I'm going to spend the next couple of review posts on two papers by Abhijit Banerjee and Esther Duflo, two development economists at the Massachusetts Institute of Technology in the Abdul Latif Jameel Poverty Action Lab (J-PAL). The paper I review today is their 'The Economic Lives of the Poor' (2007), JEP article. The second post will be on their JEP article 'What is middle class about the middle classes around the world?' (both links are to ungated working paper versions of the published papers).


Banerjee and Duflo (2007) 'The Economic Lives of the Poor'
This paper constitutes a review of a few different datasets to give researchers an insight into the everyday lives of the poor through specific statistical commonalities and differences between groups of the poor in different nations. The researchers use the LSMS as well as a detailed dataset that they themselves have compiled in work on Udaipur, India. My main interest is in the African sample.


Banerjee and Duflo first comment on their use of the $1/day extreme poor and $2/day poor poverty lines and the fact that these can be considered arbitrary depending on the way in which we conceive of poverty, but they use them to be consistent with the majority of the literature. All well and good.


They go on to quote statistic after statistic of the types of consumption, access to savings, urban-rural contrasts and differences between countries in terms of the behaviour of the poor and extreme poor. One thing that struck me was the way in which, yet again, 'the poor' and 'the extreme poor' are lumped into categories that seem to be devoid of cultural impact. When there are differences in consumption of, say, radio vs. television the first thing to which they refer is the possibility of lack of access to broadcasting. Whereas one of the first things I would think of would be: what do we associate with owning televisions? Specifically, Veblen's idea of conspicuous consumption. Yes, even when you are poor or extremely poor you too could be interested in conspicuously consuming (maybe even more so?) in order to gain reputation, or maybe to make people think things aren't as hard as they seem. This motivation seems entirely absent from Banerjee and Duflo's analysis, even when they admit that their sample shows the fairly typical pattern of patterns of food consumption that are not as high as we would expect ex ante. Many of the people in the sample eat less than they might do, while consuming other non-necessity goods.


Nevertheless, here are some of the common factors that they show: (HH = household)

• extended families live under one roof is common
• the poor of the world are young, either because as they get older they get wealthier or because the poor die sooner
• food represents between 56-78% of consumption of poor HHs.
• of non-food items, alcohol and tobacco are prominent
• spending on festivals is an important part of the budget for the poor, including weddings, funerals and religious festivals - in SA specifically, 90% of the poor allocated funding to festivals.
• the poor spend little on entertainment (movies & theatre)
• Pertinent to my considerations, even the extremely poor do not buy the most calorie-cost efficient grains they often choose to buy more costly grains.
  
• Share of rural HHs owning TV is substantially higher for poor than extreme poor.
  
• many poor run own businesses but with very few assets
• self-reported levels of happiness by the poor are not particularly low
  
• the poor experience high levels of stress: many poor have higher morbidity and can give up meals (both adults and children) in order to have more consistent meals.
  
• education spending is low for most poor, mainly because education is paid for by their governments
• the poor are engaged in multiple occupations (although not in SA) and do not accumulate specific skills & their businesses often operate at too small a scale
  
• the poor often temporarily migrate for work
• the poor have low access to credit and low access to adequate (formal) savings mechanisms, markets for land and insurance are inadequately structured to suit the poor
  
• access to water and electricity is better for the urban poor than the rural poor.
  

Subsequent to which they consider 6 questions:

1. Why so little specialization? Risk is a problem - diversity spreads risk and/or the poor cannot raise the capital to occupy themselves fully in one trade.
2. Why so many entrepreneurs? few skills, no capital & female being an entrepreneur is easier than finding a job, and borrowing is risky and unlikely to find individuals willing to lend.
3. Why don't the poor eat more? Eating more wouldn't help much, but some improvements are linked to increased productivity. Is it lack of self-control? No, because they seem to save for festivals, or they are somehow credit-worthy when it comes to festivals.
4. Why don't the poor spend more on education? Parents don't seem to react to low school quality, possibly because many parents are themselves illiterate and unable to assess quality of education.
5. Why don't the poor save more? They spend on tobacco & alcohol, festivals, tea, coffee and sugar, why don't they cut down and save, or buy machines to improve productivity? But saving at home is hard, & difficult to resist temptation of spending on goods that others have, households report wanting to cut expenditure, but don't. Is it lack of self-control?
6. Why don't the poor migrate for longer? Social networks and connections to home, informal insurance networks & other such. High social value of remaining connected.
   

There comments on these seem inconsistent - the poor have self-control when it comes to saving for festivals, but otherwise not. The poor could make more money if they stayed away from home for longer, but they don't because of lack of commitment (165). What I don't understand is that they do not admit culture at all into the ways in which they deal with 'the economic lives', as though 'economics lives' are independent of cultural factors that are extant in a given society. This is, I think, strange. I know that Economics has not traditionally admitted 'culture', but it is becoming more and more a salient feature when discussing economics, especially development economics, as culture impacts directly on the types of institutions that groups have.


Notwithstanding my culture gripe, I have a bit of a problem with claims emanating from the data. Data for 8 of the 14 countries is from the 1990s. For South Africa the data is from 1993 (the same year in which the comprehensive PSLSD, or SALDRU93, dataset was undertaken). In South Africa, for example, it is well understood that the lives of the poor, specifically with respect to access to infrastructure, housing and land reform has improved (Bhorat and Kanbur, 2005). It has not improved as well as we might like to hope, nevertheless it has improved. I cannot legitimately comment on the situations in the other countries. In lieu of this, I don't know if the piece can legitimately be called 'the economic lives of the poor' as that gives the reader the incorrect idea that it is 'the economic lives of the poor as they are living now' which is not what the paper manages to relate (in my opinion this is not assisted by the convention to write in the present tense, for example, "Infant mortality among the extremely poor is 8.7% in South Africa" (160, my emphasis)). It is not legitimate to claim that infant mortality at the end of Apartheid is the same as infant mortality in South Africa now. I understand that it is a linguistic rather than technical issue, but it still carries weight for me.


Nevertheless, I think that the piece fills a relevant gap in the literature. It highlights many issues in the lives of the poor over the last two decades. It does however have many flaws, not the least of which is the lack of discussion of the salience of culture in different societies. Lastly, I want to comment on the lack of references to researchers in the countries that the research, India aside. This is a consistent issue for development researchers in countries outside the US whose work seems not to attract as much attention as development economists in the US and, as such, is often deeply problematic because they fail to recognise the fantastic work (and use of RECENT DATA) that goes on in those countries. I'll leave it at that. Good paper, but could be better.


References
Banerjee, A.V., Duflo, E. (2007). The Economic Lives of the Poor. Journal of Economic Perspectives, 21(1), 141-167. DOI: 10.1257/jep.21.1.141

Banerjee, A.V., Duflo, E. (2008). What is Middle Class about the Middle Classes around the World?. Journal of Economic Perspectives, 22(2), 3-28. DOI: 10.1257/jep.22.2.3


Bhorat, H., Kanbur, R. (2005), Poverty and Well-Being In Post-Apartheid South Africa: An Overview of Data, Outcomes and Policy, Working Papers 9620, University of Cape Town, Development Policy Research Unit (this is actually the introductory chapter to the book, Poverty and Policy in Post-Apartheid South Africa)