Results from Trickle Up Client Monitoring System January through December 2009

Transcription

1 Results from Trickle Up Client Monitoring System January through December 29 Karishma Huda and Sheila Chanani Introduction The traditional response to alleviating poverty for the extreme poor has been food aid. Food aid, however, is a temporary crisis management instrument that does not create a sustainable platform for change. Stand alone promotional programs such as microfinance also do not meet the varied needs of the extreme poor. In response, the Consultative Group for Assisting the Poor (CGAP) and the Ford Foundation initiated a multi-pronged livelihoods protection and promotion scheme to financially and socially uplift the extreme poor. Adapted from BRAC s Targeting Ultra Poor (TUP) program in Bangladesh, this graduation model is designed to provide extremely poor households with both promotional elements that uplift them to a higher socio-economic space, as well as protective elements that provide them with basic social safety nets when faced with crises and shocks. Promotional elements include assets for entrepreneurial use, enterprise training, and basic financial services (usually in the form of savings; the protective elements include basic health services, consumption support, and individualized hand-holding support from a designated field agent. Upon receiving 24 months of holistic support, the model posits that members will have graduated into a sustainable livelihood and out of extreme poverty. 2 nd Round: completed July 29 for period April-July, 29 3 rd Round: completed January 21 for period August to December 29 1

2 Nine pilots have been implemented in seven countries to test the feasibility of the model in diverse economic settings. The pilots are all being carefully monitored and evaluated; most are accompanied by rigorous randomized control trials as well as qualitative research. This allows for close monitoring of changes in the lives of the beneficiaries as a result of the program. While the evaluations and qualitative research provide insights at different points of the program cycle, it became clear that a monitoring tool to track the regular performance of the program inputs was essential. Program staff need to know how their program is evolving, and the kind of impact they are making on a constant basis, not simply at the baseline and end line. The Client Monitoring System (CMS), therefore, is a set of simple indicators to regularly track changes in clients social and economic conditions. These indicators also help to understand how program beneficiaries are progressing in comparison to members of other pilots. Background on Trickle Up Trickle Up s TUP program is one of the nine global pilots, and is being implemented in South 24 Parganas, Canning, in West Bengal, India. It consists of three main components: Economic Component: this includes the transfer of an entrepreneurial asset, animal sheds, veterinary support, a food support allowance, a savings and credit component within the SHG framework, and asset management training to transfer basic livelihood skills to members Social Development Component: intends to build social safety nets through awareness training and confidence building in weekly SHG meetings and individual home visits, includes fruit tree distribution for food security, and a Village Assistance Committee for access to vertical social networks Health Component: encourages members to seek free treatment through government health services, provides health/nutritional consultations through staff health worker, distributes sanitary latrines for preventative health Critical to Trickle Up s approach is working through local partner organizations to implement their programs on the ground. Human Development Cooperation (HDC) is Trickle Up s local partner implementing the TUP program. This package of inputs was extended to 3 women over a period of 27 months (May 27 August 29) with the intent of graduating them out of extreme poverty and into a sustainable livelihood. In August 29, Trickle Up graduated 86% of their beneficiaries based on them being financially active (having adequate savings and taking SHG loans), diversifying income sources, having awareness of government services, health and sanitation. Trickle Up is now scaling up to 3 additional beneficiaries in Canning, and over 1, in 4 states in India. Methodology 2 nd Round: completed July 29 for period April-July, 29 3 rd Round: completed January 21 for period August to December 29 2

3 The CMS is an Excel-based format that covers basic indicators for individual clients and their assets, savings behavior, stipends, other sources of income, illnesses, and children s vaccinations. The CMS tracks a few other social changes related to clients ability to participate in community activities, purchase more clothes and gifts, and educate their children. A subset of 2 clients was randomly selected to track their food consumption during each round of data collection. Additionally, each section includes space for field agents to include their own qualitative insights on these various indicators. Currently 3 rounds of data collection have been completed (April 29, July 29, and January 21, with one more round to be completed in June 21). The April data covers the period of January through March, the July data covers the period from April through early July, and the January data from August to December. While the April and July rounds give us an idea of programmatic results, the January and upcoming June rounds will shed light upon the sustainability of these results since the program ended. Field agents were the data collectors for the beneficiaries for which they are programmatically responsible. They were given one half-day training on data collection. TU decided to give the agents ten days to collect the information and complete all CMS forms. Each field assistant collected data for approximately 6 members. In some cases, data collection was combined with other work e.g., data collected during self-help group (SHG) meetings, but in others, no activities were planned for that week and field agents went to the village specifically to collect the data. Single entry of the data has been completed by IQ System, a survey and research unit based in Kolkata, India. Software for the data entry is a web-based platform developed by Intellecap, based in Mumbai, India. There were some limitations to implementing the CMS which should be considered when reviewing this report. The CMS was rolled out for the first time with Trickle Up, meaning there were implementation challenges both in data collection and data entry. Field staff experienced a learning curve in completing the forms; staff were new to supervising the data collection process; data entry operators faced challenges in using a newly developed system. There could also be biases in responses, given that the field staff were the data collectors (e.g. beneficiaries withholding information in anticipation of receiving more benefits), but this was also offset by the staff s extensive knowledge of members and their household situations. Round 3 of the data collection was completed while HDC field workers were learning about Trickle Up s new monitoring and evaluation system, which resulted in some gaps in data with the CMS system. Overall, these challenges led to delays in final output, but compromises have not been made to the integrity of the data or the report. Results The graphs below outline the findings from the first three rounds of data collection. 2 nd Round: completed July 29 for period April-July, 29 3 rd Round: completed January 21 for period August to December 29 3

4 Graph 1: Quantity of Primary Asset Holdings Number of Asset 1 8 Average Quantity of Primary Asset Holdings January-December 29 Provided by Program Apr-9 Jul-9 Jan Goat Sheep As Graph 1 shows, the primary assets that were provided to members were goats and sheep. The graph below (Graph 2) illustrates total asset holdings, and displays the diversity in asset choices and livelihoods of members 1. There was a great deal of distress selling of sheep and goats in the first round, hence the drop in assets in April 29. Members were preparing for the agricultural season and sold off some of their livestock in order to pay for agricultural inputs. April also marked the end of winter. Due to the sudden change in the climate and the scarcity of fodder, the livestock got sick. While many died, some members decided to sell their livestock when they got sick rather than pay for medicine. During the second round (July 9), members were more knowledgeable of asset management, explaining the lower mortality rate. In July, there was also less distress selling since many members had worked as day laborers in government programs to buffer their incomes. In the last round (Jan 1), diseases from the monsoon, followed by a particularly harsh winter, led to more livestock mortality and distress selling among goats and sheep. As the program also ended in August 9, members stopped receiving veterinary care (de-worming and vaccinations) which exacerbated this decline. Although Trickle Up tried to set up government linkages so members can continue to treat their livestock, members typically did not utilize this service due to the distance of the facilities and transportation costs incurred. 1 Although one critical goal of the project was to see members diversify their assets into a variety of livelihoods, for the purposes of discussion, we are focusing upon the primary assets: sheep and goats. This is for the following reasons: the majority of participants received these two assets from the project; due to avian flu, ducks were culled in the beginning of the program and not replaced, thus having little impact; very few members chose petty trade as an option ; there was unreliable data on the quantities of fish. 2 nd Round: completed July 29 for period April-July, 29 3 rd Round: completed January 21 for period August to December 29 4

5 It s also important to note that participants were encouraged over the course of the project to diversify their asset holdings, which explains some of the decline in sheep and goats. Graph 2: Quantity of all asset holdings Clients Number of Clients Holding Primary Asset January - December Apr-9 Jul-9 Jan Goat Sheep Fish Duck Pig Trade Paddy Husking Agriculture Graph 3: Members with multiple assets Clients Number of Clients with Multiple Assets Apr-9 Jul-9 Jan-1 Graphs 2 and 3 above show that over time members have increasingly diversified into managing multiple income-generating activities. The program predominantly provided members with goat, sheep at the program s inception. Over time, in order to ensure that everyone had 6, Rs worth of diverse assets, the program provided members with fish (to those who already had excavated ponds), trade (small businesses such as selling bangles, vegetables, etc), and inputs for paddy 2 nd Round: completed July 29 for period April-July, 29 3 rd Round: completed January 21 for period August to December 29 5

6 husking (this includes 2 sacks of paddy, and pots to boil the paddy; members had to secure their own husker.) Over time, members also diversified on their own into rearing pigs (this is particular to the tribal communities and has proved to be very lucrative); some purchased ducks again after the initial government culling; and a few were able to secure land leases for agriculture (monsoon and winter cultivations correspond to the second and third data collection rounds.) An important graduation criterion was having diverse asset choices, so the program heavily encouraged this through social messaging. Importantly, end-of-project data shows that 82% of participants (246 members) diversified into one or more livelihood activity beyond the original activity supported through the project. This is not reflected in the CMS data because not all diversification, particularly that related to agricultural production, was captured. Graph 4: Birth increases for primary assets Total Birth Increases for Primary Assets January-December 29 Apr-9 Jul-9 Jan Goat (No.) Sheep (No.) Fish (kg) Duck (No.) As explained above, asset mortality and distress selling caused relatively fewer assets to be born. When members were selling the goats and sheep, many were selling the mothers, causing a low birth increase across all livestock. Between July 9 and Jan 1, we would expect to see a birthrate of over 35 goats (given the gestation period of goats), and birthrates have clearly fallen far below that. Fish, however, have been thriving over the program period. Fishlings were distributed mainly in July 9, and have begun to reproduce in the last round of data collection 2. So few members retained ducks after the bird flu epidemic that the birthrates are negligible. Graph 5: Client Sources of Income 2 Fish quantities are calculated in kilograms, while goats, sheep and duck are counted individually. It is difficult to weigh the fish at different points, as they are cultivating in the fish pond. While the weight may not be entirely accurate, it is useful to look at the estimated increase between rounds 2 and 3. 2 nd Round: completed July 29 for period April-July, 29 3 rd Round: completed January 21 for period August to December 29 6

7 Clients & Hhold Members Clients Sources of Income January-December 29 Apr-9 Jul-9 Jan Day Labour Agriculture Fish Caught Catching Crabs Fish Sold 6 Jari Work Graph 5 highlights the most frequent sources of income to the household, other than those listed in Graph 4 3. Insufficient data was collected in Round 3 to fully understand the relative importance of household sources of income, including those associated with program supplied assets and the diversification thereof. This is because recurrent income from trading and agriculture in particular was not adequately captured, and because comparing income from the sale of assets to other sorts of income generating activities, such as wage labor, is difficult due to lack of data on the number of weeks worked during the reporting period. These limitations in monitoring changes in income and income sources are expected to be addressed in the July 21 data collection. Nevertheless, it is believed that the original assets provided by the program did not constitute a major source of income. However, other activities such as agriculture and pisciculture that were diversified into using profits and SHG loans are believed by the program to have provided a fairly significant income for many participants. However, not all such income was recorded. Activities for all household members were recorded, which explains why the numbers in the first column add up to more than 3. During the second round, there were fewer alternative activities that members could engage in because of seasonal limitations. Day labor refers to agricultural day labor, but can also refer to earth work (digging ponds, building roads, etc), and can include workers who participate in the National Rural Employment Guarantee Scheme. Agricultural work refers to cultivating one s own land. While fish caught refers to members catching the fish and selling it at the market, fish sold refers to members buying fish at the market and then selling it. Jari work is a form of embroidery that the women do at home and sell to agents. Members are likely to do Jari work when they are not engaged in agriculture or other paid work. 3 It should be noted that this does not include in-kind income. Many members diversified into land and fish cultivation for increased consumption, and only sold the excess. 2 nd Round: completed July 29 for period April-July, 29 3 rd Round: completed January 21 for period August to December 29 7

8 What is clear from this graph is that the primary assets given to members (sheep and goats) were not the main sources of income for beneficiaries. This is partly due to high levels of unanticipated asset mortality, and the fact that income from these primary assets are meant to supplement members existing sources of income. Sheep and goats cannot produce a daily source of income for members rather they were intended to be a medium-term source of income, and their pre-existing activities a daily source of income. However, Trickle Up has learnt from this, and in scale-up are distributing a more diverse basket of livelihood inputs. Most families engaged in day labor at the beginning of the program as their daily source of income. The second round of data corresponds with dry season, where there is little day labor work available due to lack of irrigation water. This is generally the lean season, when members rely on credit from shops and stored harvest to survive. The third round of data corresponded with monsoons (July Oct), and families again engaged in day labor. Caring for assets during this period did not deter from this, as someone in the household engages in daily labor at this time since the wages are the highest. Also, the assets did not generate enough income to substitute earnings from daily wage labor. Cultivating agricultural land is a much desired economic activity. Crop yields result in household food security as well as income, and people prefer working on their own land to working on others land (as with daily labour). However, it is highly dependent on land availability, and access to cultivable land remains a challenge for this population. A handful of participants were able to lease land from local landlords on a seasonal basis at the beginning of the program (this is also dependent upon having strong social networks with landlords). Trickle Up also strongly encouraged members who were able to lease land, and taught them the System of Rice Intensification (SRI) method for higher yield. Programmatic encouragement contributed to greater agricultural investments in the first round. Given that the second round corresponded with dry season, few continued with this as they could not afford irrigation inputs. Very few members engaged in crab-catching and jari work to begin with. Jari work is lucrative, but it is a vocational skill that requires training and existing relationships with traders. Crab catching is specific to a certain tribal group. It is a perilous activity and not very lucrative thus, as other income generating sources developed through the program, members have developed an aversion to this and stopped. Fish catching and selling is seasonal, with the season being from July to December. Thus we see an increase in this activity in the last round. Graph 6: Pregnant members allowances 2 nd Round: completed July 29 for period April-July, 29 3 rd Round: completed January 21 for period August to December 29 8

9 No. of Clients Subsistence Allowance for Pregnant Women and Allowance Expenditures Towards Food January-December 29 Food Other Expenditures Apr-9 Jul-9 Jan-1 22 Stipends are typically Rs. 91. per week. Trickle Up was issuing weekly stipends that terminated for most members by April 9. Since then, the stipend continued only for pregnant women and included both a combination of in-kind food and cash 4. As the graph above illustrates, the majority of this stipend was used as it was intended to boost food consumption of pregnant women. This leads into the next graph, which illustrates food consumption patterns over the data collection period. Graph 7: Food consumption patterns 4 This includes some food supplements and cash to make up 91 Rs per week for pregnant women. This scheme continued until March 29 2 nd Round: completed July 29 for period April-July, 29 3 rd Round: completed January 21 for period August to December 29 9

10 Average Number of Days Eaten From Food Groups Over Three Day Period January-December 29 Days Apr-9 Jul-9 Jan Consumption of fruits/vegetables steadily increased, which coincides with the program s objective of improving food security for members. The program provided nutritional social messaging (staff encouraged members to consume every color of the Indian flag in their diet) and to grow kitchen gardens/plant fruit trees. Inflation has influenced the consumption of different foods. While it has caused fruits and vegetables to become a more affordable option, it has also contributed to the decline in fish/meat/poultry consumption in the last round. The price of fish, for instance, has increased from 6 Rs/kg to 9Rs/kg, meaning only those who own ponds and are doing fish cultivation can afford to consume it. As a result of inflation, members began to replace meat with lentils for protein, increasing the consumption of lentils/legumes. Starches such as rice, potatoes, and roti (bread) remain staples in the Indian diet, and thus have remained fairly constant. In the second round, members generally ate the best across all food groups. This seems contradictory, as this coincides with the lean season. It could be explained by the fact that this period also coincides with the Hindu marriage season, and a time when members attend more social events. Members may simply have. The improvement in consumption of meat/fish/poultry, in this case, was not necessarily due to an increase in income, and thus does not represent a sustainable change. Graph 8: Formal savings 2 nd Round: completed July 29 for period April-July, 29 3 rd Round: completed January 21 for period August to December 29 1

11 Clients Formal Savings: Number of Clients and Average Amount January-December 29 Clients with Formal Savings Average Amount of Formal Savings (Rs.) Rupees Apr-9 Jul-9 Jan-1 Formal savings refer to savings in bank accounts, post offices and government SHGs. The program has not closely followed savings patterns in non-program accounts, but staff believe that the number of savers have been decreasing over time because people who formerly saved in government sources are now saving in the Trickle Up SHG. Although the number of savers has decreased over time, the average savings amount has increased. A possible explanation could be that members with high household disposable income continued to save in both sources. As the program ended in August 21 and a few SHGs failed to operate successfully post-program, perhaps these 4 members began to save more in private savings sources. Graph 9: Informal Savings 2 nd Round: completed July 29 for period April-July, 29 3 rd Round: completed January 21 for period August to December 29 11

12 Clients Informal Savings: Number of Clients and Average Amount January-December Apr-9 Jul-9 Jan-1 Clients with Informal Savings Average Amount of Informal Savings (Rs.) 97 Rupees Informal savings refer to Trickle Up SHGs. Most members continued to save throughout and post the program cycle. As noted earlier, there was a lot of distress selling of livestock during the first round (April 9). While this was partly due to the need for cash in order to invest in agricultural inputs, it was also a result of high livestock mortality in February/March. When assets got sick, many members sold some of the livestock rather than seeking treatment. Some of the money from those sales was saved so that they could purchase new assets after the monsoon when livestock mortality is lower. This has been an important lesson learnt by members through the course of the program: to sell off sick livestock before they die, save the money, and reinvest back into livestock once the turbulent monsoons have passed. The gradual increase in savings is also explained by the fact that Trickle Up provided a lot of training on the importance of savings during the second round. During the weekly household visits and SHG meetings, the social mobilizers encouraged members to save regularly during the weekly household visits and centre meetings. Graph 1: Number of withdrawals from formal savings 5 5 Staff also did not collect data for formal withdrawals in the second round, which is why July 9 is blank. Also, the SGSY is structured so that members cannot withdraw their savings. They must take a loan against cumulative group savings and repay that amount. 2 nd Round: completed July 29 for period April-July, 29 3 rd Round: completed January 21 for period August to December 29 12

13 Withdrawals Number of Withdrawals from Formal Savings January-December Apr-9 Jul-9 Jan Data for formal withdrawals was not collected in the second round. Rupees Agriculture Pisciculture Medical Other Average Amount of Withdrawals As formal savings were not a part of the project and therefore not very carefully tracked, the reasons for withdrawals from formal savings accounts could not be fully explained by staff. It is also important to remember from Graph 8 that the data in Graph 1 refers to 4 of 3 members who have formal savings. However, it is clear that agriculture was a big expense in the first round, as land was more widely available and SRI cultivation was encouraged. Capital from the project was not provided for this activity, so members had to use their savings for this initial investment, which can explain the high withdrawal amount in April 29. While the demand for agricultural activities was there in the last round, land availability remained scarce. Medical expenses increased in the final round, as the monsoon brings more illnesses. Graph 11: Informal Loans 2 nd Round: completed July 29 for period April-July, 29 3 rd Round: completed January 21 for period August to December 29 13

14 Loans from Informal Sources (SHGs) for Primary Reasons January-December 29 Withdrawals Rupees Apr-9 Jul-9 Jan-1 Agriculture Medical Food Pisciculture House Repair Average Amount of Withdrawals Informal withdrawals (loans) from the Trickle Up SHG could be more easily explained by staff. As with formal withdrawals, there was a great deal of investment in agricultural land in the first round. It was mentioned earlier that many members sold off their assets during the first round in anticipation of purchasing agricultural inputs in the future. While the amount received from the assets could pay for part of the total cost of inputs, the results suggest that members still had to borrow some money to cover agricultural costs. These withdrawals decreased with the dry season, and increased during monsoons for those who could afford the inputs (typically those who had their own land and simply needed the additional inputs). Medical expenses (typically from the local doctor) were also high in the beginning due to a relatively high prevalence of pox and diarrhea at the time. As these illnesses cleared, medical expenses decreased. It could also be that in the second round, members borrowed from other sources for medical purposes because they had outstanding SHG loans that needed to be repaid. Medicalrelated loans shot back up in the last round due to monsoon-related illnesses (fever, cough, etc), and perhaps due to the fact that members had repaid their loans and thus could borrow again from the SHG. Given that food security improved over the course of the program, loans for food consumption sharply decreased. This is most likely a result of diversification of income, kitchen gardens and fruit trees, benefits from land cultivation and social messaging about nutrition. Hurricane Aila hit in May 29, making home improvements an exigent need in the second round. People normally complete home repairs before the monsoons in July, to avoid the challenges of working in the rain. Some used SHG loans to finance the home repairs, but most borrowed from 2 nd Round: completed July 29 for period April-July, 29 3 rd Round: completed January 21 for period August to December 29 14

15 moneylenders. The reasons for this are two-fold: firstly, the costs associated with rebuilding are extremely high and could not be covered by an SHG loan. Secondly, many members had outstanding loans, and are not allowed to borrow again from the SHG until they have repaid their existing loan. This explains the low number of SHG loans for this purpose. Pisciculture loans increased in the second and third rounds. The program pushed members to diversify in pisciculture in July 9 by giving them fishlings or grants to invest in this activity. But the distribution to some was delayed and, given the critical timeliness of fish cultivation (July December), many members grew impatient and took SHG loans to purchase fishlings. Graph 12: Frequency of most prevalent illnesses Clients Frequency of Most Prevalent Illnesses January-December Pox Fever Diarrhea Cold & Cough Apr-9 Jul-9 Jan Headache Malaria Appendix Jaundice According to Graph 12, which illustrates the frequency of illnesses among TUP members, we can see that fever has shown the greatest increase. This is to be expected, as fever is normally associated with the monsoon period. Seasonality can also explain the relatively high increase in malaria. Pox was high in Round 1 due to an outbreak and subsided in subsequent rounds. A decrease in the prevalence of diarrhea surprised the staff, as this typically increases during monsoon period when people are more likely to drink contaminated water. They attribute this decrease to programmatic interventions, particularly social messaging around consuming boiled water, the emphasis upon health/hygiene through provision of sanitary latrines, etc. They also attribute the decrease in jaundice to the program, as they would expect the prevalence to be higher in monsoon. The fact that the most prevalent illnesses are easily treatable and not terribly serious is also noteworthy. Graph 13: Most prevalent sources of treatment 2 nd Round: completed July 29 for period April-July, 29 3 rd Round: completed January 21 for period August to December 29 15

16 Clients Most Prevalent Sources of Treatment January-December Apr-9 Jul-9 Jan Local Healer Hospital Nursing Home PHC Given that the most prevalent illnesses are minor, we see that most members treated their ailments through local healers (which refers to village barefoot doctors). This is also an easier option for members who do not have a primary healthcare clinic (PHC) nearby. For those who do have access, visits to the PHC for minor ailments have also increased. The staff feel this improvement in healthseeking behavior is a consequence of the TUP health worker s efforts to encourage people to visit local health facilities rather than ignoring symptoms and ailments. The program also offered a health fund where heads of household could claim reimbursements for medical expenses, which also incentivized visits to the doctor when ill. Members reserve visits to the government hospital for serious illnesses, as the hospital is located in the city and transport remains a problem. The staff are unsure of why hospital visits fluctuated between the second and third rounds. Graph 14: Social changes 2 nd Round: completed July 29 for period April-July, 29 3 rd Round: completed January 21 for period August to December 29 16

17 Social Changes January-December 29 Apr-9 Jul-9 Jan-1 78% 83% 81% 59% 57% 56% 58% 67% 65% 51% 46% 51% Percentage of School aged Children Going to School Clients Purchasing More Clothes Clients Attending more events Clients Buying more presents for others Generally, we see that social changes have remained fairly constant and have not fluctuated much over the data collection periods. According to the staff, some of this can be explained by seasonality and custom. In terms of attending more social events, for instance, round 2 coincided with a major religious festival/holiday (Durga Puja), and Round 3 coincides with Hindu marriage season. During Durga Puja, it is not typical to give gifts, but it is for wedding season, explaining the slight increase between rounds 2 and 3. Qualitative interviews highlight that improved social standing attributes to the fact that members are now receiving more invitations during wedding/festival time, and greater economic security may also lead them to accepting more invitations. The staff explain that even with greater disposable income, members will not typically purchase new clothes outside of important occasions. Even though members had more disposable income this past year, given the expenses of reconstructing their homes after Aila, staff expected this figure to go down in the last round rather than stay fairly constant. Overall, this is not an indicator they would expect to increase with economic improvement. In terms of school attendance, staff attribute the overall increase to social messaging about the importance of education. The slight decrease in the last round is due to the fact that children in primary school graduated, and have no secondary school services available in the vicinity. Social messaging also stopped with the ending of the program, so staff were unable to push members to enroll them into secondary school. 2 nd Round: completed July 29 for period April-July, 29 3 rd Round: completed January 21 for period August to December 29 17