Short Notes Ravi Shanker Sustainable Livelihood Analysis1

Understanding livelihoods in rural India: diversity, change and exclusion: Key concepts and Issues; Manage June 2003. Compiled by Ravi Shanker for National Workshop on Watershed Management, Ahmedabad.2004. CRD Gandhinagar. Ministry of Rural Development. Government of Gujarat.

Sustainable Livelihood Analysis2

For any project to be adopted on the lines of SLA the project has to invest good amount of time in understanding project community, environment and conditions in which the project population is living. Through investigation one would understand the aspirations, processes and intended change in the environment. Existing livelihood patterns need to be studied and options must be analysed. During this process livelihood research purpose however needs to be clear from start as it should guide the planing of methods. 1. Responsive, short term program related assesments 2. Policy research 3. strategic indepth academic research Four steps of Research process Useful elements 1. Planning and getting going Define purpose Prioritize topics and questions to answer. Plan data gathering and practicalities 2. Information gathering Existing literature Fieldwork variety of methods Include on-the-spot data-checking,

documentation 3. Analysis and interpretation Aggregate data Compare data Tabulate data Interpret and summaries key findings Identify target audiences Main messages for each feedback

4.

Dissemination

Followed by

Combining qualitative and quantitative data While qualitative data is rich in details, it is difficult to make meaningful comparisions. Quantitative data enables us to compare and correlate. Linking qualitative and quantitative analysis enables in-depth research. Qualitative methods, such as key informant interviews and oral histories are often very rich in details. They put information in context, and are able to explain issues such as why.

Understanding livelihoods in rural India: diversity, change and exclusion: Key concepts and Issues; Manage June 2003. Compiled by Ravi Shanker for National Workshop on Watershed Management, Ahmedabad.2004. CRD Gandhinagar. Ministry of Rural Development. Government of Gujarat.

Quantitative methods such as questionnaire can provide a large amount of information, relatively quickly, for a large representative sample. A combination of methods is essential. One key step in SL analysis is to cluster respondents into different groups and compre across them (caste, land holding, livelihood class, and village). Another step is to assess the cause of change, economic success, impoverishment for each household and again aggregate these to derive larger patterns. Sampling Both representative and purpose sampling can be useful, depending on context and objectives. A representative sample-usually selected reandomly with only small fraction- often 30% or less of the population sampled, the paremeters of the population can be estimated. When there are different categories of overall population, a stratified sample is needed to ensure that the proportion of respondents from each in the sample. Cluster could be adopted when scope for above methods is leniar. Purposive sampling internationally includes respondents from key groups because they are necessary example; atleat 10 landholders in a sample of 50 respondents to investigate land issues. Most important information to cover in a survey Productive assets (land, livestock, education, household labor) Deomestic assets (house, consumer durables) Liabilities (debts) and interest rates Crops, costs of inputs and returns Labor market work Domestic work Enterprises Government receipts Average daily expendeture on food Average daily expendeture on food Average monthly expendeture on cloths, health etc Typical annual taxes Plus key variables that define the type of household, size, caste etc. Analysing data Questions that request data on income of family can be both sensitive and liable to produce unreliable replies. Such questions could be divided in parts usually by separating the quantiity of activity from typical unit returns. For example; how many days worked in season, at what wage rate, or how much land cropped, at what yeild, price at the time of harvest. Similarly expendeture questions could be divided in to short term spending items and annual spending items.

Livelihood tracking identified the main causes and patterns of shifts. Findings in one MP village demonstrated that privatization of the local tank/pond had forced most of the village households out of their occupation of fishing and into agriculture. Most did not have access to land were pushed into agriculture labor at lower wages because they did not have agriculture-labor skills. Repeated analysis for a number of households would lead to identification of patterns within and between villages. This is called livelihood tracking. This would also reveal the interplay between internal forces (death of productive household member) and external forces (drought). Synthesising and aggregating data are key elements of analysis. The qualitative and quantitative data need to be trhoughly corelated and analysed. Aggregating results for different types of households, then comparing across groups- caste, land-holding to identify differences in livelihood options. Synthesising data on assets, activities and incomes to identifying major livelihood types or classes that represent policy-relevant groups for further analysis. Project people themselves will participate in the process. The findings need to be shared with project people and their concent need to be secured. Combining analysis by caste, land and income helps explain why some different castes are in similar income groups- landed SCs and landed OBCs are similar, as are landless SCs and landless OBCs-or why some landless are not the poorest, given their income source from non-farm income. For example in AP, local livelihood support systems are divided on caste lines. A village common land could be given to poor OBCs by Panchayat instead of poorer SCs. Hence the caste, class and gender dynamics should be analysed.