12.07.2015 Views

Organization and Performance of Cotton Sectors in Africa ... - infoDev

Organization and Performance of Cotton Sectors in Africa ... - infoDev

Organization and Performance of Cotton Sectors in Africa ... - infoDev

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

ANNEX A10. METHODOLOGY FOR FOCUS GROUP DISCUSSIONON FARMER TYPESIn several <strong>of</strong> the study countries, formal household surveys provide data oncotton yields <strong>and</strong> <strong>in</strong>put use that can be stratified to show the performance <strong>of</strong>different categories <strong>of</strong> farmers. However, the difficulties <strong>of</strong> collect<strong>in</strong>g reliablelabor data <strong>in</strong> such surveys means that most <strong>of</strong> them do not conta<strong>in</strong> <strong>in</strong>formationon labor use by different categories <strong>of</strong> cotton farmers.To provide <strong>in</strong>sights <strong>in</strong>to labor use at modest cost, focus groups undertookdiscussions <strong>in</strong> seven <strong>of</strong> the n<strong>in</strong>e study countries (all except Cameroon <strong>and</strong>Ben<strong>in</strong>). Discussions were undertaken <strong>in</strong> two (Burk<strong>in</strong>a Faso, Mali) to six(Mozambique) villages per country, with efforts made to compare acrossregions or districts where there were considered to be important geographicaldifferences <strong>in</strong> performance (Mozambique, Tanzania, Zimbabwe, Zambia) <strong>and</strong>also between more <strong>and</strong> less accessible villages with<strong>in</strong> an area (Tanzania,Mozambique).In all cases except Mozambique, a s<strong>in</strong>gle group <strong>of</strong> <strong>in</strong>formants provided<strong>in</strong>formation on cotton farmers <strong>in</strong> their village. There were commonly 5–10<strong>in</strong>formants per village. In Mozambique, a larger number <strong>of</strong> respondents weredivided <strong>in</strong>to groups (based on level <strong>of</strong> cotton production) with each groupprovid<strong>in</strong>g <strong>in</strong>formation on its own activities, albeit <strong>in</strong> the presence <strong>of</strong> peoplefrom other groups.The first activity <strong>in</strong> each focus group discussion was a participatory rank<strong>in</strong>gexercise based on the pr<strong>in</strong>ciples <strong>of</strong> wealth rank<strong>in</strong>g. Where possible, 76 the name<strong>of</strong> every head <strong>of</strong> household <strong>in</strong> the village was recorded; then <strong>in</strong>formants wereasked to place the cards <strong>in</strong> piles accord<strong>in</strong>g to the level <strong>of</strong> cotton productionachieved by the household <strong>in</strong> a “normal” year. In most cases, this group<strong>in</strong>g producedthree or four piles. If one pile (typically the lower producers <strong>in</strong> ESA) wasmuch larger than the others, the researchers asked for a further disaggregation<strong>of</strong> this group. In Mozambique, farmers were asked to divide themselves <strong>in</strong>togroups; after verification <strong>of</strong> each farmer’s yield, each group responded <strong>in</strong> turn.Once the groups had been identified, the <strong>in</strong>formants were asked to describethe characteristics <strong>of</strong> households <strong>in</strong> each group, thereby cover<strong>in</strong>g demography,<strong>in</strong>come sources, <strong>and</strong> food production, as well as cotton production. This analysisgave a picture <strong>of</strong> a typical household <strong>in</strong> each group. A crop budget was thendrawn up for each group for one hectare (or acre) <strong>of</strong> cotton <strong>in</strong> a typical recentseason. 77The overall assessment is that the focus group discussions were a cost-effectiveway <strong>of</strong> collect<strong>in</strong>g reasonably reliable data on cotton production activities by differentgroups <strong>of</strong> farmers. However, the follow<strong>in</strong>g issues are noted:■The method tends to accentuate <strong>in</strong>tergroup differences at the expense <strong>of</strong><strong>in</strong>tragroup variation. Thus, top groups are characterized as be<strong>in</strong>g able torise above many <strong>of</strong> the problems that constra<strong>in</strong> poorer households, <strong>and</strong> the134 POULTON, LABASTE, AND BOUGHTON

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!