Gaap.ifpri.info

Case Study 2. International Food Policy Research Institute (IFPRI): Evaluating the long-
term impact of anti-poverty interventions in Bangladesh


Countries: Bangladesh

Year(s) of project/ study: 1994-present

Contact: Agnes Quisumbin; Neha Kumar
Background: While many evaluations have attempted to assess the short-term impacts of poverty reduction
programs, relatively little is known about their long-term impact. To address this gap in knowledge, IFPRI,
together with Data Analysis and Technical Assistance (DATA), Ltd. and the Chronic Poverty Research Centre
(CPRC), collected gender-disaggregated assets data spanning over 15 years (1994- 2010) and assessed the long-
term impact of three anti-poverty interventions in Bangladesh: i) the introduction of new agricultural technolo-
gies, ii) educational transfers, and iii) microfinance – on a range of monetary and non-monetary measures of
well-being (Quisumbing, Baulch and Kumar, 2011). The impact evaluation of the introduction and dissemination
of vegetable and fish technologies in Bangladesh builds on an existing IFPRI data set, collected in 1996-97, with
detailed gender-disaggregated assets data, which made it possible to estimate the impacts of technology dissemi-
nation on men’s and women’s assets (Quisumbing and Kumar 2011; Kumar and Quisumbing 2011).
Methodology: These impact evaluation studies drew from the IFPRI Chronic Poverty and Long Term Impact
Study in Bangladesh dataset, which used integrated and iterative qualitative and quantitative methods. The
study builds on three surveys conducted by IFPRI in Bangladesh to evaluate the short-term impacts of micro-
finance (1994), the new agricultural technologies (1996-97) and the introduction of educational transfers (2000
and 2003) and a follow up conducted in 2006-07. While information on many gender-disaggregated variables
was collected in all the evaluation studies, gender-disaggregated assets data was collected only in the agricultural
technology sites. In 2006, IFPRI, DATA and the CPRC began a major study to resurvey the households surveyed
in all three of the evaluations. While the focus of this study was on understanding of the drivers and maintainers
of chronic poverty in rural Bangladesh, the intervention-comparison groups were maintained from the previous
study, and greater attention was placed on obtaining gender-disaggregated data in all the sites. The resurvey
involved both qualitative studies and a follow-up longitudinal survey of households included in the IFPRI studies,
and involves three sequenced and integrated phases.
Another round of data collection in the educational transfers and agricultural technology sites was undertaken in
2010, focusing on the impacts of the food price increases in 2007-2008. A gender-disaggregated assets module
was administered to all surveyed households, focusing on gendered responses to the food price crisis.

Findings: These studies in Bangladesh indicate that household-level and individual impacts of anti-poverty
interventions differ in the short term and the long term because of differences in the time path of net benefits
from the interventions and spillover effects. Divergence between short-term and long-term impacts may be
especially important in interventions that seek to bring about behavioral change, where spillover effects and
learning from others may be significant.
In the case of improved vegetable and fish technologies, Kumar and Quisumbing (2011) found that long-term
impacts on household-level consumption expenditures and asset accumulation were insignificant in a site where
improved vegetables were targeted to women’s groups for cultivation in their own homesteads, but positive and
significant in the site where polyculture fishpond technologies were targeted to households, with minimal consid-
eration of gender dynamics. However, the impacts on individual nutrient intake, nutrient adequacy, and nutri-
tional status do not follow the pattern of household-level impacts. For example, despite the minimal monetary
gains, early adopters of improved vegetables, particularly women and children, achieved sustained improvements
in nutritional status.
Quisumbing and Kumar (2011) found additionally that women’s assets increase more relative to men’s when
technologies are disseminated through women’s groups, suggesting that implementation modalities are im-
portant in determining the gendered impact of new technologies. Results also suggest that social capital, when
embodied through women’s groups, not only serves as a substitute for physical assets in the short run, but helps
to build up women’s asset portfolios in the long run.
For more information:
Kumar, Neha, and Agnes R.Quisumbing. 2011. Access, adoption, and diffusion: understanding the long-
term impacts of improved vegetable and fish technologies in Bangladesh. Journal of Development
Effectiveness 3(2): 193-219. Available at:
Quisumbing, Agnes.R., Bob Baulch, and Neha Kumar. 2011. Evaluating the long-term impact of anti-
poverty interventions in Bangladesh: an overview. Journal of Development Effectiveness 3(2): 153-174.
(also IFPRI DP 995). Available at:
Quisumbing, Agnes R., and Neha Kumar. 2011. Does social capital build women’s assets? The long-term
impacts of group-based and individual dissemination of agricultural technology in Bangladesh.
Journal of Development Effectiveness 3(2): 220-242. (also CAPRi WP97). Available at:

Quisumbing, Agnes R. 2011. Do men and women accumulate assets in different ways? Evidence from
Bangladesh. IFPRI Discussion Paper 01096. Available at:

Quisumbing, Agnes R., Neha Kumar, and Julia A. Behrman. 2011. Do shocks affect men’s and women’s
assets differently? A review of literature and new evidence from Bangladesh and Uganda. IFPRI
Discussion Paper 01113. Available at:
The IFPRI Chronic Poverty and Long Term Impact Study in Bangladesh dataset and other related
research papers are available at:

Feedback on case study 2 methodology based on an interview with Agnes Quisumbing and Neha
Kumar:
1. What are the unique gender-asset questions/indicators you collected in your survey instru-
ment that were particularly valuable or reflective of methodologies you would like to see rep-
licated in future work and why?
The gender-disaggregated assets module builds on an existing data set (see case study 1) for the agricultural technology sites, but is now administered to all surveyed households. The major innovation is the collection of gender-disaggregated assets data over time, which allows analysis of gendered patterns of asset accumulation. In the agricultural technology panel, we have observations in 1996/97, 2006/7, and 2010. In the educational trans-
fers sites, we have observations in 2006/7 and 2010. New data collection efforts may want to be forward-looking
in terms of creating the possibility of revisiting households to build up panel data sets on individual and joint
asset accumulation. So this means obtaining information with which to track households and individuals over
time. We also updated the community questionnaire to capture changes in local facilities, institutions, and even
cultural norms (for example, the extent to which women can travel—whether limited to the village, the town
center, etc—has expanded greatly over time, partly because of the need to go outside of the village for NGO
training).
2. What are the unique gender-asset questions/indicators you either collected in your survey
instrument that you would have implemented differently or you were not able to collect, but
which you would have liked to collect and why?
 Perceptions of what men’s and women’s “ownership” of assets really means, what “jointness” really means (respondents did identify most of their assets as joint assets, although they also identified individually-owned assets);  Collect gender-disaggregated shocks data. Subsequent analysis shows that shocks affect men and women differently, but it would have been good to investigate whether illness (for example) of a man or a woman had different effects on households;  Collect better indicators of social capital and group dynamics. We have individual information on group membership and types of groups (from the 2006/7 survey), but not information on the groups themselves, and;  Do qualitative work, and then build quantitative modules, to examine portfolio substitutions (for example, when having one asset helps to acquire another one) and discern whether new types of assets (or uses of assets) have emerged.
3. Asset-gender dynamics are heterogeneous, complex and rooted in social, economic and insti-
tutional factors—are there any background factors that relate strongly to gender-asset dynam-
ics that you either collected or wish you had collected?

Since we and our local collaborators, DATA have been working in these communities for a long time (more than
10 years), we have a good grasp of local conditions.
4. Are there any particularities about the region or country of implementation which you think
are important to recognize in relation to the gender-asset indicators you collected which are
important for other researchers to be aware of? Did any of these context- or country-specific
factors influence your survey implementation methodology, and how?

We continued to follow DATA’s field protocols in Bangladesh, which is to field a team of both a male and a female
enumerator. The male interviews the husband, while the female enumerator interviews the wife. They typically
field two male and two female enumerators in an area for ease of travel, particularly safety, and accommodation.
5. What do you see as the largest methodological challenges in collecting gender-asset data in
general and how can we as a research community work towards filling this gap?
A big challenge continues to be making sure that gender disaggregated data is collected at baseline. Going forward, we need to be able to keep up with new categories of assets that emerge (for example, term insurance, new savings instruments, etc.) as well as new uses for incomes earned from assets. We also need to be able to capture changes in ownership and control of assets over time, especially as the relative value of assets change (land may become less important as incomes become more diversified, for example).

Source: http://gaap.ifpri.info/files/2011/12/Case_Study-2.pdf