The importance of qualitative research

In a previous blog post, I discussed the content of Tariq Thachil’s book “Elite Parties, poor voters. How social services win votes in India?” In this book, he tries to understand how the Hindu nationalist Bharatiya Janata Party (BJP), identified as an elite party, massively gained votes from marginal and poor citizens in the national elections of 2004 compared to the national elections of 1996. This party has been so successful among poor voters that the party’s national leader is the current prime minister of India since 2014.

In this blog post, I want to zoom on a particular chapter (chapter 5), which is a nice illustration of how qualitative and quantitative tools can be used as complements to each other to answer a question.

Chapter 5 is entitled “How services win votes”. The goal of Tariq Thachil in this chapter is to pin down the mechanisms underlying the observed correlation between votes for BJP and services to the poor. To do so, he focuses on a specific State, Chhattisgarh, where there was a surge in services targeted to the poor as well as a surge in the vote share for the BJP. Chhattisgarh has a very large tribal marginalized community (31.8 % of the total population), which has historically largely voted for the Congress, the party of Independence.

To understand if the observed increase in votes for the BJP is a consequence of the increase of services to the poor, Tariq Thachil divides the big question “how services win votes” into several logical subquestions.

1) He first tries to understand how the services to the poor by an organization that is related to a political party got accepted by the tribal villagers.

2) He then explores how the tribal villagers moved from acceptance to use of the services.

3) Then he underlines the strategy used by the organization to translate the use of services into votes.

4) He finally shows that tribal people that benefited from the services offered by the organization tend to vote more for the BJP, as well as other people in the villages where the organization was providing services.

While point 4 is a fact that can be “easily” observed using survey data, points 1 to 3 are mechanisms or “channels” that are hard to identify without fieldwork. To understand point 1, Tariq Thachil interviewed several activists from the organization, who revealed that they were introducing themselves as “ideologically neutral service providers”, so that villagers would accept their presence in the village. To answer question 2, he asked villagers what was different in the services provided by the organization compared to public services. For schooling services they answered that the infrastructures were not as good as the public schools, that the cost was higher, but that the quality of education was much better. In particular, the teachers were present all the time, and they were focused on transmitting knowledge to the kids. Finally, to pin down point 3, he again interviewed activists, who detailed their strategy to convince people to vote for the BJP. They explained that thanks to their provision of services they had acquired a high status among villagers, and villagers would listen to their advice. So they would organize meetings where they would voice their opinion about politics, not as represents of the BJP, but as “non-political individual citizens”.

This book is not only interesting because it provides convincing answers to a puzzle, namely “why poor people vote for an elite party”, but also because it illustrates how several research tools can be complementary to each other. In this context, Tariq Thachil would not have been able to properly answer the question without a mix of fieldwork and econometric analysis.  Mixing fieldwork and econometric analysis enables him to go beyond mere correlations between service use and votes. In fact, the observed fact that service users vote more for the BJP does not mean that services have an impact on votes. The correlation could be driven by unobserved characteristics – BJP voters have different unobserved characteristics and these characteristics also impact the probability to use services- or by reversal causality: BJP voters could enroll more into services provided by organization that are close to the BJP. This mix usage of methodologies is particularly important because the observed fact (poor people voting for the BJP) is counter-intuitive. As economists, we do not always have the opportunity or the institutional incentives to gather first-hand qualitative observations. It is why, for some questions, collaboration with researchers from other fields, such as political scientists, sociologists, historians or anthropologists can help figuring out the mechanisms underlying observed facts.

Posted in Books | Tagged , , , , | Leave a comment

Poverty and the present bias

It is well documented that people living in material scarcity conditions make different choices from other people. One example of such differences is the behavior in situations of intertemporal choices. A key concept is that of present bias. When subjects have to choose between a benefit at some point in time and a larger benefit later on, the choice does not only depend on the amount of the benefits and the time difference between them, as would be consistent with the value of time consistency, but also on whether the first benefit arrives today or not. For instance, a lot of subjects would prefer receiving €120 in two weeks over €100 in one week, but would prefer receiving €100 today over €120 in a week from now. A vast majority of people exhibit this present bias in all social groups. The present bias (measured by the additional interest rate at which subjects are ready to give up an immediate benefit), however, has been shown to be larger among people living in material scarcity conditions.
Continue reading

Posted in Uncategorized | Leave a comment

Why would poor people vote for an elite party? The BJP victory puzzle in India

In a previous blog post “Electoral turnout, why should we care“, I mentioned that India was an outlier with respect to the voting behavior of poor people, because poor people tend to have a higher turnout than rich people, whereas it is the contrary in other countries. Another puzzle in Indian politics is the victory of a so-called “elite party”, the Bharatiya Janata Party (BJP) in the national elections in 2014. This puzzle is addressed in the book of Tariq Thachil “Elite Parties, poor voters. How social services win votes in India?” published in Cambridge University Press. In this book Tariq Thachil tries to understand how the BJP massively gained votes from marginal and poor citizens, without doing a programmatic shift, and without losing votes from the rich.

His hypothesis is that the BJP gained votes from the poor by privately providing them with services such as education or health through affiliated welfare organizations. Outsourcing service provision is a way to gain votes from the poor while retaining votes from the rich because no programmatic change is needed to capture this new electorate. To test this hypothesis Tariq Thachil first rules out alternative explanations. He first shows that there was no programmatic shift towards lower classes’ interests.  For example he shows that the BJP had “a reduced commitment to pro-poor public spending”. Secondly, it is also commonly believed that the BJP gained votes by crystallizing the attention of people around religious questions. He rules out this potential explanation by showing that the prevalence of religious riots in Indian States is not correlated to vote gains from lower castes for the BJP. Third, he demonstrates that the gain in votes is not driven by a strategy of inclusion of lower castes candidates or by joining forces with parties that usually appeal to low-caste voters.

He then turns to a direct test of his hypothesis. Using econometric analysis and data from the National Election Study, a post-poll survey, he analyses the determinants of voting behavior. He tests the hypothesis that poor and rich voters vote for the BJP for different reasons. His hypothesis is that welfare organizations were key determinants of the votes from the poor for the BJP but did not affect the votes of the rich. So poor voters should be influenced in their vote by the presence of welfare organizations but not by the program of the BJP. Rich voters, which did not benefit from services, should be influenced by ideology and not by welfare organizations. And this is exactly what he finds. Non-party organization affiliation, which includes religious or welfare organization affiliation, strongly predicts voting for BJP for lower castes, but not for higher castes. And ideology does not predict voting BJP for lower castes, whereas it does so for higher castes.

Next, focusing on a State where the progression of the BJP was particularly important, Chhattisgarh, he studies if the vote share for BJP was higher in villages that have services from welfare organizations and within villages if villagers that benefited from the services have a higher probability to vote for the BJP. He finds robust evidence that this is the case.

While the question that Tariq Thachil addresses is specific to India, the answer he provides is relevant for other contexts and explains similar patterns that were observed in other countries. In particular, Tariq Thachil argues that the strategy of the Muslim Brotherhood in Egypt, Yemen and Indonesia was similar to the one of the BJP. More generally, it shows how an increase in votes for a specific party does not always mean that people are more attracted by the ideology of this party.



Posted in Books, Poverty, Welfare | Tagged , , , | 1 Comment

Child poverty and policies.

In many countries there is a higher incidence of income poverty among children than in the adult population.[1] This is particularly acute in developing areas like Latin America but affects developed nations too.[2] In the United States (US), for instance, the official income poverty for 2014 shows that more than 1 in 5 children are poor whereas adult poverty is well below 15%.[3] A growing literature shows that adverse individual circumstances during childhood can have long-lasting effects. Besides efficiency considerations, this is unacceptable from an ethical point of view if we embrace the equality of opportunity approach, and we consider that decisions that affect individuals during childhood, at least the most relevant ones, are out of children’s control. Given all this, it is arguably key to accumulate evidence on the long run effect of childhood circumstances into adulthood. In this post, I would briefly review three recent empirical papers that show how different policies targeting or affecting low-income children have had important long-term effects on the beneficiaries.

Continue reading

Posted in Poverty, Scientific Papers, Uncategorized | Tagged , , , | Leave a comment

Millennium Development Goals: evaluation and beyond

In the year 2000, at the dawn of the twenty-first century, world leaders gathered at the United Nations in order to establish an ambitious project aimed at fighting global poverty in all its forms. They adopted the so-called eight Millennium Development Goals (MDG). These goals are quantified anti-poverty targets that the world committed to achieve by 2015. The spectrum of these international development goals was not restricted to resource poverty, but also covered other key aspects of human well-being. The eight millenium goals were defined as follows:

  1. Eradicate extreme poverty and hunger
  2. Achieve universal primary education
  3. Promote gender equality and empower women
  4. Reduce child mortality
  5. Improve maternal health
  6. Combat HIV/AIDS, malaria, and other diseases
  7. Ensure environmental sustainability
  8. Develop a global partnership for development

The United Nations’ 2015 report on the MDG [1] offers an evaluation of the progress achieved since the MDG’s adoption. In a nutshell, there are reasons to celebrate the success of the MDG as millions of lives have been saved and the living conditions of many people have been improved. Substantial improvements have been registered in the eight different goals. Nevertheless, the targets have not all been reached. Furthermore, some countries and regions did much better than others and many among the most vulnerable individuals have been left behind. This means that extra efforts are necessary. Continue reading

Posted in Inequality, Poverty | Tagged , , , | 1 Comment

Inheritance, dependence and poverty

A few weeks ago, I attended a conference [i] which gathered social scientists (sociologists, historians, economists, legal experts) working on a similar topic using different approaches. This year’s topic was “Inheritance and Bequest”[ii]. I had the opportunity to listen to Emilia Schijman, a young sociologist, and I found her talk fascinating.

Her talk was about bequests in a context of poverty in Argentina. At first sight, it seems that bequests and poverty do not get along very well: poor people do not own anything, so what can they leave as bequest? But this is a matter of scale. People living in social housing in Buenos Aires do own something: an occupancy title. Although the occupancy title is supposed to belong to one person, a sub-market has emerged and the owner of the title can yield it, sell it out, lease-out the apartment and even leave it as bequest.

Continue reading

Posted in Events, Poverty | Tagged , , , , | Leave a comment

When does unequal growth reduce income poverty? A new index combining absolute and relative dimensions of poverty

We are happy to share a guest post published by one of our regular contributors, Benoit Decerf, on the World Bank’s Development Impact website.

Read more about his Job Market paper here:


Posted in Inequality, Poverty, Scientific Papers, Welfare | Leave a comment