«Deborah Duveskog Faculty of Natural Resources and Agricultural Sciences Department of Urban and Rural Development Uppsala Doctoral Thesis Swedish ...»
7.2 Research sites The research was undertaken in the context of the IFAD FFS project in Kenya, Uganda and Tanzania and field sites defined accordingly to the two or three districts in each country that were part of the project. All sites were fairly high in agricultural potential, high-populated locations with rainfall in the range of 1000-2000 mm/year.
Western Kenya The field sites in Kenya were located in the districts of Kakamega, Busia and Bungoma, which are fairly similar to each other in terms of agro-ecological and socio-economic situation. Agriculture is the main economic activity in these districts with maize, beans, groundnuts, vegetables and livestock, especially poultry, as predominate production enterprises. Cash crops such as coffee, tea and sugar are also grown. Much of western Kenya is considered to have good potential for agriculture, however the area is relatively highly populated and land holdings often small. The history of farming in the area, however, is characterized by low input – low output farming. The lack of land has led to overexploitation of land resources with highly nutrient poor soils as result. Much of the tree cover has been removed. Rainfall is seasonal, reliable and range between 1,000-2,400 mm per annum, which allows two cropping seasons. Topography is rolling hills with scarps, and with potential for irrigation. The economy is largely public sector and subsistence driven and the districts have limited infrastructure facilities in terms of roads, information resources, value addition plants etc. Population densities in the region are among the highest in rural Kenya at an average of 950/km2. There is an estimated 1.8 million people in the districts of Kakamega, Busia and Bungoma (1999), and of the order of 75 % of people under the age of 30 years. Luhya is the most common ethnic group found in the districts. Poverty levels are high at an estimated 50 % in absolute poverty. A national study of poverty found Western Province to be one of the poorest in the country (Republic of Kenya 1997). It was estimated that 31.5% of households in western Kenya are among the hardcore poor, as opposed to 19.6% for all rural areas. Western Kenya is centrally located within the country and within East Africa, it is on the main trading routes between the coast/Nairobi and the hinterland of Uganda, Rwanda and the DR of the Congo, and adjacent to Kisumu – a main lake trading centre.
Eastern Uganda In Uganda the field sites were located in Soroi, Kaberamaido, and Busia districts, where the two first are adjacent district in the north east, while Busia is located on the border to Kenya on the east. The situation in all district are fairly similar with Busia providing a higher potential context, and is also more favourable located in terms of trade etc. than the other two districts. Rainfall ranges between 1000-2000 mm/year with Soroti having the driest conditions.
There are two rainy seasons per year between April-June and AugustNovmber. Infrastructure in terms of roads is fairly well developed.
Kaberamaido is one of the districts with fastest growing population in the country with a 99 people per Sq km of land and has many up-coming trading centres scattered all over the distinct. Soroti is much more sparsely populated with only half of the population density of Kaberamaido.
During the eighties agriculture was depressed by civil war, but following the peace in the 1990s the area has experienced extensive agricultural growth.
Agriculture in the area is fairly high potential, despite often poor and shallow soils, with most of the population depending on farming for food and income.
They use animal traction (oxen) to plough the land while hand hoe is the basic tool for cultivation. Crops grown in all districts include maize, potatoes, cassava, groundnuts and beans. In Soroti and Busia cotton and coffee are also important crops, and in Kaberamaido and Soroti; millet, rice and potatoes are grown. Soroti is one of the leading suppliers in the country for sweet potatoes.
There is a high livestock population in all districts, in particular in Busia. In Soroti the cattle population was reduced to nearly zero following extensive cattle rustling during the war (Government of Uganda 2003), but is slowly increasing again.
Kagera region in Tanzania The field sites in Tanzania include the districts of Bukoba rural, Muleba and Karagwe which all are located in the Kagera region. Kagera is the most remote region from the administrative centre of Dar es Salaam along with Kigoma.
The isolation is further compounded by poor roads into the region and by being squeezed between the neighbouring countries of Uganda, Rwanda, Burundi and Lake Victoria in the east. The geographical isolation and the proximity to three foreign countries have made Kagera vulnerable to foreign influence and in particular influx of refugees. Kagera has thereby suffered severely from refugee damage, including severe deforestation, poaching of game reserves, and overload of infrastructure and service facilities. Education levels are high due to the history of early European missionaries The region’s climate is influenced by its proximity to Lake Victoria, with higher rainfall on the shore strips and the highlands close to the shores, The rains are bimodal; March-May and October to December, with an average annual rainfall of 800-2000 mm. The region is generally considered as the banana and plantain country and the land of coffee. Soils in the area have high iron and clay content, but low in nitrogen, phosphorus and are acidic. Soil erosion is a serious problem especially near the lake.
The farming system is divided in three distinct agro-ecological zones: Lake shore and islands, receive the highest rainfall, growing mainly bananas, cassava, beans, coffee and tea and where farm size range between 1-2 acres.
The Plateau area; with moderate rainfall growing mainly bananas, beans, maize, cassava and coffee, and where farm sizes are 2-10 acres. Lowlands; flat plains with low rainfall and only one rainy season, with main crops being cassava, rice, sorghum, millet and maize and with cotton as the main cash crop, and farm sizes ranging between 3-5 acres. Kagera region has further a long history of the development of cooperatives, with over 222 agricultural marketing cooperatives in 2002 and 115 saving and credit cooperatives. The cooperatives in Kagera have not suffered the large collapse as compared to as for example Kenya, and continue to grow even though often faced with management problems (United Republic of Tanzania 2003).
Figure 10. Location of research study sites in Kenya, Uganda and Tanzania 8 Methodology
8.1 A combined methods approach After considering the broad scope of my research and the questions and issues that needed exploration I decided that no one single methodology would adequately capture all of the required information. I needed to apply a variety of methods and tools, and therefore chose to combine qualitative and quantitative measures in order to capture the depth of issues while at the same time achieve some degree of generalizability. Carvalho and White (1997) elaborate on qualitative and quantitative approaches in relation to poverty related analysis and concludes that quantitative approaches can be characterised as having breadth, while qualitative having depth. There is a growing recognition that to understand social phenomenon, a combination of data collection methods are necessary, despite that these differs substantially with respect to their epistemological foundations. Quantitative strategies relate more to positivism and objectivism, through deductive testing of theory approach while qualitative strategies relate to constructivism and interpretivism through an inductive generation of theory approach (Bryman 2004).
Despite applying a range of quantitative tools, I have considered my research predominately to lie within a constructivist and qualitative perspective. Theoretical understanding evolved during actual research, through continuous interplay between analysis and data collection as described by Strauss and Corbin (1994). Bateson (1979) explains there are fundamental differences between the world of non-living things and living processes, where order arises from the patterns of information flow rather than from physical relationships of cause and effect and where differences in quality and more profoundly important than differences in quantity.
The constructivist philosophy is generally inherent in community-based action research, where the researcher and the community work together to generate new knowledge. Similarly, this research seeks to engage “subjects” as equal participants in the research process (Stringer 1999) and scientific objectivity is not the purpose of the research. Knowledge is thereby socially constructed and objective and value-free science is impossible (Bryman 2004).
Practically the research combined qualitative and quantitative approaches in a manner consistent with what Carvalho and White (1997) indicated brings out the best of both. Much of the study was conducted in this way drawing on participative methods of inquiry. Qualitative processes such as exploratory workshops and focus groups were used to help frame indicators of well-being or empowerment that then were applied through qualitative measures in household surveys. Qualitative processes were also used for enriching and confirming findings generated from quantitative tools. Yet, the quantitative aspects of the study have been important for the theoretical development as well. Quantitative data in form of household surveys were used to focus in on particular sub-groups or individuals for sub-sequent follow-up qualitative study. Finally combining findings of qualitative and quantitative measures helped in gaining a more holistic view and understanding. That transformative learning emerged as a fitting interpretative framework for the learning processes were not given, but an outcome of the analysis.
The table below illustrates the purpose of using both qualitative and
quantitative approaches in my research:
8.2 Data collection tools and methods Household surveys The quantitative data source for this study comprise of a combination of faceto-face questionnaire surveys of a total 1203 households carried out in Kenya, Uganda and Tanzania during 2004-2007. The samples in the different countries included non-FFS members, FFS pre-members (enrolled for FFS but not yet started) and FFS members (FFS/NAADS group members in Uganda). One major impact survey was carried out within the scope of the research while data from several additional surveys were used in the analysis. The face-to-face impact survey undertaken included about 300 graduated FFS members in Kenya and Tanzania and was carried out in 2007 with randomly sampled FFS participants from FFS groups started back in the years 1999-2002.This dataset was used to compare the post FFS situation to the pre-FFS scenario. The pre
situation was defined by the following datasets:
In Kenya and Tanzania: Data in each country were collected in 2006 through a stratified random sampled survey with about 280 individuals signed up for FFS (but not yet having commenced participation in FFS, i.e. FFS premembers) and 120 non-FFS households. A two-stage random sampling technique was applied to select FFS pre-members with 20 FFS groups per country randomly selected, divided proportionally per district. Thereafter household members were selected based on lists of households in the selected FFS groups. Non-FFS participants were randomly sampled in neighbouring villages (without FFS activities ongoing) to the selected FFS groups, by means of village and household name lists obtained from the local administration.
In Uganda (Soroti district): A survey questionnaire was implemented in 2007, managed by Danish Institute of International Studies (DIIS), with 403 respondents. Respondents were randomly selected in the district irrespective of FFS membership status. During data analysis, groups of FFS graduates and non-members were then separated for comparisons. In Uganda the NAADs program provides the dominant framework for collective activities among farmers, and most FFS groups that started in 1998-2001 had turned into NAADs groups. Therefore the sample of FFS members in practice included both FFS and NAADs participants, while the non-member groups included neither FFS nor NAADs members All survey interviews were conducted with the help of a formally structured questionnaire under the supervision of the researchers. Trained enumerators, knowledgeable in the local language, carried out all surveys. The surveys were field-tested before being implemented in the countries. The questionnaire included a range of aspects such as poverty indicators, the adoption of agricultural technologies, economic and institutional issues, personal and collective agency, attitudes, perceptions of power etc. The content of the surveys in Kenya and Tanzania was largely identical, while there were a few variations in the Uganda survey format. Data analysis was carried out using SPSS software.
Well-being ranking methodology One part of the survey related to capturing the overall well-being of the households. This part builds on innovative experience in East Africa with developing well-being indicators identified by farmers (Ravnborg et al. 2004;
Friis-Hansen 2005), initially tested on large scale in Uganda and later verified in Tanzania. Multidimensional and participatory poverty well-being indicators were identified by farmers through small groups of community members, through household ranking and description, statistical testing of the indicators and finally translation into 13 categories of farmers perception of well-being.
Based on these indicators a household poverty index was computed.
Explorative participatory seminars Larger stakeholder group events, such as participatory and interactive seminars were used particularly in the design stage of more in-depth study for framing of indicators for quantitative tools or for framing of checklists for focus group discussions or key informant interviews.