93916 from EVIDENCE to POLICY Learning what works for better programs and policies January 2015 BANGLADESH: Can Conditional Cash Transfers Improve Nutrition? Conditional cash transfer programs have proven successful at to maximize the impact and understanding whether electronic CONDITIONAL CASH TRANSFERS encouraging poor families to use health services and send chil- payments can reduce administrative costs, improve transpar- dren to school. Across the world, these programs are helping ency and give women more say over family finances. break the cycle of pov- Policymakers and development groups are interested in erty by boosting edu- innovative ways to deliver cash transfer programs and make cation and improving an impact where it’s needed most. In Bangladesh, World health for the most vul- Bank researchers worked with the government to evaluate a nerable. Less is known pilot social protection program that conditioned transfers on about how to best primary school attendance, monthly growth monitoring for structure cash transfers young children and monthly nutrition sessions for mothers. to improve nutrition Payments were made to mothers via electronic cash cards. in young children, and While the program had no effect on school attendance, it sig- some programs are now including regular meetings for moth- nificantly boosted consumption of protein and other nutri- ers to boost their knowledge of nutritious foods and the impor- tion-rich foods, and reduced wasting. Based on the results, tance. As countries work to develop sustainable and effective the nutrition-related part of the program is being scaled social protection programs, understanding the possibilities of up with World Bank assistance to reach 600,000 of the conditional cash programs will help create safety nets that de- poorest households across Bangladesh, potentially ben- liver. Part of this is learning how to structure these programs efitting 2.7 million people. Context Bangladesh reduced the percentage of families living in pov- ings on nutrition. Poor mothers with children aged six to 15 erty by almost a third over the past decade to 30 percent, but got the same amount if the children attended at least 80 per- the country hasn’t been as successful at cutting the rate of cent of primary school classes. The payment was set regardless wasting and stunting among children. Bangladesh has one of of the number of children and the maximum a family could the highest rates of malnutrition in the world and 41 percent receive was 800 Taka a month. The program ran from April of children below the age of 5 are stunted. About five million 2012 through December 2013, with payments made every two children are also out of school. months. In January 2012, the Government began piloting a new The program relied on electronic transfers rather than conditional cash transfer program called Shombhob, which handing out cash directly. Electronic payments not only help means “possible” in Bangla. The pilot took place in Jaldhaka bring people into the financial system, but they’re a good way and Hatibandha, two rural areas in the Rangpur Division, the to track that money reaches the intended beneficiary and they country’s poorest region, and in the slums of Narayanganj City. have the added plus of giving women more control over financ- Poor mothers with children up to three years old received es. Mothers received cash cards to access their account, which 400 Taka (about $5) a month for taking their children for was held by the Bangladesh Post Office. They could either trav- monthly growth monitoring and for attending monthly meet- el to the nearest post office, or they could wait and withdraw money from mobile Point-of-Sales machines brought in on proxy means tests score, compiled through a survey of their designated days. The program had to pay a commission (to the assets, access to facilities, and household demographics, post office) when money was withdrawn—either one percent among other things. The list of selected families was then vet- of the transferred amount, or 10 Taka, whichever was more. ted by community leaders, who also played a role in explain- More than 37,000 families applied for the program and ing the program to the targeted families. The final number of 15,952 of the poorest families were selected based on their enrolled beneficiary households was 14,125. Evaluation CONDITIONAL CASH TRANSFERS The evaluation was carried out in two of the three pilot ar- with both. Researchers used regression discontinuity design eas, Jaldhaka and Narayanganj. First, a census was carried combined with the difference-in-difference methodology to out in these two areas of all 44,590 household. The census compare outcomes between eligible and participating families questions were designed to measure poverty using the proxy just below the selection cutoff with outcomes from those just means test. Households with a poverty score above the 50th above the cutoff, who were almost eligible but did not par- percentile were excluded as they were unlikely to be benefi- ticipate in the program. The same households were included ciaries of the cash transfer program. Those scoring below the in the follow-up endline survey in May-June 2013, about 13 25th percentile were viewed as potential beneficiaries, and months after their first Shombhob payment. A total of 2,718 those between the 25th and 50th percentile were marked as households were in the analysis. Results from Narayanganj, potential controls. A random sample of 3,000 households which had high migration typical of urban slums, were ex- was then interviewed to ensure representation of three groups cluded from the impact evaluation because researchers had a qualifying for cash transfers: those with children under the difficult time tracking down nearly 30 percent of the house- age of three, those with primary school age children and those holds in the baseline survey. Results The program succeeded in improving nutrition for which reflects chronic malnutrition and is based on a child’s young children, with cases of wasting declining by height-for-age. The lack of impact on stunting may reflect 40 percent. the need for sustained, longer term programs to halt chronic malnutrition. For children aged 10 to 22 months at the time of enrollment in the program, wasting declined by 12.5 percentage points. Overall, families who qualified for the cash transfers This translates into a 40 percent decline in the number of spent more on healthy, high-protein food, including children who were wasted, which uses weight-for-height to eggs, meat, dairy, fish and beans, compared with measure nutritional status. There wasn’t a similar impact for similar families who weren’t in the program. children aged 22 to 46 months at enrollment, which may be because nutrition programs are most effective before a The transfer helped families to increase their spending by child turns age two. Nor was there any decline in stunting, about 70 percent of the money they received, and close to This policy note is based on “Can Conditional Cash Transfers Improve Education and Nutrition outcomes for Poor Children in Bangladesh? Evidence from a Pilot Project,” Celine Ferre, Iffath Sharif, June 1, 2014 https://openknowledge.worldbank.com/handle/10986/20511 90 percent of that was spent on food. More importantly, The flexible mobile electronic payment system they bought more nutritious and protein-rich food. Moth- used by the program was efficient and transparent, ers in the nutrition group spent 174 Taka more a month reducing travel time for poor mothers and making on protein-rich food—out of a total average transfer of 542 it easy for them to claim their money. Taka a month—than mothers from similar families who did not receive these program benefits. Spending on protein- In total, Shombhob delivered some US $1.78 million dol- rich food was even higher among mothers who went for lars in cash transfers to more than 14,000 families over the monthly nutrition meetings by about 56 Taka, compared 13 months period, including administrative costs. The ad- with mothers who received cash transfers if their children ministrative cost for delivering the program worked out to attended primary school. be about US $1.40 a month per child, which is considered reasonably cost effective for a conditional cash transfer pro- The program also led to an increase in knowledge gram in largely rural areas. about the importance of breastfeeding. In order to give mothers a better understanding of child- hood nutrition and health, women with young children were required to attend monthly nutrition counseling ses- sions on how to care for and feed children. They also had to bring their young children for monthly growth monitor- ing sessions, during which the children were weighed and measured and their growth was documented on growth charts that mothers took home. This led to an eight percent increase in the number of women—from 83.1 percent to 89.4 percent—who knew that babies should be exclusively breastfed until six months of age. The program, however, didn’t improve school attendance or enrollment. Because the program started after the new school year had begun, researchers knew it would be too late to boost en- rollment, but they hoped for an improvement in atten- dance. That didn’t happen, which is inconsistent with results from similar cash transfer programs around the The program targeted the right people and reached world. Researchers pointed to a few possible reasons for the area’s most vulnerable families. the lack of impact. Children were supposed to attend pri- mary school at least 80 percent of the time, but program The surveys found that recipients were actually poorer than administrators had a problem collecting the information non-recipients, and that those who applied and were accept- because of the large number of schools children attended. ed into the program were poorer than those rejected, based Also, transfers were made regardless of attendance dur- on their proxy means test scores. Nevertheless, between ing the flooding of September-October 2012. This could baseline and endline, researchers were unable to track many have confused parents about the need to send children to families—particularly the most vulnerable slum dwellers school to receive the payments. Another possibility is that among the very poor—highlighting the difficulty that even the money wasn’t enough to promote change, given that well-administered programs can have when poor families there’s already an existing national cash transfer program don’t have stability to stay in one place. linked to school enrollment and attendance. Conclusion Shombhob improved the welfare of extremely poor house- it hard for them to meet requirements for payments, be holds and boosted nutrition outcomes for young chil- tracked to qualify for payments or even receive payments. dren—a significant accomplishment that highlights the As researchers continue to explore ways to help the poor, potential for programs like Shombhob to meet the nutri- the results of this evaluation will guide the creation of new tional needs of poor children during a critical development programs that give families the tools they need to help window. As policy makers in Bangladesh and elsewhere their children grow, at a time when their brains and bodies look to expand conditional cash transfer programs, the need it most. CONDITIONAL CASH TRANSFERS strong nutritional outcomes from this program will surely provide a guide for how to boost nutrition, particularly In Jaldhaka… during a child’s first 1,000 days. • Average household size is 4.5 people Nevertheless, the program did face operational chal- • 40 percent of household heads are casual agricultural laborers lenges—especially in the education component—that with no education need to be considered when creating new programs and • Half of total expenditures are on food revising existing ones to help society’s most vulnerable. • Average size of land owned is 1.8 acres • Few households have durable assets like televisions or fans… and One question underlined by this evaluation is how to reach in the rural areas, most don’t have access to electricity. the very poor families that move between slums, making The Strategic Impact Evaluation Fund, part of the World Bank Group, supports and disseminates research evaluating the impact of development projects to help alleviate poverty. The goal is to collect and build empirical evidence that can help governments and development organizations design and implement the most appropriate and effective policies for better educational, health and job opportunities for people in developing countries. For more information about who we are and what we do, go to: http://www.worldbank.org/sief. The Evidence to Policy note series is produced by SIEF with generous support from the British government’s Department for International Development. THE WORLD BANK, STRATEGIC IMPACT EVALUATION FUND 1818 H STREET, NW WASHINGTON, DC 20433 Produced by the Strategic Impact Evaluation Fund Series Editor: Aliza Marcus; Writer: Daphna Berman The original had problem with text extraction. pdftotext Unable to extract text.