Thursday, December 18, 2014

Does ethnic diversity decrease economic interactions? Evidence from exchange networks in rural Gambia

As the end of the year is approaching, a little summary of my research work during 2014. 

This year I spent a lot of time in the field collecting the data for the follow-up survey of the Gambia Networks Project. We are currently cleaning the new data and looking forward to have a panel data of economic networks of our beloved villages. 

In the meantime, I kept working on the papers that use the baseline data. In a new working paper, joint with my Maestro Jean-Louis Arcand, we present evidence that ethnic diversity does not reduce economic interaction in rural Gambian villages:   

Using a unique dataset collected in 59 rural Gambian villages, we study how ethnic heterogeneity is related to the structure of four economic exchange networks: land, labor, inputs and credit. We find that different measures of village-level ethnic fragmentation are mostly uncorrelated with network structure. At a more disaggregated level, household heads belonging to ethnic minorities are not less central than those from the predominant ethnicity in any of the networks and, at the dyadic level, the fact that two households share ethnicity is not an economically significant predictor of link formation. Our results indicate that, in the particular setting of our study, the structure of the exchange networks is better defined by other variables than ethnicity, and that ethnic heterogeneity is unlikely to be a driver for sub-optimal economic exchanges. We argue that our findings can be interpreted in a causal way as the current distribution of ethnic groups in rural Gambia is largely influenced by specific historical features of the British colonial administration. Moreover, the network structure of our data allow us to include fixed effects at different levels as well as to precisely measure kinship ties, a confounding variable often omitted in previous studies

Wednesday, December 10, 2014

Missing Links, Missing Markets: Evidence of the Transformation Process in the Economic Networks of Gambian Villages

As the end of the year is approaching, a little summary of my research work of 2014.

After more than four years of work, I am very proud that eventually "missing links, missing markets", one of the papers from the Gambia Networks Project, will be published in 2015 at World Development:

The aim of this paper is to contribute to the empirical analysis of the transformation process in traditional rural societies using a network perspective. A unique database collected in 60 villages in rural Gambia is used to study the ways in which households with links outside the village (a proxy for market connections) behave in the locally available exchange networks for land, labor, inputs, and credit. The econometric results at different levels of disaggregation provide suggestive evidence of substitutability between internal and external economic interactions, particularly in the case of reciprocal exchanges.

Here the original post about the paper, here Marc Bellemare's review.

Friday, December 5, 2014

Multilateral determinants of regionalism revisited

As the end of the year is approaching, a little summary of my research work of 2014.

In June, with Katerina Gradeva we published our "revisit" of the almost classic paper "Multilateral determinants of regionalism" (Mansfield and Reinhardt, 2003), and we found little support for the original results:

The idea that some features of the multilateral trading system create incentives for countries to join preferential trade agreements (PTAs) is among the first and most influential explanations for the wave of regionalism in the last decades. Until recently, only a few empirical studies have explored this hypothesis and their results have been accepted by many researchers and policy-makers to be a fact. In this study we revisit the question of whether multilateral events are important determinants of regionalism. We use an extended dataset and implement several empirical specifications in the analysis. Unlike previous work, our results provide little support for the relevance of variables such as the number of GATT/WTO members, ongoing trade negotiation rounds, and trade disputes as predictors of PTA formation.

Tuesday, April 9, 2013

The Asian roots of cultural change in The Gambia

In the last 3 weeks I have been conducting fieldwork in The Gambia and Senegal in the framework of the project AFRASO, an initiative of University of Frankfurt to understand Asia-Africa interactions in a multidisciplinary perspective. My focus is on development cooperation in the agricultural sector. And in the case of The Gambia, the Republic of China (ROC, Taiwan) is one of the most important bilateral donors, particularly after 2005. When then president of Senegal Abdoulaye Wade decided to switch diplomatic recognition from ROC to Mainland China, the Taiwanese technical mission was forced to leave the country and reallocated part of their staff to the "smiling coast".  

Some local cultural aspects have become a big headache for the Taiwanese efforts to promote development cooperation as a way to keep on their side one of the few diplomatic allies they still have in Africa. Since British colonial times, one of the problems for agricultural endeavours in The Gambia has been the deeply ingrained gender division of labor in Mandinka culture, the predominant ethnicity, which has also been adopted by other groups, like the Fulani. While men cultivate groundnut, millet, and fruits, rice and horticulture are the exclusive domain of women. 

The main elements of the Taiwanese agricultural technical cooperation in The Gambia are rice production (which I will describe in detail in future posts) and "women gardens". Previous research has shown that indeed one of the main problems for the sustainability of the production in the gardens is the lack of men's involvement in any of the tasks. But this seems to be changing... 

As can be seen in the video below, captured during our visit to the most famous of the gardens (Bajulunding, close the the capital city and the airport), now there are many men involved in the production. It seems like in recent years the Taiwanese have convinced some guys to work there by directly hiring them for a portion of the garden dedicated to the consumption of the embassy staff and for some concrete tasks as operating the machines. 

But this is not the only change. In the video it is possible to see boys working side-by-side with their mothers in the gardens! This is unthinkable in rural villages and was in this same garden some years ago. Maybe it is too early to say, but this can be part of an Asian-driven cultural change in The Gambia. 

Friday, March 15, 2013

A socio-economic characterization of returnee households in the Nuba Mountains

This weekend will start the CSAE conference 2013 in Oxford, the largest gathering of experts on Africa related economic topics and one of the most important conferences about economic development in general.

My co-author Asha Abdel Rahim (University of Juba) will be presenting our work (also joint with Aleksi Ylonen) about households returning to their communities of origin in the Nuba Mountains after the signature of the Comprehensive Peace Agreement that ended Sudan's civil war in 2005.

The session will be available in live web broadcast  here.

Asha collected this data in 2008, when expectations of a long lasting peace were high and the coexistence  between former enemies was not utopical. Unfortunately, shortly after the data was collected the situation deteriorated and the war resumed in the area in 2011, the same period when the former Nuba allies in the south were preparing to secede and create South Sudan. Therefore the Nuba area was transformed in the southern border of Sudan instead of being the (still always isolated) center of the country.

The inherent problems of data collection in post conflict areas implied that some important information (like place of displacement or ethnicity) was not registered, and the restart of the conflict impeded any follow-up. Given these shortcomings, it is difficult to attempt a causal analysis of the effects of displacement  but the uniqueness of the data make it worthy to be described in detail.

In particular, we are interested in compare returnee households with those who stayed in the villages during the war. Around 300,000 persons returned to Nuba communities after the CPA, and in the 8 villages of our sample (344 households), around 40% of the households are returnees.

We actually find many differences:

- Returnees tend to own less assets than stayers. (This is not surprising, and in line with previous findings by Ibanez and Moya, 2010; Fiala, 2012, among others).

- Returnee households cultivate more agricultural varieties, but mainly staples. Stayers are more likely to cultivate cash crops. (Previous studies had shown differences in production structure, but I am not aware of previous findings mentioning this staple/cash crop differences)

Returnees have better health outcomes, in the sense of household members having a lower probability of contracting serious diseases. (This is somehow in contradiction to Verwimp, 2012, but in line with the findings of Hynes et al. 2002)

- We actually relate the last finding to two possible explanations (apart from self selection into displacement): The fact that returnees tend to have better hygiene habits (in terms of washing their hands and other attitudes, potentially learned during displaced) and the targeted support of NGOs in the post conflict period.

We hope that our contribution can help to the post conflict efforts when the current devastating,  and largely unknown and neglected, conflict in the Nuba Mountains comes to and end. Hopefully soon.

A good source of information of the current events in the Nuba Mountains is NubaReports:

Monday, March 4, 2013

Conflicts and Economic Development

I am currently preparing the course that I will be teaching next semester at the IIEP Master of Goethe University:  Conflicts and Economic Development.

It is actually a "block seminar", meaning students pick up a topic and I will be guiding them to write a term-paper during the semester.

This is first time I will be responsible for this course, so I am sharing the syllabus, with the hope to receive some feedback about relevant topics and papers I may be missing. My students (maybe) and I (for sure) will be very grateful to any suggestions!

Topics: Conflicts and Economic Development (MIIEP-Goethe University, Summer 2013)

General readings

Economic determinants of conflicts
Ethnicity, polarization and conflict
Long run effects of conflicts
Effects of conflicts on human capital
Effects of conflicts on health
Violent conflicts and behavior change
International trade and conflicts
Commodity prices and conflict

Thursday, November 29, 2012

The deadliest weapon in Nuba?

I am preparing a first draft of a joint paper with Asha Abdel Rahim and Aleksi Ylonen. We analyze data that Asha collected in the Nuba Mountains in 2008, when a short interwar period made possible for international institutions to enter the area.

Given I am not expert in the topic (but Asha and Aleksi are), I am trying to understand the situation from different sources. Nuba Mountains is one of the most isolated places on earth, so there is no so much information. One of the first things you find on the web is the fantastic photographic work of the (in)famous Leni Riefenstahl.

She was prized worldwide from her work in Nuba, and for some it was her redemption after a dark past... but then I found this really amazing documentary made in 1998 by the Slovenian writer, documentary film maker, human rights activist, journalist and worldwide traveler Tomo Kriznar, some of the few that was in the area during the civil war in the 90s. He went back to the village in which Leni did most of her work, and discovered something really disturbing. There were no more naked painted people dancing and wresting. Instead, they were dressed and reading the Koran... what had happened? One of the elders told him: "IT WAS THE BOOK". Which book??? After Khartoum discovered  Riefenstahl's book , decided to increase the Arabization and military control in Kau village and the rest of Nuba... her camera was a deadly weapon, again.

Monday, November 19, 2012

Enlaces perdidos - mercados inexistentes: El proceso de transformación de las economías rurales

Versión en español del post sobre el estudio en que se resumen más de tres años de investigación acerca de la transformación de las economías primitivas, con datos recolectados en aldeas rurales de Gambia: "Missing links, missing markets: Internal exchanges, reciprocity and external connections in the economic networks of Gambian villages". 

La transición desde actividades económicas primitivas a intercambios más complejos que eventualmente han dado origen a economías de mercado y otros sistemas económicos modernos fue un elemento importante en la estructura de las teorías de los economistas clásicos y una punto fundamental para los primeros sociólogos económicos como Thorsten Veblen, Max Weber y, en particular, Karl Polanyi. En la "Gran transformación" de Polanyi, las economías modernas son formadas a partir de una red de intercambios recíprocos de comunidades aisladas que evoluciona hasta un sistema que institucionaliza el mercado. El concepto de economías primitivas como intercambios recíprocos se basa principalmente en la influyente descripción de Malinowski del sistema de producción de los habitantes de las islas Trobriand, la que a su vez es la base para el análisis de Mauss sobre la economía del don

El proceso de transformación fue formalizado por Rachel Kranton (AER, 1996).  En su modelo los agentes deciden entre intercambios recíprocos (economía del don) con "agentes conocidos" (en términos de preferencias, costos de producción y otras características relevantes), o transacciones  de mercado con agentes anónimos, utilizando dinero como medio de intercambio. Si el costo de búsqueda de agentes anónimos es mayor que el beneficio que se obtiene por la diversificación de consumo que ofrecen los mercados, los agentes preferirán intercambios recíprocos. Uno de los principales resultados es que la reciprocidad se puede mantener incluso si existe la alternativa de usar el mercado para las transacciones.

El objetivo de mi investigación es contribuir al estudio empírico del proceso de transformación de las sociedades rurales tradicionales utilizando una perspectiva de análisis de redes. Para esto he tenido la fortuna de recolectar, luego de de 3 meses de trabajo de campo, una base de datos única sobre redes económicas (tierra, trabajo, insumos y crédito) en 60 comunidades rurales de Gambia. En estas aldeas predominan los intercambios económicos tradicionales no monetarios, pero unos pocos hogares participan en transacciones de mercado (en torno al 10%). El objetivo es explorar si estos hogares tienen un comportamiento distinto dentro de la economía de la comunidad.

La red de intercambios económicos en una de las aldeas 

Como se puede ver en la figura anterior, la mayoría de los hogares de la aldea están conectados con al menos un enlace en uno o más intercambios económicos. Además, muchos de estos intercambios son recíprocos (el enlace es bidireccional). Si el proceso de transformación es cierto, los hogares con conexiones de mercado debiesen tender a abandonar las transacciones dentro de la aldea y en particular aquellas que implican reciprocidad. 

En el análisis empírico los datos son estudiados tanto a nivel de hogares y como nivel de cada enlace (link-level), utilizando técnicas de propensity score matching, MCO en modelos lineales y regresiones diádicas. En todas las especificaciones econométricas se encuentra evidencia para las dos hipótesis principales: (i)  Sustituibilidad entre intercambios internos y externos, es decir, hogares con relaciones económicas externas son menos propensos a estar involucrados en interacciones económicas dentro de la aldea, y (ii) reciprocidad versus mercado, o sea  que hogares con relaciones económicas externas son menos propensos a estar involucrados en intercambios recíprocos con otros miembros de la comunidad.

Si bien los resultados muestran una relación negativa entre participación en transacciones de mercado y en la economía de la aldea, no es evidente interpretar ésta en forma causal. Para eso se deben cumplir ciertos supuestos especificados en el paper. Básicamente, se debe esperar que las características no observables que determinan la creación de enlaces internos afecten a la formación de los enlaces externos en una misma dirección. Yo sostengo que esto es probable, pero un potencial sesgo  en la estimación sigue siendo un problema no resuelto totalmente, y que permanece como desafío a tratar en futuras investigaciones.

Incluso si los resultados se consideran como correlaciones parciales,  se pueden deducir importantes implicaciones para políticas públicas y de cooperación internacional. Muchos programas de desarrollo rural tienen como objetivo aumentar la integración de aldeas remotas al mercado. Estos programas podrían tener efectos secundarios no deseados, tales como la reducción de interacciones dentro de la comunidad y la destrucción del sistema de intercambio en que se basa la economía del don. Por lo tanto, es necesario tener en cuenta la complejidad de los intercambios de éstas comunidades para poder comprender los efectos de intervenciones orientadas al mercado. Por ejemplo, Von Braun and Webb (1989) y  Carney and Watts (1990) han mostrado como en Gambia programas que intentaron aumentar la productividad agrícola y la producción de cultivos comerciales fallaron porque el sistema económico tradicional no fue considerado en el diseño.

Reunión de aldea en que los datos de redes  de intercambio económico fueron recolectados  

Friday, November 9, 2012

NEUDC 2012: Papers I liked

A lot of fun and a lot of jet lag. Short but meaningful. It was a great pleasure to participate for first time at the NEUDC conference last weekend. Many papers blew up my mind and gave me good new ideas, here a selection of my favorites (at least from the presentations I assisted):

"Aid Under Fire: Development Projects and Civil Conflict"
Presented by: Benjamin Crost (University of Colorado Denver).
Main message: Elegible munipalities for a CDD program in Phillipines experienced a large increase in conflict casualties compared. This is likely to be related with rebel groups trying to impede increase for government support in these areas.

"Preferences over leisure and consumption of siblings and intra-household allocation"
Presented by: Martina Kirchberger (University of Oxford)
Main message: While most models consider children as passive agents, they are agents with their own preferences over leisure and consumption. A model of Bart versus Lisa is presented and supported with data from many countries.

"Violence, Emotional Distress, and Induced Changes in Risk Attitudes Among the Displaced Population in Colombia"
Presented by: Andres Moya (UC Davis)
Main message: Data collected in a group of internally displaced rural households and a group of non-displaced rural households in Colombia provide evidence that more severe and more recent episodes of violence and the incidence of anxiety disorders induce higher levels of risk aversion.

"Water Supply and Water handling-Complements or Substitutes"
Presented by: Elena Gross (University of Göttingen)
Main message: Households in rural Benin consider improved water supply and water handling as substitutes of water provision. This implies a neutralization of the effects of public water infrastructure programs, given households reduce water filtration and disinfection.

"The Value of Advice: Evidence from Mobile Phone-Based Agricultural Extension"
Presented by: A. Nilesh Fernando (Harvard University)
Main message: Indian farmers that received the option to get assistance of a call center and receive other information related to agricultural information in their cell phones changed their behavior, adopting more effective and less hazardous pesticides.

"The Effect of Financial Access on Networks: Evidence from a Field Experiment in Nepal"
Presented by: Margherita Comola (Paris School of Economics)
Main message: Using panel data on the network of financial transactions before and after a field experiment in rural Nepal, evidence of the endogeneity of the networks is provided and estimates of the bias of the exogenous assumption are provided.

And finally, two nice RCTs in Chilito:

"Micro Entrepreneurship Training and Assets Transfers: Short Term Impact on the Poor"
Presented by: Claudia Martinez (University of Chile)
Main message: Business training and asset transfers to micro-entrepreneurs increase significantly employment and income in the short term.

"Savings as Insurance: Evidence from a Randomized Field Experiment among Low-Income Micro-Entrepreneurs in Chile"
Presented by: Dina Pomeranz (Harvard University)
Main message: Women micro-entrepreneurs significantly increase savings when a free saving account is offered. Some of the main effects come from helping the women to confront the hostile environment of a banking institution.

Wednesday, October 31, 2012

Missing links, missing markets: The transformation process of rural societies

This weekend I will be going for first time to the NEUDC conference, to be held at Dartmouth College this year. I am excited, because this is probably the biggest development economics conference in the USA and I will have the opportunity to present my favorite outcome from the Gambian networks project, a paper called  "Missing links, missing markets: Internal exchanges, reciprocity and external connections in the economic networks of Gambian villages". 

The transition from primitive economic activities  to more complex exchanges that eventually lead to market economies or alternative modern economic systems was a relevant element in the structure of theories of the classic economic authors and a key issue for the early economic sociologists like Thorsten Veblen, Max Weber and, in particular, Karl Polanyi. In Polanyi's  great transformation, modern societies are shaped in the transition from a network of communitarian reciprocal exchanges  to institutionalized market interactions. The concept of primitive economies as reciprocal exchanges is largely based on Malinowski's  influential description of the production system of the Trobriand islanders, that is also the foundation for Mauss' analysis of a gift economy.

The transformation process  was formalized by Rachel Kranton (AER, 1996). In her model, agents can choose either reciprocal exchanges with other agents whose preferences, production costs and other relevant characteristics are known, or  market transactions with anonymous agents, using money as medium of exchange. If the cost of searching for trading partners is higher than the benefit obtained from consumption diversification offered by markets, then agents will prefer reciprocal exchanges. One of her main results is that reciprocity can be enforced even if markets exist as an alternative for transactions. 

The aim of the paper is to contribute to the empirical analysis of the process of transformation in traditional rural societies using a network perspective. A unique database on economic networks (land, labor, inputs and credit) collected in 60 villages of rural Gambia, where traditional non-monetary economic exchanges -gift economy- prevail, is used to study  the  behavior of  households involved in market transactions. 

The network of economic exchanges in one village

As can be seen in the figure above, most of the households in the village are connected with a link in one or more economic exchanges. And many of these exchanges are reciprocated (the link is bidirectional). If the transformation process is true, household with connections to the market will tend to abandon transactions inside the village and particularly those that imply reciprocation. Given the Gambian network data have information regarding the existence of links external to the village in each of the networks, I can compare if households with links to the market (that are very few, around 10% of all the households in the village) behave differently. 

The empirical analysis is conducted at both household- and link-level, using propensity score matching techniques, OLS linear models and dyadic regressions.  In all the econometric specifications I find support for the two main hypotheses: (i) Substitutability between internal and external exchanges, i.e. households with external economic links are less likely to be involved in economic interactions within the village; and (ii) Reciprocation  versus market, i.e. households with external economic links are less likely to be involved in reciprocated exchanges with fellow villagers.

In the paper I discuss the assumptions required for a causal interpretation of the results, basically that unobservable characteristics determining the creation of internal links affect the the formation of external links in the same direction. I argue that this is plausible, but the potential bias remains as a not fully solved issue to be addressed in future research. 

Even if taken as partial correlations, there are relevant policy implications related to the findings. Rural development programs that aim to increase market integration of isolated villages can have undesired effects, such as the reduction in community interactions and destruction of the gift exchanging system. Therefore, it is necessary to consider the complexities of community exchanges in order to understand the effects of market-oriented interventions. For instance, Von Braun and Webb (1989) and Carney and Watts (1990) have shown how in The Gambia programs that attempted to increase agricultural productivity and cash crops production failed because the traditional economic system was not considered in the design. 

Village gathering where the data about network of economic exchanges was collected