer-term adverse effects for their social inclusion and citizenship.

This paper explores the implications of using two methodological approaches to study poverty dynamics in rural Bangladesh. Using data from a unique longitudinal study, we show how different methods lead to very different assessments of socio-economic mobility. We suggest five ways of reconciling these differences: considering assets in addition to expenditures, proximity to the poverty line, other aspects of well-being, household division, and qualitative recall errors. Considering assets and proximity to the poverty line along with expenditures resolves three-fifths of the qualitative and quantitative differences. Use of such integrated mixed-methods can therefore improve the reliability of poverty dynamics research.

Using controlled experiments to compare the risk attitude and willingness to compete of husbands and wives in 500 couples in rural Vietnam, we find that women are more risk averse than men and that, compared to men, women are less likely to choose to compete, irrespective of how likely they are to succeed. Relevant to development programmes concerned with lifting women out of poverty, our findings suggest that women may be more reluctant to adopt new technologies, take out loans, or engage in economic activities that offer higher expected returns, in order to avoid setups that require them to be more competitive or that have less predictable outcomes.

Drawing on a participatory study of integrated organic waste management, this article explores the local political barriers and preconditions for its implementation in Diadema, Brazil. Solid waste management in Brazil is embedded in and mediated by a political framework that is characterised by uneven power geometries. This article explores how the local political context affects the potential for integrated organic waste management in Diadema, paying particular attention to relations between stakeholders. The discussion addresses the contested nature of deliberative decision-making spaces and the need for pro-active socio-environmental policies. The findings underline the importance of a praxis of everyday public participation that goes beyond rhetoric.

Using survival models, we test whether short-term changes in the labour market affect poverty duration. Data are from the Brazilian Monthly Employment Survey. Such a monthly dataset permits more accurate estimations of events than using annual data, but its panel follows households for a short period. Then methods that control for both right- and left-censoring should be used. The results are as follows: households with zero income are not those with the lowest chances of exiting; changes in aggregate unemployment do not affect poverty duration; and increasing wages in the informal sector has a negative effect on poverty duration.

There is a renewed interest in whether land reforms can contribute to market development and poverty reduction in Africa. This paper assesses effects on the allocative efficiency of the land rental market of the low-cost approach to land registration and certification of restricted property rights that was implemented in Ethiopia in the late 1990s. Four rounds of a balanced household panel from 16 villages in northern Ethiopia are analysed, showing that land certification initially enhanced land rental market participation of (potential) tenant and landlord households, especially those that are headed by females.

Few papers in the literature provide quantitative analysis of the difficult circumstances faced by children of short-term labour migrants. This paper uses new survey data from rural northwest India to study both children who migrate and those left behind. It finds that, unlike in other contexts, children who migrate rarely work when they accompany adult migrants. Additionally, this paper reports a robust, previously unquantified negative relationship between children's migration and educational outcomes and investments. It calls for further research about externalities of migration for children and suggests that expansion of a large public employment program might help these children.

Why is it that couples who have a son or whose last child is a son earn higher conditional income? To solve this curious case we tell a detective story: evidence of a phenomenon to be explained, a parade of suspects, a process of elimination from the enquiry, and then the denouement. Given the draconian family planning policy and a common perception that there is strong son preference in rural China, we postulate two main hypotheses: income-based sex selection making it more likely that richer households have sons, and an incentive for households with sons to raise their income. Tests of each hypothesis are conducted. Taken as a whole, the tests cannot reject either hypothesis but they tend to favour the incentive hypothesis; and there is evidence in support of the channels through which the incentive effect might operate. To our knowledge, this is the first study to test these hypotheses against each other in rural China and more generally in developing countries.

Development trustees have increasingly sought to challenge chronic poverty by promoting citizenship amongst poor people, a move that frames citizenship formation as central to overcoming the exclusions and inequalities associated with uneven development. For sceptics, this move within inclusive neoliberalism is inevitably depoliticising and disempowering, and our cases do suggest that citizenship-based strategies rarely alter the underlying basis of poverty. However, our evidence also offers some support to those optimists who suggest that progressive moves towards poverty reduction and citizenship formation have become more rather than less likely at the current juncture. The promotion of citizenship emerges here as a significant but incomplete effort to challenge poverty that persists over time.

The paper "High Noon for Microfinance Impact Evaluations" by Duvendack and Palmer-Jones replicates the papers of Chemin (2008) and Pitt and Khandker (1998) that estimate the impact of microfinance in Bangladesh. My paper replicates the Duvendack and Palmer-Jones replication and finds so many serious errors in their code and misrepresentations of the methods described in their paper that I conclude that their results are spurious and provide no evidence about the validity of either the papers of Chemin or Pitt and Khandker or on the effectiveness of microfinance.

The significance of social movements for pro-poor political and social change is widely acknowledged. Poverty reduction has assumed increasing significance within development debates, discourses and programmes - how do social movement leaders and activists respond? This paper explores this question through the mapping of social movement organisations in Peru and South Africa. We conclude that for movement activists 'poverty' is rarely a central concern. Instead, they represent their actions as challenging injustice, inequality and/or development models with which they disagree, and reject the simplifying and sectoral orientation of poverty reduction interventions. In today's engagement with the poverty-reducing state, their challenge is to secure resources and influence without becoming themselves subject to, or even the subjects of, the practices of government.

The article argues for what can be called a 'relational' approach to poverty: one that first views persistent poverty as the consequence of historically developed economic and political relations, and second, that emphasises poverty and inequality as an effect of social categorisation and identity, drawing in particular on the experience of adivasis ("tribals") and dalits ("untouchables") subordinated in Indian society. The approach follows Charles Tilly's Durable Inequality in combining Marxian ideas of exploitation and dispossession with Weberian notions of social closure. The article then draws on the work of Steven Lukes, Pierre Bourdieu and Arjun Appadurai to argue for the need to incorporate a multidimensional conception of power; including not only power as the direct assertion of will but also 'agenda-setting power' that sets the terms in which poverty becomes (or fails to become) politicised, and closely related to power as political representation. This sets the basis for discussion of the politics of poverty and exclusion.

This paper evaluates effects of community-level women's property and inheritance rights on women's economic outcomes using a 13 year longitudinal panel from rural Tanzania. In the preferred model specification, inverse probability weighting is applied to a woman-level fixed effects model to control for individual-level time invariant heterogeneity and attrition. Results indicate that changes in women's property and inheritance rights are significantly associated with women's employment outside the home, self-employment and earnings. Results are not limited to sub-groups of marginalised women. Findings indicate lack of gender equity in sub-Saharan Africa may inhibit economic development for women and society as a whole.

This study utilises eight alternative measures of institutions and the instrumental variable method to examine the impacts of institutions on poverty. The estimates show that an economy with a robust system to control corruption, an effective government, and a stable political system will create the conditions to promote economic growth, minimise income distribution conflicts, and reduce poverty. Corruption, ineffective governments, and political instability will not only hurt income levels through market inefficiencies, but also escalate poverty incidence via increased income inequality. The results also imply that the quality of the regulatory system, rule of law, voice and accountability, and expropriation risk are inversely related to poverty but their effect on poverty is via average income rather than income distribution.

Strategic interaction between public and private actors is increasingly recognised as an important determinant of agricultural market performance in Africa and elsewhere. Trust and consultation tends to positively affect private activity while uncertainty of government behaviour impedes it. This paper reports on a laboratory experiment based on a stylised model of the Zambian maize market. The experiment facilitates a comparison between discretionary interventionism and a rules-based policy in which the government pre-commits itself to a future course of action. A simple precommitment rule can, in theory, overcome the prevailing strategic dilemma by encouraging private sector participation. Although this result is also borne out in the economic experiment, the improvement in private sector activity is surprisingly small and not statistically significant due to irrationally cautious choices by experimental governments. Encouragingly, a rules-based policy promotes a much more stable market outcome thereby substantially reducing the risk of severe food shortages. These results underscore the importance of predictable and transparent rules for the state's involvement in agricultural markets.

The training of indigenous women in hospital nursing is directly modelled on western services. Hence training does not address social problems attendant on implicit cultural conflict between indigenous and western healing systems and conceptualisations of illness. Because these problems are not faced in the curriculum, nurses are imperfectly socialised into the modern health care setting. They, and the public, continue to accept the efficacy of both health care systems, assuming complementarity when in fact in many respects they conflict. The involvement of young women in healing, a radical innovation, became institutionalized. Because of the high visibility of the nurse and a radical change in the role of young women as healers, and because of the vulnerability of the patients, conflict in treatment systems focused on nurses who became scapegoats in hospitals, which came to be called 'houses of death'.

This paper aims to explore the changing determinants of child progress through school over the last two decades using unique long-term household-panel data from four villages in the Philippines. In a regime of low income in the late 1980s, income from farming is the most important source of funds to finance child schooling. As households shift away from farm to non-farm activities and their children pursue higher education, non-farm income and revenues from pawning of land have emerged as main sources of schooling funds in the early 2000s. In this process, farm income has lost its prime importance as a determinant of schooling investments among rural households.

A two-sector model of terms of trade (TOT) determination is developed and tested using time-series data for Turkey. Empirical results support the structuralist 'flex-price agriculture fix-price industry' models. TOT is found to be sensitive to changes in nominal demand and the exchange rate. Rising nominal demand turns the TOT in favour of the agricultural sector provided that there are no supply constraints in the industrial sector. If industrial supply is constrained by import bottlenecks, then aggregate demand expansion turns the TOT in favour of the industrial sector. Devaluation turns the TOT against the agricultural sector primarily via the cost-push factors in the industrial sector.

This paper examines the phenomenon of real-income stagnation (in which real-income growth is uninterruptedly negligible or negative for a sizable sequence of years). We analyse data for four decades from a large cross-section of countries. Real income stagnation is a conceptually distinct phenomenon from low average growth and other features of the growth sequence that have been previously considered. We find that real income stagnation has affected a significant number of countries (103 out of 168), and resulted in substantial income loss. Countries that suffered spells of real income stagnation were more likely to be poor, in Latin America or sub-Saharan Africa, conflict ridden and dependent on primary commodity exports. Stagnation is also very likely to persist over time. Countries that were afflicted with stagnation in the 1960s had a likelihood of 75 per cent of also being afflicted with stagnation in the 1990s.

Given shortcomings in basic data collection and insufficient resources in preparing official statistics African growth data are unlikely to be very reliable. Estimates of an annual growth rate of 3 per cent may be consistent with a reality between 0 and 6 per cent growth. Although data from international databases are widely used in an expanding literature on African growth there has been no research into how serious these data inaccuracies are. This paper addresses the reliability of the available growth evidence for a selection of countries and offers concrete measures of inaccuracies. It examines the reasons for discrepancies and shows that they can be quite large.

This study analyses how changes in factor abundance and openness have affected relative factor prices in Kenya since 1965, using cointegration analysis and error correction models of relative factor prices. We find that factor proportions determined relative factor prices in the long run, while openness, measured by three different proxies, possibly had a short run effect on relative factor returns. The only deviation from this pattern occurred during the latter half of the 1990s when there was rapid wage growth, mainly due to labour market deregulation.

During the 1970s and early 1980s Paraguay experienced relatively high rates of economic growth as well as a boom in primary goods production destined for export. The question which this research addresses concerns the relationship between these events and the applicability of the so-called export-led growth (ELG) hypothesis. The hypothesis is investigated via the use of modern time series methods including Granger causality tests, error correction modeling, and vector autoregression. The basic conclusion reached is that the ELG does not have much relevance to the Paraguayan case.

This article presents estimates of capital flight from 25 low-income sub-Saharan African countries in the period 1970 to 1996. Capital flight totaled more than $193 billion (in 1996 dollars); with imputed interest earnings, the accumulated stock of flight capital amounts to $285 billion. The combined external debt of these countries stood at $178 billion in 1996. Taking capital flight as a measure of private external assets, and calculating net external assets as private external assets minus public external debts, sub-Saharan Africa thus appears to be a net creditor vis-a-vis the rest of the world.

Whether the plight of OECD unskilled labour is due to trade or technology is examined with a general equilibrium Heckscher-Ohlin calibrated model of North and South. Technology transfer from North to South via direct foreign investment in unskilled-labour-intensive manufacturing industries is identified as the trade shock. Simulations are carried out for this and other relevant shocks, and compared with the facts of the 1970-90 period. Weights on the shocks are selected by least squares. It is found that the roles of trade and technology in the plight of Northern unskilled labour are roughly equal.

We examine India's regional disparities in economic performance between 1970-97. Our preliminary analysis shows that, in absolute terms, initially poorer states grew at slower rates than initially wealthier ones and that there is also evidence of increasing dispersion of income levels across the states. Our econometric analysis investigates the possibility of club convergence and conditional convergence. Although we do not find evidence of the former, we can suggest some of the factors associated in the latter. Our research also indicates that the onset of economic policy reform in 1991 significantly intensified growth differentials between the states.

West Bengal, a major state of eastern India, is conspicuous not only for being ruled by an elected Leftist coalition since 1977 (often described as sound 'political stability'), but also for its widely acknowledged successes in fertility transition, execution of redistributive land reform and political decentralisation programmes. Ironically, however, the state, in almost all comparative assessments of social, human and infrastructural developments occupies a lagged position vis-a-vis many other states, especially in the south and even against all-India records. This paper seeks to examine this paradox by comprehensively evaluating West Bengal's relative performance in demographic and socio-economic transformations. A well-disciplined grassroots political mobilisation network, and the machinery of the Left Front parties, have been highly instrumental for comparatively fast declines of fertility and population growth and for lasting political stability in an otherwise 'laggard' development regime. However, a government geared to ensuring mass electoral support overwhelmingly via a grassroots mobilisation network but, with a relative neglect of social movements, economic infrastructure and human development, is likely to suffer adverse consequences in the longer term.

If parents expect higher market returns to schooling or additional transfers from their children, they invest more in their children's schooling. Results from models of schooling demand using data from the Malaysian Family Life Surveys of 1976 and 1989 suggest that market returns of mothers but not fathers positively affected schooling. The propensity for parents to spend time with their parents had a small positive effect on education of daughters, but other transfers had a weak negative effect. The results suggest that if one generation perceives a low return to schooling, then the next generation - especially daughters - pays the price of lower schooling.

Since 1980, income inequality has risen faster in Taiwan than in the United States. Inequality rose despite a rapid increase in the share of educated workers in the labour market that might have been expected to depress returns to education. Returns to a college education rose in Taiwan for all but the least experienced college graduates who were the most substitutable by the large new cohorts of college graduates. This pattern of changes in relative employment and relative wages is consistent with persistent shifts in relative demand toward skilled labour. The shifts are not sector-specific as might have been the case if shifts in trade flows were responsible for the shifts in relative wages. Growth of relative employment of more-educated workers occurred in all sectors of the economy, consistent with the hypothesis of skill-biased technical change. These results are similar to findings reported for OECD countries, suggesting that Taiwan has been exposed to the same types of skill-biased shifts in relative labour demand as in Europe and North America.

After declining from the mid-1970s to the mid-1980s, inequality in monthly earnings in Costa Rica stabilised from 1987 to 1992 and then increased from 1992 to 1999. In this article, we use recently developed techniques to measure the extent to which these changes in earnings inequality were the result of changes associated with the distributions of personal and workplace characteristics of workers or the earnings differences associated with those characteristics. We present evidence that the most important cause of the fall in inequality prior to 1987 was a decline in returns to education. Inequality stopped falling in Costa Rica in the 1990s in part because returns to education stopped falling. The most important cause of rising inequality in monthly earnings in the 1990s was an increase in the proportion of workers working a non-standard work week (part-time or over-time).

This paper advances a particular measure of deprivation - called the Capability Failure Ratio (CFR) - on the dimensions of longevity, knowledge and income, and suggests how an (inequality-) 'adjusted' version of the CFR may be derived. These measurement concerns are explored in the context of relevant State-wise data for the Indian Union. The paper thus makes an attempt to enrich the analysis of human predicament by sensitising 'measures of central tendency' to distributional concerns.

This article investigates whether the Chilean banking reforms of the 1980s have contributed to reducing market imperfections in Chilean financial markets in the late 1980 and early 1990s. To analyse this issue, patterns of investment and its finance for different types of firms are studied, based on balance sheet information of a panel of 70 firms. The most important conclusion is that those specific reforms, aimed at reducing intra-conglomerate lending, seem to have been successful, since access of non-conglomerate firms has increased, indicating a reduction of existing market imperfections for such firms.

Economic reform programmes assume that major goods are tradable, such that depreciation of the real exchange rate raises the value of output compared to factor costs in domestic currency. In Tanzania, major food staples that account for most real income are non-tradables in at least one-quarter of the country. This conclusion is demonstrated and the implications are assessed for the constraints imposed on macroeconomic-led adjustment strategies.

Cote d'Ivoire's economy declined drastically in the second half of the 1980s. The incidence of poverty climbed from 30 percent in 1985 to 35 percent in 1987, and jumped to 46 percent in 1988. But how widespread was the collapse in living standards? Did a lucky few escape the decline? Using panels of data from the Cote d'Ivoire Living Standards Survey (for 1985-86, 1986-87, and 1987-88) allowed the authors to track the level of living for the same households over successive years. These panels had not yet been used to examine the dynamics of poverty in the second half of the 1980s. They find that two-period poverty was generally less than poverty measured from single-period snapshots. Surprisingly, a significant number of the poorest of the poor improve their status over the two years of the panel, even though there was a downturn in the average fortunes of the poor. The authors find that thelucky few are not so few. They were wide-spread regionally - though in some socioeconomic groupings, the poor had a greater chance to escape poverty amidst the general decline in living standards. Finer investigation of the characteristics of these groupings is hampered somewhat by the small sample sizes of the panels.

Over the course of the 1988-97 period, the Brazilian steel sector was subject to an unprecedented programme of trade liberalisation and privatisation while becoming far more open towards foreign investment and technology inflows. Using various indicators of competitive performance, it is established that the impact of this drive towards globalisation was broadly favourable. Productivity and the technological level of production processes improved while import penetration remained minimal despite the scale of trade liberalisation. However, these improvements have to be set beside the fact that the competitive performance of the sector still lagged behind that of major international producers and technological weaknesses persisted. The evolution of the sector's competitive behaviour can only be fully understood when pre-existent internal factors, as well as the externally imposed forces of globalisation are taken into account.

All parts of a country are rarely equally affected by political violence. Yet statistical studies largely fail to address sub-national conflict dynamics. We address this gap studying variations in 'routine' and 'episodic' violence between Indonesian provinces from 1990 to 2003. Within a grievance framework, the article focuses on the violence potential of resource scarcity and population pressure, as well as inter-group dynamics related to polarisation and horizontal inequality. Demographic pressure and inequality seem to have little effect in isolation. However, in provinces where population growth is high, greater levels of inequality between religious groups appear to increase the violence risk.