ABSTRACT
Introduction Krueger (1998) said that ‘(trade) liberalisation was good for growth’ and Dollar and Kraay (2000) argued in turn that ‘growth was good for the poor’. Together, such sentiments comprise the most optimistic views of neo-liberal economists, what has become known as the Washington consensus – that not only will economic liberalisation increase economic growth, but the poor are likely to benefit disproportionately. There is no trade-off between equity and efficiency; when growth increases, poverty or even inequality will decline (Wolf 2004: ch. 9). This is the first of three broad ideas about the notion of convergence in economic theory. This orthodox (neo-classical) view is that a free market economy will generate a strong tendency to (unconditional) convergence. Convergence could here be judged as being between income levels of states or regions within a single economy, or income levels between separate countries. This is exemplified by the idea that capital should flow from rich to poor areas in pursuit of higher profits, and labour from poor to rich areas in pursuit of higher wages. The relocation of factors should drive profit and wage rates into equality, and in an environment of decreasing returns to scale equalise income levels. Related is the theory of comparative advantage, which states that liberalisation will push a country (state) to produce and trade according to its relative factor endowments. For a developing country (poorer state) this will imply producing goods and services that are intensive in its abundant factor. The abundant factor typical of developing countries or poor states is typically unskilled labour – whether in conjunction with land (agriculture and agricultural processing) or unskilled labour-intensive manufactured goods. The theory of comparative advantage demonstrates that liberalisation will increase the returns to the abundant factor (wages) at the expense of returns to the scarce factor (capital). Rising wages and increased demand for unskilled labour will thus reduce inequality. As developing countries increasingly shifted towards neo-liberalism during the 1980s there was increasing dissatisfaction with the social and growth outcomes of liberalisation. This was encapsulated by direct concerns for the social costs of adjustment, or that even where growth had occurred, it had not been translated into wider human development. Such concerns were mostly obvious
when the rapid transition to a free market in the former Soviet Union led to economic collapse and a dramatic increase in inequality (Nolan 1995). These concerns were encapsulated in the views that the market had to be supplemented in order for it best to work its magic – the augmented or post-Washington consensus. This second framework is a conditional version of the first hypothesis. The theory of conditional convergence states that income differences between countries or states are determined by various social and economic conditions. Once these have been accounted for, there will be ‘conditional convergence’ dependent on initial conditions. Countries or states will converge on income levels conditional on having certain factors in common; these could be having a high initial level of education, being located close to a port, having strongly enforced property rights, having good institutions of governance, etc. There is an enormous literature and no agreement on what exactly that elusive factor is. The academic chaos is captured in the bewildering variety of ideas contained within the crosscountry growth literature started by Barro (1991). The third theory remains more of a set of ideas than a single whole. These ideas deal with the theory of divergence: that richer states or countries should grow faster than their poorer counterparts. Factors potentially generating divergence are numerous and include a brain drain, a surplus drain, unequal exchange, declining terms of trade and dependency relations. This chapter explores the idea of convergence and divergence using the empirical context of state-level growth in India. The first theme of this chapter is a methodological one. The policies of the post-Washington consensus – domestic and international trade liberalisation, improved infrastructure, better education, and so on – have received widespread support. Such policies are justified for being causally related to India’s current (2003-8) high levels of economic growth. Equally widespread is the view that lack of progress in implementing such policies has led to some states lagging in terms of economic growth and poverty reduction. This chapter notes that the strident rhetoric has very little empirical support. There is almost no clear evidence at state level in India connecting such policies with improved economic growth, and despite the continued failure of lagging states to pursue liberalisation, there are clear signs of convergence in measures of human development between states. Supporters of neoliberalism argue that this empirical paradox simply reflects an empirical problem. Supporting empirical evidence, they argue, awaits only better and longer data sets, more reliable proxy measures for factors such as ‘governance’ or ‘institutions’, and further refinements to econometric techniques. The second theme of this chapter explores the underlying philosophical implications of neo-liberal economic theory. The use of cross-country growth regressions as a methodological tool to analyse economic growth is loaded with ideological presumption. Cross-country growth regressions assume that economic growth operates according to universal laws across all economies through time and space. The policies of the Washington consensus are the most explicit: the notion that all economies are similar and will respond in the same way to the same policy change. In order to run large cross-state regressions, researchers have tightly constrained them-
selves to the assumption of universalism. Conventional growth analysis assumes that growth parameters are homogeneous, that they are identical across countries and states. Each individual state provides evidence used to elucidate this one underlying universal economic relation. An increase in education, for example, is hypothesised to have the same effect on growth in all countries and states. This chapter uses a case study of growth in the state of Kerala to support a more general argument that the growth process does work differently over time and space, and in doing so questions the universalising presumptions of neo-liberal economics. There is an important distinction here. This chapter is exploring inequality between states. It would also be possible to measure intra-state inequality.1 This chapter is organised as follows. Its second section reviews existing empirical studies. The majority argue that divergence in state-level incomes is occurring over time – rich states are growing faster than poor states. A generation of empirical work has focused on using cross-state regression analysis to try to explain this phenomenon. Results remain extremely poor. Conventional responses to this problem in both the wider literature and specific literature on India are discussed here. Such responses tend to assume that the problem is solely an empirical one. The third section argues that there are significant and intrinsic theoretical problems with using cross-state regression analysis to examine the process of growth. The fourth section makes a case for an alternative methodology – the use of case studies – to explore growth between states. The case study of the episode of growth in Kerala in the 1990s is used to highlight a variety of important explanations for the increase in growth that would be missed by the traditional regression methodology. The chapter ends with a concluding section.
