Index > Publications > Publication Details

Publication Details

Bringing politics back into poverty analysis: Why understanding social relations matters more for policy on chronic poverty than measurement

John Harriss


Mainstream poverty research - even after experts had generally accepted the need for a multidimensional view of poverty that goes beyond income/consumption measures to take account of holdings of assets and hence of longer run security (see Chambers 1988, 1992) - has generally failed to address the dynamic, structural and relational factors that give rise to poverty. There is a great deal of technically sophisticated research, much of it based on household surveys, that has provided ever more detailed profiles of poverty in different countries and regions. This research has also produced a number of studies of 'poverty dynamics' that show the implications, for example, of the distribution of assets in a society or of access to human capital. One such study is by the World Bank entitled 'Why some people escape from poverty and others don't' (Grootaert et al 1995). In general this research tends to converge around much the same conclusions: household characteristics, especially dependency ratios, matter; ownership of assets is highly significant; access to insurance such as that provided by holding a regular, secure job, or through being able to claim such a resource as a ration card, matters; and education counts for a lot. Lately these studies have been extended to take account into social relationships through the concept of social capital. But it is perhaps still the case - as exemplified by Guatemala (Ibanez et al. 2002) - that it is the better educated, relatively wealthy, middle-aged men who enjoy most of the social capital. Being very poor, on the other hand, seriously constrains people's abilities to invest in social capital, even within the family (Cleaver 2005). So it is far from being clear that this factor cuts through the self-reinforcing circle of factors that are associated with movements out of poverty (see also Adato, Carter and May 2006; Kumar and Corbridge 2002). Little, if any, of the earlier research aims to address the questions of how and why. Why it is that the factors under consideration are distributed through a society in the manner that they are? These questions refer to the political economy of contemporary capitalism, and to cultural politics. The fact that they are largely ignored shows that poverty research plays a role in depoliticizing what are in essence political problems. It is a part of what James Ferguson (1990) memorably describes as the 'anti-politics machine'. Poverty research in international development shares in 'the idea that scientific knowledge holds the key to solving social problems' which, according to O'Connor, 'has long been an article of faith in American liberalism' (2001: 3). If only - the implicit reasoning runs - 'we' can build a good scientific understanding of poverty, then 'we' will be able to solve the problem. But the reality is that poverty knowledge is profoundly political, as the contemporary debates over poverty trends in India in the 1990s so clearly show (Deaton and Kozel 2004). The problem is that even in the most sophisticated poverty measurements, long chains of assumptions are necessarily made so that these are always open to question. And the assumptions specialists most readily accept depend on value judgements. As O'Connor argues, poverty research, dominated in the case of international development by people educated in a small number of mainly American universities, is an exercise in power. This has been recognized in recent years at the centre of poverty knowledge, the World Bank, in its celebrated study, Voices of the Poor. But that study, which argues for a different model of knowledge as the basis for poverty action, has been ignored. Poverty research seems to indicate that the social sciences should not try to emulate the natural sciences (Flyvberg 2001). They are more effective in generating the kind of knowledge that develops from familiarity with practice in particular contexts, helping people to question relationships of knowledge and power, such as those giving rise to poverty, and subsequently to work to produce change. Such a view has quite profound implications for the design of poverty research.

Publication Type(s)

CPRC Working Paper


intergenerational transmission of poverty policy social relations data childhood poverty

ISBN: 1-904049-76-1


Bringing politics back into poverty analysis PDF 139.1 KB

Back to previous page