The 1990s and Beyond

The 1990s and 2000s saw continued work by Oxford development econo­mists on macro issues; there was also a renewed focus on sectoral issues and on technology; and micro research—both quantitative and qualitative— blossomed.

Macro-Analysis: Chris Adam, working both at CSAE and ODID, and David Bevan at CSAE, continued the analysis of the macroeconomy of devel­oping countries, incorporating both a neoclassical framework and sensitivity to institutions and political economy. They were partly inspired by the Bevan, Collier and Gunning approach—in particular, adopting conventional tech­niques of modern macroeconomics to understand the dynamic behaviour of small, open, shock-prone economies, starting from a recognition of the criti­cal importance of understanding the structure and political economy of fiscal and monetary institutions in the relevant countries. This research led to a range of policy applications, and Adam has worked with the Department for International Development (DFID) in Britain and the IMF, among others, advising on macro policies and institutions in low-income countries. Significant and influential policy conclusions covered policy requirements for fiscal adjustment and sustainability in low-income countries, and macro pol­icy responses to sudden aid inflows, such as those associated with the Heavily Indebted Poor Countries (HIPC) initiative.

On fiscal reforms—on the basis of work in Zambia in the 1990s—Adam and Bevan concluded that, in low-income countries, reforms which directly addressed revenue mobilisation underpinned sustained fiscal adjustment, as against a focus on expenditure control which was central to reforms in Latin America and Europe (Adam and Bevan 2004, 2005). They also stressed the critical role of the political economy for protecting recurrent spending on operations and maintenance (Adam and Bevan 2014).

In parallel, Vijay Joshi worked on India's macroeconomy for many decades, as well as advising central economic institutions. With Robert Cassen and others, he analysed India's economic reforms of the early 1990s (Cassen and Joshi 1995). More recently, Joshi's 2017 book, Indids Long Road: The Search for Prosperity, provides a long-term analysis of the Indian economy, arguing that the foundations of rapid, durable and inclusive economic growth in India are weak. He suggests that for India to realise its huge potential, the relations among the State, the market and the private sector need to be comprehensively realigned. Deeper liberalisation is required but far from sufficient. In addition, the State needs to perform its core tasks much more effectively.

Matthew McCartney (who joined the Oxford South Asia Programme in

2011) has also been concerned with India's macroeconomy, taking a political economy approach, more critical of the contribution of liberalisation, explor­ing the role of class interests and the economic impact of geography, regional diversity and discrimination (McCartney 2019). Adopting a similar approach, he has also analysed the social dynamics underlying the Pakistani economy (McCartney and Zaidi 2019).

Sectoral Studies: Recent analysis of agriculture and industry has encom­passed both micro and comparative cross-country analysis.

Analysing agricultural productivity, Douglas Gollin (who joined ODID in

2012) and others showed that low agricultural productivity was a key element explaining income differences across countries (Dercon and Gollin 2014), which he found was partly due to barriers resulting from remoteness and poor spatial connectivity (Gollin and Rogerson 2014). In work with David Lagakos and Michael Waugh, Gollin showed that the gap in agricultural productivity in African economies relative to other countries is a real one and that only a fraction can be explained by omissions and measurement errors (Gollin et al. 2014).

Industry and Productivity in Africa: The 1990s saw a transformation in the fortunes of many African countries. Nonetheless, there was a pervasive failure to create a successful industrial sector. Why this was so formed the central focus of Francis Teal's survey work at CSAE which covered Ghana, Ethiopia and Tanzania. He found, unsurprisingly, that firm labour productivity was, on average, very low. Size emerged as a central issue from the research. Large firms were different in almost all respects from their smaller cousins. They were far more likely to export, they paid higher wages, their capital-labour ratios were much higher, they were far more likely to have paid employees rather than depend on apprentices, and their skill levels were much higher. How these outcomes were related to firm size and how firms became large thus emerged as key research questions.

Panel data helped explain why and how firms grow. It appeared very clearly that the answer was not increasing returns to scale, while skills, as convention­ally measured by the education and experience of the workforce, played only a small part in differences in productivity. What mattered overwhelmingly for increased labour productivity was the higher capital-labour ratio of larger firms. Panel data showed the importance of fixed effects—time-invariant unobservables—as determinants of both firm productivity and worker earnings.

Gollin has also been concerned with the relative failure of industrialisation in Africa, despite rapid urbanisation. Using cross-country data, he found that a heavy dependence on natural resources was one explanation (Gollin et al. 2016), while a high rate of urbanisation was due to the significantly higher living standards in urban than in rural areas (Gollin et al. 2018).

Xiaolan Fu brought QEH research back to the issue of technology. She extended the work of both Lall and Stewart, mainly using evidence from China. Exploring China's industrial policy, she challenged the Washington consensus of the unqualified positive role of FDI and the market and explored the transmission mechanisms through which trade and FDI affect economic development (Fu 2004, 2015). Examining the relationship between foreign technology transfer and indigenous innovation in technology upgrading, Fu et al. (2011) found, in line with the findings of Lall (1992), that ‘despite the potential offered by globalization and a liberal trade regime, the benefits of international technology diffusion can only be delivered with parallel indige­nous innovation efforts...

Fu also extended the theory of appropriate technology to the sector level taking into account sector specificity in technology intensity and dynamics, and demonstrating the critical importance of indigenous innovation in tech­nological upgrading in developing (especially middle-income) countries (Fu and Gong 2011). Technologies created in the South are shown to be more appropriate for developing countries which have similar economic and tech­nological resources and industrial structure (Fu et al. 2011). She has also explored the determinants of innovation in low-income countries, extending her investigations to the informal sector, a seriously under-researched area, for example in Fu et al. (2018). Her work on innovation suggests that countries should adopt an “open national innovation system”, with multiple drivers, market and State, national and international (Fu 2015).

International Dimensions: Developing countries have always been greatly affected by the behaviour of the global economy. Until the 1970s, this influ­ence was mainly through aid flows, FDI and the terms of trade. The debts of developing countries became a major issue in the 1980s as a result of excessive lending and borrowing in the 1970s, leading to stringent adjustment policies and, eventually, some debt write-off. Accelerated growth of trade and capital flows after 1980—“globalisation”—gave developing countries more influence on the world economy, but also created new concerns. How could private capital flows be managed to avoid the major financial crises observed in Asia in the 1990s and recurrently in Mexico? How would changes in the global division of labour affect income inequalities? These were topics explored at QEH by Valpy FitzGerald and Adrian Wood.

FitzGerald came to Oxford in 1992, bringing to ODID and LAC an aca­demic interest in the macroeconomics and finance of middle-income “open economy” developing countries, and professional experience as an economic adviser to Latin American governments.

FitzGerald’s interest in income distribution led him to focus on the role of the progressive taxation of corporations and private wealth as central to the construction of a stable and equitable macroeconomic strategy in support of sustainable development, which in turn had significant implications for inter­national tax cooperation (FitzGerald 2012; FitzGerald and Dayle Siu 2018). He showed that higher taxation of profits would not depress investment or growth, but that achieving it would require much greater collaboration in the regulation of offshore financial centres—his work, including advice to UNCTAD, the OECD and the G24, contributed to the ongoing process of the reform of global tax rules (FitzGerald 2002, 2012). From these theoretical and policy analyses, a fundamental rethinking of the conventional approach to public finance in developing countries followed, taking into account the consequences of their integration in global capital markets (FitzGerald 2019).

Wood became a Professor of International Development at QEH in 2005, working mainly on the causes and effects of the global division of labour. In Wood (1994), he had analysed North-South trade in a Heckscher-Ohlin model in which comparative advantage depended on endowments of four factors of production—skilled labour, literate unskilled labour, illiterate labour, and natural resources—but not on capital because of its international mobility. At QEH, he extended this analysis to show how cheaper communi­cation and travel, by assisting highly skilled workers in the North to cooperate with workers in the South, had widened the wage gap in the North between highly skilled and other workers while narrowing the North-South gap in the wages of other workers (Anderson et al. 2006). He proposed improvements to Heckscher-Ohlin theory (Wood 2011) and estimated the size of the shift in the comparative advantage of the rest of the South away from manufacturing and towards primary production as a result of China's entry into world mar­kets (Wood and Mayer 2011). In another work, Wood showed how the impact of aid on world poverty could be increased by allocating it in a way that took account of differences among developing countries not only in terms of current poverty levels but also in probable rates of poverty decline in the absence of aid (Wood 2008).

Micro Studies: Two approaches can again be distinguished: on the one hand, use of large samples and econometric analysis; on the other, more qualitative and multidisciplinary approaches. In the more quantitative vein, CSAE and the Department of Economics made a large contribution in this area. Prominent economists here were Paul Collier, Marcel Fafchamps, Francis Teal, John Knight, Stefan Dercon and Abigail Barr. Research methods included analysis of cross-section and panel data; randomised experiments; and experimental games. Here, we only give some highlights of the very extensive output.

Household Econometrics, Risk and Uncertainty: Fafchamps (in Oxford from 1999 to 2013) covered a large range of issues, including agricultural markets and market institutions, risk sharing, and analysis of self-help groups, social networks and political violence, as well as child labour, rural poverty and crime. He used a variety of methods, including surveys, randomised experi­ments and game playing (with Barr), often investigating underexplored areas. For example, with others he showed that social networks were effective in transmitting information (Fafchamps and Vicente 2013); that firms in Africa learn by exporting (Bigsten et al. 2004); that there were negative effects of workday taboos on agricultural output (Stifel et al. 2011); that, controlling for population composition and other risk factors, crime in Madagascar ‘increases with distance from urban centres and, with few exceptions, decreases with population density' (Fafchamps and Moser 2003: 625; see also Fafchamps and Ferrara 2012); that informal associations have an insurance and redis­tributive function, using evidence from urban Kenya; and other research showed that child labour decreases with urban proximity, using data from Nepal (Fafchamps and Wahba 2006).

Knight (with Richard Sabot) undertook major surveys of education in Kenya and Tanzania, contrasting secondary school education policies. The research showed that returns to education were not simply due to “signalling”, then a fashionable view, but to the cognitive skills acquired (Knight and Sabot 1990). Analysis of education in Ethiopia found high production externalities to rural education while internal returns were relatively low, helping to explain the low level of education (Weir and Knight 2007). Further work in Africa included analysis of the very high open unemployment rate in South Africa, which he and Kingdon explained by barriers to entry to the small informal sector (Kingdon and Knight 2004).

In the 1990s, partly stimulated by Chinese students at Oxford, Knight joined the China Household Income Project (CHIP), then co-directed by Keith Griffin. His research on China covered poverty, inequality, the labour market, migration, education and happiness, based largely on microdata. Knight and Lina Song were among the first to provide comprehensive cover­age of the widening rural-urban divide in incomes and services, a paradox in an economy emerging from a peasant revolution. Once the early rural reforms had narrowed the income gap, the divide grew for a quarter of a century, assisted by institutional constraints and government policies. The high ratio of urban to rural income per capita began to decline about 2010, essentially because of emerging rural labour scarcity (Knight and Song 1999, 2005).

A key focus of Dercon's extensive research has been the relationship between risks and poverty, conducting surveys in a number of East African countries. He shows that poor rural households adopt a variety of methods to cope with risks, but these are often insufficient to prevent adverse events, leading to extreme poverty among vulnerable households. Women often bear the brunt of the resulting hardship. Macro policies to reduce fluctuations and a variety of State-supported safety nets can be helpful, though care is needed as the lat­ter may undermine self-insurance mechanisms and worsen the position of those not covered (Dercon and Krishnan 2000; Dercon 2002, 2004). Dercon has extended his work on risk to the analysis of the role of social protection mechanisms in humanitarian crises (Clarke and Dercon 2016). Analysis over time has shown that the deleterious effects of shocks are long-lasting (Dercon et al. 2005; Dercon and Porter 2014; Weerdt et al. 2017). Dercon has also made contributions in many other areas, including African agricul­tural development strategies, the impact of industrial jobs on health and incomes, and informational flows and social externalities, among others. He also made a direct contribution to British aid and development policy as Chief Economist at DFID from 2011 to 2017.

Research by Barr (in Oxford, mainly at CSAE, from 1996 to 2011) focused on issues such as how to set up mutually beneficial agreements, what deter­mines individual values and the role of “other-regarding preferences” in deci­sions. She researched these topics by designing incentivised games to generate data: conducting experiments among students, villagers, nurses and doctors in a large number of countries in Africa, Latin America and Europe. For example, she investigated the role of cultural background, or social norms, in determining the propensity to be corrupt; the influence of networks on enter­prise productivity; and the determinants of trust in communities, differentiat­ing between local communities and new settlements (Barr 2002, 2003).

A Case-Study Approach—Village and Town Studies in India: A more qualita­tive approach, developed over several decades, blossomed in this period as findings accumulated. These came from a number of path-breaking long-term village studies in India, revisiting the same village(s) over decades and thereby capturing the process of dynamic change. Again, differences in approach can be seen, with some primarily concerned with economic relationships and outcomes—the Palanpur studies and those of Dercon—and others—Heyer and Harriss-White—adopting a more multidisciplinary approach and viewing village and town economies as embedded in social relations and the surrounding area.

Bliss and Stern first surveyed Palanpur, a small village in Uttar Pradesh in 1974-1975 (Bliss and Stern 1982). This was the baseline survey which pro­vided a benchmark source of comparison with subsequent surveys, conducted at roughly decadal intervals, but with new participants from outside Oxford, many still involving Stern, who had moved to LSE. To date, seven successive surveys have shown rising incomes at a steady 2% per annum, changing con­tractual arrangements with a decline in sharecropping and an increase in rent­als, but no clear difference in productivity between the two types of contract, while cropping intensity and mechanisation in agriculture increased. There was a shift of households into a plurality of activities with agriculture provid­ing a declining share of income and sources outside the village a rising share. Inequality decreased up to the mid-1980s but subsequently increased and there was some shift in average incomes among castes, with some castes seizing new opportunities to a greater extent than others. Intergenerational mobility declined (Himanshu et al. 2018).

Dercon and others examined changes in living conditions in six villages in Andhra Pradesh and Maharashtra, initially surveyed during 1975-1984. They found that both monetary and non-monetary indicators of well-being improved considerably. Migrants experienced faster welfare improvements than non-migrants, but more analysis is needed to confirm whether this was due to their initial characteristics or to migration. Surprisingly, lower caste groups experienced faster poverty declines, although this effect was largely confined to Mahbubnagar in Andhra Pradesh (Badiani et al. 2007).

Judith Heyer studied villages near to Tiruppur, a small town in Western Tamil Nadu in South India, over a thirty-year period, visiting them every decade from 1980. Initially, her concern was with the impact of government- targeted interventions aimed at increasing the assets of poor households, but later examining the drivers of asset distribution over time. She found that the 1970s and 1980s programmes worked reasonably well for small farmers but not for agricultural labourers because they did not have enough of the other resources needed to make assets productive. Food for work and other employ­ment programmes that were also a feature of the 1970s and early 1980s largely excluded Dalits, or included them on less advantageous terms. Poorer house­holds did better under food subsidy programmes which started in the early 1980s (Heyer 2013). From 2008, poorer households also benefited substan­tially from the national employment guarantee programme (MGNREGA— Mahatma Gandhi National Rural Employment Guarantee Act). The fact that Dalits and other poorer households benefited more in the later years was partly a result of changing hierarchies of power at the village level which were less strongly loaded against Dalits as time went on.

In later work, the role of caste and gender became central. The village study was important in demonstrating that caste discrimination did not weaken significantly with industrialisation, as had been widely expected. Moreover, this occurred in a State that was known for its political and social movements in the 1920s and 1930s which were celebrated for breaking down caste power.

The surprising strength of gender discrimination also emerged from Heyer's fieldwork. It had generally been assumed that gender discrimination was much less serious in South India than in the rest of India. Heyer showed both that patriarchal structures put very strong limits on what women could do, and also that female infanticide and foeticide were both significant correlates with capitalist development in the area studied (Heyer 1992, 2016).

Heyer's village-level work showed the importance of rural-urban interac­tions and the need to understand the regional context of village developments.

The villages studied were part of a regional economy in which industrialisation was proceeding apace, and the villages became increasingly integrated into these economic processes of change (Heyer 2013).

Barbara Harriss-White came to Oxford in 1987, already known for her work on agricultural markets. Unlike the earlier research of Griffin and Heyer, she focused on the unequal relationship of producers and traders and showed that traders were responsible for much of the inequality in India's agricultural sector (Harriss-White 1996; Harriss-White and Janakarajan 2004).

From the early 1970s, Harriss-White also undertook long-term studies of ten villages and a small town, Arni, in northern Tamil Nadu. These village studies showed the growth in productivity due to the Green Revolution and subsequent agricultural diversification and then stagnation as a water crisis hit village economies, as well as many aspects of the economics of agrarian pov­erty, and the drivers and impact of rapid institutional change on rural devel­opment (ibid.).

Her research into the long-term development of a small-town economy and its rural hinterland in India from 1973 to 2012 is a rare case study of small-town change. She tracked the changing nature of local business and the workforce, their urban-rural relations, their regulation through civil society organisations and social practices, and their relations to the State and to India's accelerating and dynamic growth (ibid.; Harriss-White 2015). This research showed that plans do not reflect realities on the ground; and, like Heyer, it also showed the persistence of institutions, such as caste and gender, which had been expected to wither, and the ways that rural development is dyna­mised by the urban economy.

Arising from these studies, Harriss-White produced an influential body of work on the importance of the informal economy and the way that social structures regulated accumulation in both the informal and the formal econo­mies (Harriss-White 2003).

3.5