2005 IND: Evaluation of UNICEF’s Drought Mitigation Project in Madhya Pradesh, 2001-2003
Author: Winrock International India, New Delhi, in association with the Centre for Advanced Research and Development, Bhopal
Large tracts of India are prone to droughts. 68 percent of India’s land mass is drought-prone to varying degrees, of which about 50 percent is chronically drought-prone. As per Government of India’s estimate, there are one or two years of droughts every five years in semi-arid and arid regions of India. Every year, GoI spends millions on relief for creating employment, provisioning water supply, foodgrains and fodder to mitigate the impacts of droughts. With a view to enhance drought proofing, GoI has accorded high priority to its watershed development program. This is evident from the fact that GoI has developed a perspective plan of 20 years (2002-03 to 2021-22) for treating around 88.5 Mha of land with a total investment of Rs. 727.5 billion.
Between 2001 and 2003 UNICEF assisted the Government of Madhya Pradesh (GoMP) for mitigating the impacts of drought in 34 of 45 districts of the state. This evaluation was commissioned by UNICEF to draw lessons to from this experience to feed into better design, planning and implementation of such interventions. It is worth noting that this evaluation looks at UNICEF’s support within the larger framework of GoMP’s approach to and efforts towards mitigating droughts and drought relief work in the state.
Purpose / Objective
The specific objective of this evaluation was to assess, primarily from the perspective of various socio-economic sections of drought-affected communities, to what extent the project objectives of long-term drought proofing and short-term livelihood sustenance through wage employment were achieved.
The districts were selected using two levels of stratification. Firstly, based on data from the MP State Human Development Report 2002, the drought resilience index2 of various districts was computed. Subsequently, the vulnerability index of various districts was calculated by aggregating drought resilience index, proportion of SC/ST population and percentage of land degradation. Based on the vulnerability index of various districts to droughts, 11 districts were shortlisted from the 34 intervention districts. Secondly, based on this shortlist, 10 districts were finally selected through meetings with UNICEF and the Government of MP, which provided information on project performance in the shortlisted districts. Based on suggestions received, 6 of the 10 districts were selected from Phase II of the UNICEF Drought Relief project and four from the Phase I.
The survey tools used for the evaluation helped capture both the qualitative and quantitative dimensions of data. These included Semi-structured interviews, Participatory Resource Mapping, Village Schedules, Observation of structure, Focus Group Discussion (FGD), Drought Timeline and Household Questionnaires. Separate FGDs were organised with various sections of the community—landless and SC/ST, Small and Marginal farmers, Women groups and Panchayat members. For each key issue, a set of primary tools and secondary tools were selected from the above list. While Primary Tools were the main source of information and reflected the emphasis on data analysis and presentation, Secondary Tools were seen as complementing/ substantiating/ triangulating information obtained through Primary Tools.
Key Findings and Conclusions
Short-term mitigation of drought impacts
Livelihood & Food Security:
• Across 142 structures that were constructed under the UNICEF Drought Mitigation Program a total of 282,861 person days of wage employment was generated.
• The construction of ponds and tanks, and renovation of ponds provided significant wage employment benefits especially to SC/ST and OBC communities.
• Women received 1,900 person days of wage employment more than men only in the case of ‘construction of ponds’. In the case of all the other works, men received a larger share of the total wage employment that was generated. Highest disparity was noted in the case of ‘construction of wells’ (67% less than men) followed by ‘renovation of ponds’ (53% less than men).
• In terms of support to livelihoods, the ‘construction of tanks’ provided the maximum relief (cash and cereal), followed by the ‘construction of ponds’, and finally the ‘construction of wells’.
• UNICEF assistance was effective in targeting SC and ST communities. The district-wise proportion of SC beneficiaries was higher in Chattarpur (38%) and Sehore (41%), while Jhabua (90%), Mandla (90%) and Ratlam (72%) had high percentages of ST respondents.
• In UNICEF drought relief assistance a ‘cash to cereal ratio’ of ~ 2 was found.
• Over 90 percent of the beneficiaries were satisfied with the type of food grains provided, whereas with regards to quantity and quality 43 percent found them to be satisfactory. Similarly, the wage rate was satisfactory for over 87 percent of the beneficiaries and payments were made in time. Equal wages were paid to both men and women. The payment in 57 percent of cases was made on a weekly basis and in 66 percent cases made by Panchayats.
• Average of 10 percent of cost was incurred on machinery/material with highest being for well construction (33%) and practically no cost incurred on machinery/material in renovation activities (ponds and tanks).
• Of 1377 food insecure households 47.5 percent households reported that the wages and food grains received from project helped them to successfully tide over the phase of food insecurity. Migration
• In larger households (more than 8 members), on an average, the proportion of migrants (37.5%) was lower than in the case of smaller four-member households (50%)
• Overall, the number of migrants in 5-6 member households was found to be the highest. Jhabua reported the highest number of migrants per household (3-4 per household), while Mandla and Sidhi stand at the lower end with an average of 1 migrant per household.
• Migration was observed to be the highest in the vulnerable, medium income households (Rs.12,001 – 25,000 / annum), while the very poor or ‘destitute households’ (<12,000 / annum) did not have the ability to adopt migration as a coping strategy during droughts. Therefore, drought mitigation efforts need to specifically target this group of ‘destitute’ households.
• Landed households in districts that are moderately prone to regular migration were forced to migrate under drought conditions along with those dependent on wage labour (agri and non-agri based).
• Highest migration was seen among Schedule Tribes.
• Chronically poorer and high migration prone six districts, (Seoni, Jhabua, Rajgarh, Ratlam, Chhatarpur and Sidhi) reported upto 16 percent of additional migration
• Districts moderately prone to migration (Khargone and Dhar) had a moderate increase in migration (20 % to 24 %). Two districts, (Mandla & Sehore) not considered as migration prone, showed an extra ordinary rise in the migration rate.
• For communities that practiced ‘regular or traditional migration’, a distinction between distress and regular migration cannot be made as the former tends to get absorbed in the latter. However, in the case of communities that did not practice regular migration distress migration emerged as an important coping strategy.
• The degree of UNICEF assistance was the highest in Ratlam, Khargone, Chattarpur and Seoni. It was found that 91% of the households that benefited from wage employment in Ratlam belonged to the SC/ST category, followed by Khargone (83%), Seoni (70%) and Chhatarpur (54%).
• Across the 10 sample districts, on an average, 63% of the beneficiaries were landless and marginal farmers.
• A high degree of women’s participation was found (46.32%), in terms of receiving wage employment, under the UNICEF program.
• Effectively, on an average, while one man received 69 person days of wage employment, one woman beneficiary received 56 person days of employment
• 43 percent of the beneficiaries were from the ‘medium income’ bracket (Rs. 12001 – 25,000/ annum) while 18.44 percent belonged to the BPL category (below Rs 12,000/annum). The percentage of BPL households was found to well below the percentage of beneficiaries belonging to the APL category (around 29%) - Rs. 25,001-50,000/ annum income bracket.
• The program coverage, as a percentage of the population covered to the total population of the project village, ranged from 10 percent in Chattarpur to 100 percent in Jhabua with overall program average of 31 percent. Further, a minimum of 33 percent of total labourers were sourced from outside the project villages. These ‘outside’ labourers were mainly from the same Panchayat as that of the project village.
• UNICEF’s drought mitigation program has been quite successful at targeting SC/ST and BPL households. More than 80 percent of SC/ST and BPL populations of project villages were covered by program in Chhatarpur, Khargone, Mandla, Ratlam and Seoni. The average coverage of SC/ST and BPL households at the program level was 73 percent.
Long-term drought proofing
• Three types of structures (ponds, tanks and wells) were either constructed or renovated. It was found that all the structures were constructed within the estimated /sanction cost—there was no cost overrun. The average overall sanctioned cost per (new) structure worked out to Rs.1,85,873 against an average expenditure of Rs.1,51,138.
• Of 128 works studied, 80 (62%) were in the lowest cost range of below one lakh, 19 (15%) were in the 1 - 2 lakhs bracket, 20 (15%) were in the cost bracket between Rs. 2 lakhs and Rs. 5 lakhs, while 9 works (7 %) were in the uppermost cost bracket of above Rs. 5 lakhs.
• Almost two-thirds of UNICEF assistance either went towards the renovation of existing structures or for the construction of small water harvesting structures, costing less than Rs. 1 Lakh.
• It was found that in 50 percent of cases the quality of work was reported as ‘fair’ or merely satisfactory. Only 5 percent of the works were rated as ‘very good’, 17 percent were given a ‘good’ rating and 20 percent were considered to be of poor quality.
• In the case of 60 structures for which an indepth technical assessment was carried out: It was found that 43% reported a loss of the impounded water through seepage.
- In 18% of the cases water loss due to seepage was a result of poor implementation (inadequate depth and width of puddle, inadequate or no use of clayey impervious soil in puddle, and poor compaction of bunds)
- Technical specifications were not always adhered to. Some structures were under-designed and more than the required free board was kept. Others were over-designed and a greater height of embankment, than desired, was observed (in 17% of the cases the height of the embankment was greater than the stipulated 6 m)
• 80 percent of the respondents were satisfied with appropriateness of sites of structures, 73 percent were satisfied with command and coverage of structures.
• Water security improved during the lean season in 45% of the villages, which resulted in the construction of new sources of water in 13% of the villages and improvement in yield of handpumps/Wells (existing) in 52% villages. Water for livestock and domestic needs were met in 87% and 35% of the villages, respectively.
• Increase in irrigation was reported in 23 percent of villages, which enhanced the yield of the Rabi (winter) crop in half of these villages. In three-fourth of these villages the increase was upto 10 percent while in the rest the increase was upto 30 percent. Institutional environment
• Pani Roko Samitis were not constituted in 39 percent of villages and in 73 percent of cases where they were constituted they did not play any role in program implementation. Strikingly, Pani Roko Samitis were constituted in 100 percent of villages in Mandla, while in Jhabua and Sidhi they were not constituted in any village.
• In 32 percent of villages Gram Sabhas were approving authorities. In Mandla, the Gram Sabha was the sole decision maker in all the villages.
• In 41 percent of cases villagers played key role in site selection while Panchayats and line department officials were involved in site selection in 30 percent and 12 percent cases, respectively.
• Villagers’ influence in the decision making process in construction of structures was limited only to the initial planning stage - in deciding on the location of the site, where 50 percent of respondents participated. The level of participation in next stages of planning was low, viz. command and coverage (16%), type of structure (3%), cost estimates (5%) and design (0.7%). This can perhaps be attributed to the fact that these stages were considered too technical for villagers to participate. Strikingly, participation of villagers in site selection for well construction was high (68%).
• Total community contribution amounted to Rs.10.46 lakhs - 5 percent of the total cost of 128 structures. Shramdaan (36.93%) and cash (3.03%) contribution were the two major modes of contributions. Cash contribution was not an up-front payment, but rather a pre-determined portion of the wages was deducted from the wage bill of the beneficiaries. Deductions at source (wages) varied from Rs. 5 per day to Rs. 10 per day. In effect, those not involved in wage labour did not contribute at all.
Evaluation of Drought Assistance of UNICEF in MP
• Community contribution was noted in only one-third (42) of the total 128 works studied. The contribution, as a percentage of the total cost of works, was high in the case of construction of ponds and wells
• At the village level there was no formalized system of community participation for the processes of planning, implementation and monitoring. This was done in an informal manner with small groups of influential villagers getting together and discussing these aspects of the intervention.
• It was found that only in 102, 94 and 103 villages, local communities participated in planning, implementation and monitoring respectively. Further the participation was limited to men folk and that too at the group level. BPL and SC/ST communities had a minor role in the planning process. The participation levels declined further in the implementation and monitoring phases.
• Women participated in the informal planning groups in 41 out of 132 villages (31%). But their influence in the planning related decision-making was insignificant, reflected by the fact that they influenced planning related decisions only in 19 out of 132 villages.
• SC/STs communities were part of the informal planning groups in 34 villages. Given that the majority of the respondents belonged to these two categories.
• Allocation of funds to districts to be made based on drought assessment reports: Need to develop strategies and procedures for targeting BPL households
• For curbing migration, districts that have low traditional / regular migration to be targeted.
• Supporting follow up programs necessary for translating natural assets created into potential livelihood options.
• Construction of tanks and ponds (costing more than Rs.300,000) to be promoted as the gross returns in the short-term (wage employment, community participation) and long-term (water security) are higher.
• There is a need to manage groundwater resources by first, undertaking scientific district-wise studies of scenario, and secondly, assessing additional scale of rainwater harvesting by taking into consideration hydrogeological profiles.
• Scientific assessments of local geo-hydrological conditions from point of view of aquifer’s transmissivity and porosity need to be undertaken in various geological formations–Vindhyan, Gondwana, Deccan Basalts and Quaternary Alluvium, to assess the recharge potential at the local level. For this, few structures that were constructed with DFID-UNICEF’s assistance in each of the geological zones should be selected. Based on this, a map showing potential recharge and discharge zones for various geological formations in the state should be developed.
• Detail water balance studies involving the assessments of ‘local’ supplies and demands need to undertaken, again in various geological formations, in order to decide the type and scale of water harvesting interventions that will be required in the area for long-term drought proofing. For assessing supplies, a system of monitoring on meteorological, hydrological and groundwater levels should be setup in few select structures. On the other hand, for demand assessment, detail land and water use studies should be taken up.
• Groundwater movement studies using Isotype tracer techniques need to be undertaken in different geological formations of the state for better site planning of water harvesting structures.
• Good network of monitoring wells need to be established for monitoring groundwater quality in pockets, especially where excessive fluoride concentration is observed.
• Long term drought proofing strategies has to help create strong institutions and separate funds/programmes should be allocated for capacity development of CBOs in order to engender greater community participation
• Short term drought proofing strategies should however not focus on creating institutions but build upon the existing institution base
• Panchayats should be the preferred implementing agency
• Effective monitoring systems for renovation works need to be developed.
Full report in PDF
PDF files require Acrobat Reader.