ID21 - Insights Issue 31

		ID21 Home
		Insights
		Issue #31
		For richer, for fairer- poverty reduction and income distribution

		Efficiency versus equity? Wage waves in China

		Seeds of hope? Is the Green Revolution coming for Africa?

		Measuring pro-poor growth in rural India

		Earnings off the farm: magic bullet or myth?

		Ethiopia after reform: why some poor got poorer

		Storm clouds over Asia: signs of a silver lining?

		Sites for Sore Eyes

September 1999 Insights Issue #31

Back to Insights #31

Tracking pro-poor growth

New ways to spot the biases and benefits

Promoting 'pro-poor' growth has been a commanding idea in development economics since the 1970s. It was central to the concept of 'broad-based growth' in the World Development Report of 1990. Lately, major donors have explicitly committed themselves to encouraging economic growth that benefits the poor. Yet little is known about what pro-poor growth really means or how to gauge it. New work by Institute of Development Studies researchers set out to establish a clear and simple definition of pro-poor growth, using a measure called the 'poverty bias of growth' (PBG). Then they applied it to test situations surveyed in contrasting regions of India.

The importance of increasing the rates of growth of income in poverty groups was emphasised in seminal 1974 work on Redistribution with Growth by Hollis Chenery and others. Oft-repeated citations from that volume underline the need for growth to:

be labour intensive and create employment, as labour is the poor's most plentiful asset

be concentrated in rural areas, where the majority of poor people tend to live

be focused on activities/products most vital to the living standards of poor people.

The IDS study sought to establish and test a clear and simple definition of pro-poor growth. A measure called the 'poverty bias of growth' (PBG) was calculated. It was derived by subtracting changes in the poverty headcount that occurred between any two periods under actual circumstances, from the change in poverty that would have occurred if all had gained equally. This measure was then applied to test data from two Indian states, Andhra Pradesh and Uttar Pradesh. The results revealed that:

Between 1973 and 1989 growth in Andhra Pradesh was pro-poor whilst growth in Uttar Pradesh was biased against the poor (see charts, below).

Both states secured substantial reductions in the incidence, depth and severity of poverty during the period in question. The poverty headcount in Andhra Pradesh fell by 24 percentage points, that in Uttar Pradesh by 22 points.

Cumulative distribution of per capita expenditure

Chart1

The pro-poor shift in distribution in Andhra Pradesh meant reduction in the poverty headcount was one percentage point greater than if all had benefited in equal measure: thus, the PBG in this case is plus one.

By contrast, worsening distribution in Uttar Pradesh reduced gains to the poor there. Some 18 percent of the fall in the poverty headcount that would have occurred had growth been evenly spread, was forfeited on account of this anti-poor bias.

The PBG provides a simple and intuitive way to assess the extent to which economic growth translates into poverty reduction. The study report's authors warn, however, that the PBG method has shortcomings:

In common with the poverty measures used in its calculation, the PBG ignores the welfare of households above the poverty line. So if a distributional change reduced inequality amongst the poor at the same time as increasing inequality for the population as a whole judged by standard measures, the PBG would (correctly) expose this as a pro-poor shift.

Because of the nature of currently available data, the PBG is 'gender blind' - indeed it does not consider any aspect of the intra-household distribution of welfare.

Planners should be wary of using PBG to assess budgetary spending priorities or the efficacy of anti-poverty policy, because changes in mean income and distribution of income already reflect the influence of initial conditions and existing policy interventions. For example, if growth between two periods is known to have had an anti-poor bias, this effect cannot necessarily be attributed to policy interventions launched at that time. Such interventions may have forestalled an even larger anti-poor bias to growth. Conversely a pro-poor bias to growth is not necessarily indicative of sound policymaking.

Contributor(s): Neil McCulloch and Bob Baulch

Further information:
Neil McCulloch
Institute of Development Studies at the University of Sussex
Brighton
BN1 9RE
UK

Tel: +44 (0)1273 877223
Fax: +44 (0)1273 621202
Email: neilm@ids.ac.uk
Institute of Development Studies

Other related links:
Search Eldis for sources on Poverty

FREE Information Delivery services from ID21:

ID21 is enabled by the UK Government Department for International Development(www.dfid.gov.uk) and hosted by the Institute of Development Studies (www.ids.ac.uk/ids), at the University of Sussex, UK. Charitable Company No. 877338. ID21 is a oneworld.net (www.oneworld.org) partner and a mediachannel affiliate (www.mediachannel.org).

Views expressed in INSIGHTS are not necessarily those of DFID, IDS, id21 or other contributing institutions. Copyright remains with the original authors but (unless stated otherwise) articles may be copied or quoted without restriction, provided id21 and originating author(s) and
institution(s) are acknowledged.