Patterns of Change in Western Nepal: Rural Households of the 1970s and 1980s Compared

IntroductionThis paper presents the initial findings of a ruralhouseholdsurvey undertaken in Western Nepal during1996-97 as part of a larger project on u0027rural livelihoodsand long term changeu0027 (on which a preliminarymethodological note is provided by Bagchi et al. 1(98).The survey was undertaken by members of the OverseasDevelopment Group (ODG) at the University of EastAnglia together with members of Actionaid Nepal, theLutheran World Federation and the Womenu0027s Divisionof HMG Ministry of Local Government, with a view toproviding the basis for a comparison with an earlierrural household survey undertaken in what was thenWest Central Nepal by members of the ODG in themid-1970s. also as part of larger project.The Approach and MethodologyFieldwork for the Rural Livelihoods TrajectoryResearch Project (funded by ESCOR for the BritishDepartment for International Development, DtlD) wasconducted in Western Nepal during 1996 and 1997. Theapproach adopted was u0027multi-methodology, , includingboth quantitative and qualitative techniques, to gaininsights into long-run cultural, ecological, economicand political processes. The analysis of the findings ofthe research as a whole is currently being undertaken,and it is anticipated that a fuller report of the analysiswill be published in the near future.This paper, however, utilises only the results of aformal survey of rural households to allow directcomparisons with the results of a similar householdsurvey conducted in the same region in1974/75 for aprevious ESCOR funded project (the major results ofwhich were published in Blaikie, Cameron and Seddon1980). The sampling frame for the 1996/97 survey wasdeveloped with the same objective as that of the mid-1970s: to provide unbiased estimates of householdsu0027livelihood positions for the region with somereasonable claim to representativeness.The 1975 Rural Household SurveyThe 1974/75 Project included a rural surveyinvolving 667 interviews by a team of aboutfive enumerators employed for roughly a year. Thesampling frame was designed to produce an unbiased,self-weighting sample of the whole population of ruralhouseholds ofWest Central Nepal (an area more than10,000 square kilometres with about two millionpeople).Claims to representativeness were first establishedby the use of multi-dimensional stratification, allowingfive Districts to be randomly selected within the WestCentral Region from groupings with similarcharacteristics. Clustering was then necessary to meetlogistic requirements, and 44 Gaon Panchayat wards (thesmallest unit of local government at the time) wererandomly chosen as the maximum manageable number.The number of wards in three of the Districts wasproportional to population recorded in the 1971 Census;they were lower in one relatively urbanised District(Kaski) and higher in the one District with a claim to berepresentative of higher mountain conditions (Parbat).Within the Districts every ward was given an equalchance of being selected by using a random numbertable.A uniform proportion of one quarter of householdswas chosen in every selected ward to give aproportional-to-popu lation-size, self-weightingcharacteristic. The final households were randomlyselected in the field from an updated electoral register.Interview refusal or non-availability was very rare arrlso substitutions at the household level were handledan