Measuring poverty more complex than it looks
In April, Statistics South Africa released its report “Poverty trends in South Africa, 2006-2011”. The picture it describes is largely positive. After almost two decades with little progress towards the reduction of poverty, the results show notable changes.
The percentage of the population who are categorised as being poor has dropped from 58 percent in 1995 to 45 percent in 2011.
There are variations by province, with the percentage that are poor lowest in Gauteng and the Western Cape, at 25 percent and 23 percent, compared with 65 percent in KwaZulu-Natal, almost 64 percent in Limpopo and 61 percent in the Eastern Cape. The report also shows variation in the trends. Gauteng and the Western Cape experienced the largest reductions in poverty from 2006 to 2011, at 33 percent and 29 percent.
Importantly, much of the decrease has taken place among the most poor as is reflected in the indicators showing changes in the depth of poverty.
However there has been comment that the poverty lines used by Stats SA are unrealistic and are too low.
This is a common reaction to the use of poverty lines which are often thought of as being unacceptably conservative.
Further, the usefulness of dividing society into “poor” and “non-poor” is frequently questioned since most policies cannot be targeted with the same precision.
It is helpful to consider the purpose, history and methods involved in poverty measurement in order to respond to such concerns. Poverty is conceptualised in many ways, and the financial approach which makes use of poverty lines is only one approach.
Nonetheless it is the most widely used and has a long history. This can be traced back to 1894 when Charles Booth is credited as the first to make use of a poverty line.
It was fixed at between 18 and 21 shillings a week for a family of five living in London. Once adjusted for a century of inflation, this works out at around R600 per person per month.
Coincidentally this is one of the international poverty lines used by the UN.
South Africa makes use of a suite of poverty lines. In March 2011, these were the Food Poverty Line (R321 per person per month), Lower Bound Poverty Line (R443) and Upper Bound Poverty Line (R620). The Stats SA report applies these lines using data from large sample household surveys that form part of our national statistics system.
These have similar methodologies and gather detailed information on the income and expenditure patterns of South African households using a diary method.
This means that respondents are asked to keep a diary for several weeks recording all of their expenses. As is common with most official statistics, the samples are large at around 30 000 households and so are representative of the nine provinces.
They are also rigorous with a sample frame derived from Census data and data collection procedures that adhere to international standards set by the different branches of the UN statistical system.
For a statistics agency, the poverty measurements produced with these lines have a purpose similar to medical tests used to narrow down the likely causes of an ailment and so enable the correct treatment to be prescribed.
However it must be acknowledged that poverty measures are associated with contentious political choices.
These include the level of support provided by public policy, how minimum wages are determined or as the administrative threshold beyond which eligibility for public funds is withdrawn.
Poverty measures are also used in times of political competition as evidence of the success of alternative policies or administrations. This is not the intention of the lines contained in the Stats SA report.
The poverty lines are not a recommendation of what people need to get by. They are rather an indication of when we can be certain that they do not.
Below a poverty line, we can be sure that people are eating insufficient or inadequate food and, ultimately, are at risk of lives that are less healthy and shorter than those who are “non-poor”.
Further, the trends described are the result of many factors including decisions made by local and international business.
A closer reading of Stats SA’s report shows that the percentage of people categorised as poor increased in 2008. This is unsurprising given the global financial crisis experienced in that year.
Careful interpretation of measurements and the formulation of appropriate responses are necessary if we are to move beyond counting or boasting and towards the promotion of social justice.
* Julian May is the director of the DST/NRF Centre of Excellence in Food Security at the University of the Western Cape, and is a member of the South African Statistics Council. He writes in his personal capacity.