51713 The Cotton Sector of Zambia Africa Region Working Paper Series No. 124 March, 2009 Abstract T his country study is a background paper international markets. Credit recovery has typically prepared for the comparative analysis of been good, but turned suddenly worse during two organization and performance of cotton periods: 1999-2000, and 2006-07. Each credit sectors in Sub-Saharan Africa, a study carried out default crisis was occasioned by the entry of new by the World Bank, with the objective of analyzing cotton buyers, and resulted in sharp declines in the links between sector structure and observed production; yet five years after the first crisis, performance outcomes and thus draw lessons from production was double its pre-crisis high. Zambia's reform experience that can provide useful guidance cotton sector is now less strongly concentrated to policy-makers, other local stakeholders, and than during the first 10 post-reform years. Zambia interested donors agencies. It describes and thus illustrates the tendency of concentrated reviews the cotton sector situation in Zambia. sectors to become more competitively structured Since liberalization in 1994, Zambia has illustrated over time. Private stakeholders are currently the strengths and weaknesses of a concentrated working with government to institute a regulatory market-based cotton sector. Led for most of its Cotton Board to define "rules of the game" that post-liberalization history by two dominant protect input credit even in the presence of more companies, the sector has continuously expanded ginners. This regulatory challenge carries the number of smallholder farmers receiving inputs important implications for neighboring countries on credit. Yields have risen among experienced such as Zimbabwe and even for countries in West farmers, but the continual entry of new farmers has Africa, and merits close watching by researchers made overall yield growth modest. Lint quality has and policy makers interested in the future of cotton improved dramatically, with Zambia now receiving in Africa. the highest premium in Sub-Saharan Africa on Author Affiliation and Sponsorship David Tschirley Michigan State University, East Lansing "Tschirley, Dave" Stephen Kabwe Food Security Research Project, Lusaka, Zambia Stephen Kabwe The Africa Region Working Paper Series expedites dissemination of applied research and policy studies with potential for improving economic performance and social conditions in Sub-Saharan Africa. The Series publishes papers at preliminary stages to stimulate timely discussion within the Region and among client countries, donors, and the policy research community. The editorial board for the Series consists of representatives from professional families appointed by the Region`s Sector Directors. For additional information, please contact Paula White, managing editor of the series, (81131), Email: pwhite2@worldbank.org or visit the Web site: http://www.worldbank.org/afr/wps/index.htm. The findings, interpretations, and conclusions expressed in this paper are entirely those of the author(s), they do not necessarily represent the views of the World Bank Group, its Executive Directors, or the countries they represent and should not be attributed to them. COMPARATIVE ANALYSIS OF ORGANIZATION AND PERFORMANCE OF AFRICAN COTTON SECTORS THE COTTON SECTOR OF ZAMBIA Paper prepared for the World Bank by David Tschirley Stephen Kabwe March 2009 Aknowledgements T his paper is a background paper prepared for the comparative analysis of cotton sector reform in Sub-Saharan Africa, a study carried out by a World Bank team led by Patrick Labaste (Lead Agricultural Economist, SD Department, Africa Region, World Bank) and including David Tschirley (MSU), Colin Poulton (Imperial College London), Nicolas Gergely (consultant), John Baffes (DEC, World Bank), Duncan Boughton (MSU), Julie Dana (World Bank) and Gérald Estur (marketing and quality consultant). The study was funded by the World Bank and by contributions from bilateral trust funds particularly from Belgium (BPRP), the Netherlands (BNPP/CRMG), and the Swiss Secretariat for Economic Affairs (CRMG), as well as by the All-ACP Agricultural Commodities Programme (AAACP) of the European Union. This report also draws heavily on work conducted with support from the DFID-funded project Competition and Coordination in Cotton Market Systems of Southern and Eastern Africa and from United States Agency for International Development-USAID, through the Food Security III Cooperative Agreement between the USAID and the Department of Agricultural, Food, and Resource Economics at Michigan State University. The continuous financial support of the USAID mission in Zambia over eight years has been instrumental in allowing the research team to build up empirical insights grounded in a solid understanding of Zambia`s historical context. Additional financial and substantive support was provided by the Swedish International Development Agency. Special thanks to Ballard Zulu for his participation in previous research and for his continuing interest in this area. . i Contents 1. Introduction........................................................................................................................... 1 2. Historical Background ......................................................................................................... 2 2.1 Pre-Reform Institutional Set-up and Performance ......................................................... 2 2.2 Reform through 2003: Phases One to Three .................................................................. 3 3. Overview of the Cotton Sector ............................................................................................. 6 3.1 Macroeconomic Environment ........................................................................................ 6 3.2 Cotton Production: Trends, Geographical Distribution, and Farm Structure ................ 8 3.3 Cotton Ginning............................................................................................................. 12 3.4 Independent Cotton Traders ......................................................................................... 15 3.5 Cotton Spinning/Weaving/Apparel Industries ............................................................. 16 3.6 Oil and Cake ................................................................................................................ 17 3.7 Independent Input Dealers ........................................................................................... 17 4. Current Institutional Arrangements and Performance .................................................. 18 4.1 Sector Coordination ..................................................................................................... 19 4.2 Input Credit and Extension .......................................................................................... 20 4.3 Varietal Development & Agricultural Research .......................................................... 27 4.4 Seed Multiplication and Importation ........................................................................... 29 4.5 Quality Control ............................................................................................................ 29 4.6 Pricing .......................................................................................................................... 30 5. Cost Competitiveness, Returns to Farmers, and Sustainability ..................................... 32 5.1 Processing and Marketing Cost.................................................................................... 32 5.2 Cost Competitiveness at Farm Level ........................................................................... 33 5.3 Returns to farmers and poverty alleviation considerations .......................................... 35 5.4 Sector Sustainability .................................................................................................... 39 6. Lessons Learned.................................................................................................................. 41 7. References............................................................................................................................ 43 List of Tables Table 1: Summary chronology of key events in Zambia`s cotton sector, 1977 to 2007 ................ 5 Table 2: Seed cotton production estimates in Zambia .................................................................... 9 Table 3: Cotton production data in Zambia by province, 2003 harvest season ............................ 10 Table 4: Cotton and other indicators by quintile of total cotton production among small holder farmers ........................................................................................................................... 11 Table 5: Linear regression results on cotton yield in Zambia ...................................................... 13 Table 6: Ginning Companies Operating in Zambia as of 2005/06 ............................................... 14 Table 7: Spinning Mill Capacity and Throughput in Zambia, 2002............................................. 17 ii Table 8: Credit allocations and repayment rates under Cotton Outgrower Credit Fund, 2003/4 to 2005/06 .......................................................................................................................... 22 Table 9: Cotton production inputs provided by outgrower companies in 2005/06 growing season, prices charged, and prices at retail charged by input companies ................................... 28 Table 10:Estimated ginning costs in Zambia and seven other countries of SSA (US cents per kg of lint cotton) ................................................................................................................. 32 Table 11: Summary crop budget indicators by farmer type, mean of crop seasons 2004/05 ­ 2006/07 .......................................................................................................................... 34 Table 12: Indicative crop budgets for maize and cotton in Zambia .............................................. 38 List of Figures Figure 1: Seed cotton production in Zambia prior to reform (1987 ­ 1995) .................................. 3 Figure 2: Nominal and real exchange rates between Zambian Kwacha and USD, January 1996 through December 2006................................................................................................... 7 Figure 3: Seed cotton production in Zambia, 1993 - 2007 ............................................................. 9 Figure 4: Map of Zambia showing provincial boundaries and main cotton production zones..... 10 Figure 5: Median cotton yield and share of national production, by quintile of total cotton production among smallholder farmers ......................................................................... 11 Figure 6: Spatial pattern of market purchases of seed and fertilizer, and growing of cotton ....... 18 Figure 7: Credit allocations to cotton companies from Cotton Outgrower Credit Scheme relative to previous year`s cotton area, allocations for 2004/05 ................................................. 21 Figure 8: Structure of Dunavant and Cargill field operations ...................................................... 26 Figure 9: Prices paid to cotton farmers in Tanzania and Zambia, 1995 ­ 2003 (USD/kg) .......... 31 Figure 10:Zambian producer prices and Index A, 1995 - 2006 ..................................................... 31 Figure 11: Returns per day of labor by farmer group and year, compared to average rural wage rate ................................................................................................................................. 36 Figure 12: Total net earnings from cotton by farmer group and year as share of extreme poverty line for family of six ...................................................................................................... 36 Figure 13: Summary regression results from analysis of association between area planted to cotton and household income in Zambia ....................................................................... 37 iii Abbreviations ADMARC Agricultural Development and Marketing Corporation APEP Agricultural Productivity Enhancement Program (Uganda) CAZ Cotton Association of Zambia CCC CHIPATA Cotton Company CDT Cotton Development Trust CFS Crop Forecast Survey CIRAD Centre de Coopération Internationale en Recherche Agronomique pour le Développement COCF Cotton Outgrower Credit Fund CSO Central Statistical Office ECZ Environmental Council of Zambia FFD Farmer Field Demonstration FRA Food Reserve Agency GTZ The Deutsche Gesellschaft für Technische Zusammenarbeit (German Society for Technical Cooperation) HH Household KG Kilogram LF Lead Farmer LINTCO Lint Company of Zambia MACO Ministry of Agriculture and Cooperatives MT Metric ton PHS Post-Harvest Survey SCCI Seed Control and Certification Institute ULV Ultra-low volume YIELD Yield Improvement through Empowerment, Learning, and Discipline ZABS Zambia Bureau of Standards ZKW Zambia Kwacha (Zambian currency) ZNFU Zambia National Farmers Union ZPA Zambia Privatization Agency ZKW Zambian Kwacha iv Executive Summary ince liberalization in 1994, Zambia`s cotton sector has been characterized by persistent S private sector innovation in input credit supply to smallholder farmers and vigorous and sustained growth punctuated, so far, by two relatively brief credit default crises. At the present time, the country is recovering from the second of these crises, which affected it during the 2006 and 2007 seasons. Zambia is an example of a concentrated market-based cotton sector and well illustrates the strengths and weaknesses of such sectors. Led for most of its post-liberalization history by two dominant companies, the sector has continuously expanded the number of smallholder farmers receiving treated seed, pesticides, and foliar feed fertilizers on credit. Yields have risen among experienced farmers, but the continual entry of farmers new to cotton has made overall yield growth modest. Lint quality has improved dramatically, with Zambia now receiving the highest premium in Sub-Saharan Africa on international markets. Credit recovery has typically been good, but has turned suddenly worse during two periods: 1999- 2000, and 2006-07. Each credit default crisis was occasioned by the entry of new cotton buyers, and resulted in sharp declines in production; yet five years after the first crisis, production was double its pre-crisis high and 10 times higher than during the last year prior to reform. Zambia`s cotton sector is now less strongly concentrated than during the first 10 post-reform years, as the last several years have seen the entry of a number of new ginners, some of them with the resources and experience to remain significant players in the sector. Zambia thus illustrates the tendency of concentrated sectors to become more competitively structured over time in the absence of a regulatory regime that preserves the concentrated structure. While this may improve prices to farmers, it carries the risk of undermining input credit, which has been the foundation of the sector`s growth. Private stakeholders are currently working with government to institute a regulatory Cotton Board to define rules of the game that protect input credit even in the presence of more ginners. This regulatory challenge is the linchpin of Zambia`s effort to continue growing the size and productivity of its cotton sector, carries important implications for neighboring countries such as Zimbabwe and even for countries in West Africa, and thus merits close watching by researchers and policy makers interested in the future of cotton in Africa. v 1. Introduction Cotton is an unquestioned success of Zambia`s shift towards a market economy. After liberalization in late 1994, production rose from 20,000 MT to over 100,000 MT by the 1998 harvest. Production fell again to less than 50,000 MT in 2000, but rose steadily through the first half of the decade and hit nearly 200,000 MT in 2005. From 2002 through 2005, exports of cotton lint were first among all agricultural exports in value, 30 percent higher than any other agricultural export (Export Board of Zambia, 2006). Raw cane sugar, the closest competitor to cotton during this time, is primarily produced on large operations, while cotton is almost entirely a smallholder crop. Its potential role in poverty alleviation and food security is thus very large. The success of this sector has been achieved despite persistent declines in international cotton prices since 1995, serious problems of credit default during the late 1990s, the departure of the sector`s biggest company, Lonrho, in 1999 and a very recent crisis brought on by the appreciation of the Zambian Kwacha during 2006. The Zambian cotton story is of interest for policy makers for four reasons. First, the sector`s boom since 2000 has been driven entirely by private sector innovation. Understanding the details of that innovation and how it helped the cotton sector flourish can provide insights for other countries. Second, as one of the only remaining sectors that is both fully private and highly concentrated1, the Zambian cotton industry faces special challenges maintaining a balance between the coordination needed to ensure rising quality and productivity, and the effective competition typically needed to protect the interests of smallholder farmers; assessing the ways in which both government and existing firms have gone about this may also yield useful insights. Third, these challenges have been thrown into high relief during 2006 and 2007, as the entry of several new and potentially strong companies combined with the appreciation of the Kwacha may substantially alter the competitive dynamic in the sector; over the next several years, Zambia may thus be a second case study (following Zimbabwe) in how a concentrated sector responds to the challenges of new entry. Finally, government policy initiatives since 2002 which have had a decidedly mixed effect on the sector, have evolved since their last review (Tschirley and Zulu, 2004), and merit a further examination. This paper is part of a continent-wide review of cotton sector reform experience in Africa. It grows out of earlier work on cotton in Zambia by the Food Security Research Project (Govereh et al. 2000; Tschirley and Zulu, 2003; Tschirley and Zulu 2004), along with collaborative work by Michigan State University in Southern and Eastern Africa with Imperial College and colleagues in Tanzania, Zimbabwe, and Mozambique (Poulton et al 2004; Tschirley et al 2006). The paper has five main purposes: To capture the key elements and sequencing of the sector`s reform process since 1994, and suggest how these might be functionally linked to previous structural and performance characteristics of the sector; 1 A concentrated, market-based system in the classification scheme of Poulton, et al. Zimbabwe was classified in this way in the early 2000s, but has since seen the entrance of numerous smaller companies. 1 To provide an updated, detailed descriptive overview of the current organization of the sector and of the behavior of key public and private participants; To assess the sector`s current performance in multiple dimensions, including the institutional approaches used to pursue desired performance, and the extent to which chosen approaches can be linked to structural characteristics of the sector; To provide an updated critical evaluation of recent policy initiatives in the sector and suggest key modifications that might be needed; To assess the cost competitiveness of the sector at farm and ginning levels; and To identify very recent structural changes in the sector and highlight the key steps the sector needs to take to ensure its future competitiveness in regional and international markets. The next five chapters deal with each of these objectives: Chapter 2 presents historical background and reviews the reform process of the mid-1990s, while Chapter 3 provides a current overview of the sector. Chapter 4 then describes in greater detail current institutional arrangements in the sector and related performance. Chapter 5 presents farm and ginnery budgets, uses them to assess returns to farmers and competitiveness at farm and ginning levels, and discusses sector sustainability. Chapter 6 closes by identifying lessons learned. 2. Historical Background 2.1 Pre-Reform Institutional Set-up and Performance From 1977 to 1994 the state-owned cotton company LINTCO (Lint Company of Zambia), on behalf of the government, purchased seed cotton from farmers at a fixed price, provided certified seed, pesticides, sprayers, and bags and provided extension advice to farmers. LINTCO had a near monopsony in buying seed cotton and a monopoly in distributing cotton inputs on credit. Very little empirical information is available for this period, though some insight into the performance of the cotton sector during LINTCO`s tenure can be inferred from the only available data from that period: the annual crop forecast surveys conducted by the Government of Zambia`s (GoZ) Central Statistical Unit (Figure 1). The data suggest that, from 1987 to the year immediately following liberalization (1995), production was low, fluctuating, and in secular decline, falling below 20,000 MT of seed cotton in the 1995 harvest year. According to Zambia Privatization Agency (ZPA), LINTCO was also in serious financial crisis before its sale, having accumulated substantial unpaid debts. In 1994, as part of a concerted and broad-based effort to restructure Zambia`s economy by the new government of Frederick Chiluba, LINTCO was sold to Lonrho Cotton and Clark Cotton, two private companies with regional cotton interests. The sale appears to have been designed explicitly to limit competition between the companies, as Lintco`s gins in the center of the country were sold to Lonrho, and those in Eastern province were sold to Clark. 2 Figure 1: Seed cotton production in Zambia prior to reform (1987 ­ 1995) 2.2 Reform through 2003: Phases One to Three Zambia`s cotton sector since reform has passed through five overlapping phases: in phase one (1994-1997), LINTCO was sold and the sector expanded rapidly on an entirely private and unregulated basis; phase two (1998-2000) was marked by a severe credit default crisis, which was resolved in phase three (2000-2003) entirely through private sector innovation; in phase four (2002-2005) the government became increasingly involved in the sector, with mixed results. Phase five (2005-present) has seen the entry of four significant new players at the ginning level and a recurrence of the credit default crisis of the late 1990s. Table 1 presents a summary chronology of key events in the sector. We discuss phases one to three here, leaving the more recent events of phase four and five to Chapter 4. Competition between Lonrho and Clark was minimal, as they operated in different areas of the country. Each company initiated outgrower programs and had very little problem with credit repayment. From 1994 through 1998, cotton production increased by a factor of three to four, depending on the data source, facilitated by high international prices and aggressive promotion of the crop by Lonrho and Clark. In 1997 the rapidly expanding cotton production base attracted new entrants, both in ginning and assembly. At least four new ginning companies emerged and began to compete aggressively in the purchase of cotton. Beyond the farmers directly recruited by these new entrants some were recruited indirectly through agents contracted by ginner-contracted agents. There also emerged a group of independent cotton traders who obtained their own inputs, distributed them to farmers, purchased seed cotton and sold to any ginner wishing to purchase. At the time, GoZ was committed to a liberalized economic policy and made no attempt to limit this competition. As the number of ginners and assemblers expanded, several key problems came to the fore. First, ginning capacity expanded to over 150,000 MT per annum, while production peaked at about 105,000 MT in 1998 and then declined for two years. The overcapacity created a competitive scramble for cotton among ginners to increase their throughput and minimize unit ginning costs. The emergence of agents and independent traders contributed substantially to this scramble. Firms operating outgrower schemes 3 experienced increased loan default rates as competing firms, some of which did not provide input credit and hence could offer higher prices, purchased cotton from farmers participating in other firms` outgrower programs. These problems were exacerbated by a continual decline in world market prices from their peak in 1995, which was passed on to farmers. Farmers had grown accustomed to increasing prices, and with limited information on world market conditions, they found it difficult to understand the reasons for the price declines. Together with a lack of transparency in how each buyer determined its prices and deducted input costs, led many farmers and their representatives to conclude that they were being exploited. Lonrho estimated that loan repayment rates dropped more than 20% in three years. The increasing default rates created incentives for outgrower firms to capitalize their bad loans into the cost of inputs for those farmers who did repay,2 resulting in still lower net prices for cotton after deducting the cost of inputs. Farmers who remained loyal and repaid their loans were thus penalized, fueling a vicious cycle of further loan defaults or exit from participation in outgrower programs. 2 One outgrower company stated that in 1999 it attempted to offset its loan defaults by adding a 50 percent mark up to the price of inputs. 4 Table 1: Summary chronology of key events in Zambia's cotton sector, 1977 to 2007 Year Key Event Comments Pre-Reform 1977-94 State-owned LINTCO runs single Production trends downward from mid-1980s in spite of rising channel cotton system. international prices. Public debt accumulates. Phase One 1994 LINTCO sold to two private companies: Two companies operate for two years in separate areas of Lonrho and Clark Cotton. country. Production booms, aided by high international prices. Phase Two 1997-99 Four new ginning companies enter Combined Dunavant and Clark market shares fall to 80 percent. market, group of independent traders Competition for seed cotton increases. Charges that new also emerges. Government does not entrants provide few if any inputs to farmers. Credit recovery intervene. falls below 60 percent during 1997/98 season. 1999 Lonrho, citing input credit losses of Lonrho launched Distributor Scheme, Dunavant (under same USD $2m, leaves Zambia. Assets management) continues to develop it. Credit recovery over 60 purchased by private Co. Dunavant. percent. Phase Three 2000-2001 Dunavant fully develops its private Credit recovery improves to 85 percent. At least one of the Distributor Scheme. recent entrants falters but does not leave market. 2001/2 Drought in southern areas of country. Indications that credit recovery rate decreased. Phase Four 2002 New government enters late 2001, Publicly funded credit line for input provision, being developed launches Cotton Outgrower Credit Fund. in close collaboration with ginners. First direct government involvement in the sector since liberalization in 1994. 2003/04 Cotton Act proposes new Cotton Board. Regulatory functions only. Heavy policing role. 2005 Cotton Association of Zambia formed. Represents about 300,000 cotton farmers, nearly all smallholders, under aegis of Zambia National Farmers` Union. 2003/04, Cotton Outgrower Credit Fund expands, Funds increased to USD $390,000 from USD $250,000 in the 2004/05 becomes revolving fund. first year. Disbursements favor smaller players in relative sense. Recoveries suggest effective management. December President signs Cotton Act. Stakeholders were surprised by the signing, as they had raised 2005 concerns and understood that the President would delay signature. Phase Five Late 2005 Kwacha begins rapid appreciation. Appreciates 33 percent through start of cotton marketing season, then begins to decline. Late 2005 ­ Broad group of stakeholders ­ ginners, Run-up to elections prevents submission to Parliament in June late 2006 farmers, Minag, CDT ­ collaborate to 2006; Act to be presented to new government November 2006. develop proposed revisions to Cotton Act. May 2006 Cargill Cotton buys Clark Cotton. Former parent company Afgri cites low profit margins and insufficient global reach in marketing. Harvest Key beneficiaries of Cotton Outgrower Credit default returns to levels not seen since 1999. Dunavant season 2006 Credit Scheme continue to be suspected operations outside Eastern Province most affected. Cargill of promoting credit default by farmers. claims repayment rates in Eastern of more than 90 percent. March 2007 Three multi-national affiliates Meant to encourage cooperation among these companies to (Dunavant, Cargill, Great Lakes) open reduce credit default during marketing season; group invites discussions on Zambia Cotton Pre- CCC also to participate. Financiers` Association. 5 The sector reached a crisis point in 1999. Lonrho, the largest buyer, was sold to Dunavant, a privately held U.S. cotton company. Among its reasons for departing, the company cited USD $2 million per year in unpaid loans. Other outgrower firms cut back on the number of farmers they supported from the 1999/2000 season, driving production to a post-reform low of less than 50,000 MT. From this nadir, the sector underwent important structural change and recovered dramatically through the 2005 harvest season. The agents and independent buyers responsible for the credit repayment problems in the late 1990s largely disappeared. At least one of the new ginners went out of business in late 2002. These developments were associated with two parallel strategies adopted by Dunavant. First, in 1999, the US-based cotton company launched a Distributor System, which was refined over the next several years and dramatically improved credit repayment rates among farmers. Second, Dunavant used this system to aggressively expand its production network. Clark also improved its more traditional system and was able to dramatically expand its production in Eastern province while maintaining high repayment rates. Partly as a result, national production more than quadrupled between 2000 and 2005, driven by yield growth in addition to area expansion. Credit repayment for Dunavant also improved from about 65 percent to over 90 percent. Both companies largely resolved, through very different approaches, a problem of polypropylene contamination which had threatened the country`s export market, and the country began to receive a premium on world markets.3 Finally, despite operating in a much more concentrated sector, companies in Zambia paid prices nearly as high as in Tanzania. 3. Overview of the Cotton Sector 3.1 Macroeconomic Environment Movements in the real exchange rate have had an important influence on the cotton sector in Zambia over the past decade. Figure 2 shows movements in the real exchange rate between the Kwacha and the U.S. dollar from January 1996 through December 2006 4. From 1996 through 2001, the Kwacha slowly depreciated in real terms against the dollar. As a result, export sectors with a significant share of costs in local currency would have been able to earn slightly higher profits had everything else been equal. Since 2002, however, the Kwacha has steadily appreciated against the dollar. Until late 2005, this pattern may have been broadly consistent with the general decline of the U.S. dollar in international currency markets. Since that time, however, the appreciation of the Kwacha has proceeded much more rapidly and has been related to factors internal to the country. In either case, export sectors have been placed under increasing pressure by the Kwacha since 2002; and since late 2005, the currency`s rapid appreciation has caused a serious crisis in all export sectors. The slight recovery in the real rate in late 2006 left it still well below typical levels from 1996 through 2002. 3 We discuss the two companies` different approaches in Chapter 4. 4 The graph takes a purchasing power parity approach. With calendar year 1996 as the base, we calculate movements in the ZKW/USD exchange rate that would have maintained the purchasing power of the Kwacha relative to the USD. Purchasing power is based on relative movements in the Consumer Price Index in each country. A value above 100 indicates that the Kwacha had depreciated in real terms compared to 1996, while a value below 100 indicates real appreciation. Note that this approach traces out a pattern of appreciation nearly identical to that calculated by a Real Effective Exchange Rate (REER) approach incorporating prices of tradables and non-tradables, and weighting by the structure of trade (Calí and te Velde, 2007). 6 Figure 2: Nominal and real exchange rates between Zambian Kwacha and USD, January 1996 through December 2006 Fynn and Haggblade (2006) estimate that, if ginning companies were earning an 8 percent gross margin (revenue minus variable costs) at an exchange rate of ZKW4500/USD (the rate prior to the dramatic appreciation), that margin nears zero at an exchange rate of 3,500, while net profit (gross margin minus capital costs) is clearly negative at this level. The timing of the 2005/06 appreciation of the Kwacha was especially bad for agricultural export sectors. Outgrower companies had to purchase most inputs for the 2005/06 growing season no later than June or July of 2005, when the exchange rate was around ZKW4,700/USD. They based their input prices to farmers on their costs at that exchange rate. The Kwacha then began to appreciate in August, and by May 2006 was down to ZKW3,200/USD, and some in government were indicating that they were committed to continued appreciation. In this environment, Dunavant indicated that it mobilized the local currency it needed for the 2006 harvest at the low exchange rates prevailing in May, and announced that it would pay only ZKW860/kg, down from ZKW1,200/kg the previous year, and from the ZKW1,220/kg that it had indicated it would pay prior to planting in late 2005, Though it indicated that this offer was contingent on the exchange rate remaining above ZKW4,200/USD, this condition received little focus when the company did drop its price. Interviews with ginning companies and farmers indicate that smallholder cotton plantings for the 2006/07 growing season fell by 40 percent to as much as 50 percent. Dunavant announced that it would pay no less than ZKW1,050/kg of seed cotton, regardless of prevailing exchange rates, but farmer confidence in this price had been shaken by events the previous year. Other companies declined to announce pre-planting prices. By early 2007, the exchange rate had stabilized around ZKW4,200-4,300/USD, a level which should allow Dunavant to pay at least its guaranteed minimum prices, with comparable prices paid by competitors. In Dunavant`s own words, a key challenge the company now faces is to rebuild confidence in (the) pre-planting price. 7 3.2 Cotton Production: Trends, Geographical Distribution, and Farm Structure Production data in Zambia are available from several sources, including the Central Statistical Office`s (CSO) Crop Forecast Survey (CFS), CSO`s Post-Harvest Survey (PHS), and derived estimates from ginnery outturn. These estimates do not all agree with each other, but, with the exception of CFS data for 1999, paint a relatively consistent picture of production trends since liberalization (Table 2). Since reform began in 1994 with the break- up of LINTCO, the monopoly cotton parastatal, production has gone through four phases (Figure 3)5: 1) rapid expansion through 1998, with production increasing from less than 20,000 MT in 1995 to over 100,000 MT in 1998, 2) rapid decline to less than 50,000 MT in 1999 and 2000, spurred in large measure by a serious credit default crisis; 3) sustained and rapid recovery from 2000 to 2006, and 4) a sharp forecasted decline in 2007, driven by the Kwacha appreciation crisis of the previous year. Household survey data show that cotton production is heavily concentrated in Eastern province, with over one-third of all households in that province producing the crop and accounting for approximately two-thirds of national production during the 2003 harvest season. Central and Southern provinces follow, with 16 percent of farmers growing the crop in Central and accounting for 19 percent of national production, and 12 percent growing in Southern and accounting for 13 percent of national production (see Table 3 and Figure 4). Data from a similar survey for the 2000 harvest season show a comparable dominance of Eastern province, followed again by Central and Southern, but suggest that Southern Province`s national share, at 5 percent, was much lower that year than in 2003. Nationally, nearly 11 percent of all farmers grew the crop in 2003; by 2006, that figure had likely risen near 15 percent. 5 Data in Figure 3 are based on CFS estimates for 1993 through 1995, and on derived ginnery or CDT estimates since that time. All production figures are in terms of seed cotton. 8 Table 2: Seed cotton production estimates in Zambia Harvest PHS Estimates CFS Derived Ginnery Estimates2 Year1 Estimates # of house- Area (ha) Production # of households Production holds (MT) Min Max 1993 32,944 32,343 23,103 47,851 1994 30,764 28,669 18,384 33,093 1995 32,824 28,450 27,991 16,578 1996 50,981 64,084 63,859 40,824 113,333 170,000 61,200 1997 85,514 74,279 58,051 70,000 142,217 213,325 79,900 1998 85,735 79,272 72,561 110,000 179,050 268,575 104,500 1999 70,159 63,000 50,858 140,024 139,895 209,842 84,700 2000 44,196 36,681 27,500 49,498 74,449 111,674 46,700 2001 87,422 87,026 65,979 57,083 110,924 166,387 72,000 2002 47,326 172,900 259,350 116,000 2003 170,341 255,512 118,000 2004 240,712 361,069 172,000 2005 266,173 399,259 196,000 2006 244,005 366,007 185,000 2007 (Est.) 142,308 213,462 111,000 Various sources, 1993-2007 1 Harvest year refers to cotton planted late in the previous year. 2 Seed cotton production estimates through 2000 derived from lint production figures of Lonrho, Clark, and Amaka, and based on ginning outturn ration (GOR) of 0.38. Production figures since that time are based on CDT estimates, which use data from all companies. Estimate of minimum (maximum) number of households assumes average of 0.8 ha (1.2 ha) cotton per farmer, with yields increasing from 450 kg/ha in 1996 to 650 kg/ha in 2007. Figure 3: Seed cotton production in Zambia, 1993 - 2007 9 Table 3: Cotton production data in Zambia by province, 2003 harvest season Share of HHs Share of all Share of in province cotton Total cotton national producing # of cotton farming HHs production cotton Province cotton farming HHs nationally (MT) production Central 0.16 22,155 0.17 23,754 0.19 Copperbelt 0.00 127 0.00 254 0.00 Eastern 0.36 89,773 0.68 79,702 0.65 Luapula 0.00 0 0.00 0 0.00 Lusaka 0.07 2,522 0.02 2,082 0.02 Northern 0.00 0 0.00 0 0.00 NorthWestern 0.00 0 0.00 0 0.00 Southern 0.12 17,778 0.13 16,484 0.13 Western 0.00 0 0.00 0 0.00 Total 0.11 132,355 1.00 122,276 1.00 Source: PHS/FSRP Supplemental Survey Figure 4: Map of Zambia showing provincial boundaries and main cotton production zones Zambian cotton is produced almost entirely by smallscale farmers in Zambia. Among the 11 percent of farmers that grew the crop in 2003, over half of production and sales were accounted for by the largest 20 percent of farmers (Table 4 and Figure 5). These concentration levels are not high compared to a crop like maize, where the top 20 percent of producers account for nearly two-thirds of all production and a much larger share of sales. Of course, because nearly 90 percent of farmers do not produce cotton, the top 20 percent of cotton farmers represent only 2-3 percent of all farmers. These large cotton farmers cultivate 10 more total land, dedicate more of it to cotton, and achieve higher cotton yields.6 We will compare these same groups of cotton farmers to non-cotton farmers in Chapter 4 when we examine the effects of cotton farming on household incomes. Table 4: Cotton and other indicators by quintile of total cotton production among small holder farmers Cotton Indicators Other Indicators Median Median Share of Median Quintile of Median Cotton Cotton all Share of all Median Total Maize Cotton Ha in Yield Production Cotton Cotton Land Production Production Cotton (MT/ha) (MT) Area Production Cultivated (MT) 1 0.41 0.51 0.249 0.11 0.05 1.25 0.70 2 0.41 0.86 0.420 0.14 0.09 1.62 0.86 3 0.61 1.05 0.600 0.17 0.13 1.75 1.15 4 0.81 1.11 0.900 0.22 0.21 2.28 1.38 5 1.22 1.48 1.700 0.37 0.52 3.09 2.07 Total 0.81 0.96 0.600 1.00 1.00 1.92 1.15 Source: PHS/FSRP Supplemental Survey, 2004) Figure 5: Median cotton yield and share of national production, by quintile of total cotton production among smallholder farmers Source: PHS/FSRP Supplemental Survey, 2004 Yield calculations in Zambia are based primarily on seed cotton purchased by Dunavant and Cargill. These figures are thus affected by side-selling; since Dunavant and Cargill most likely lose more cotton from side-selling than they gain, the figures should be interpreted as a lower bound for the yields obtained by farmers financed by these two companies. With these caveats in mind, yields appear to have risen since reform, due to the persistent efforts of these two companies to ensure steady input supply and some level of extension training for farmers. Dunavant has suggested that their mean yields have risen from about 450 kg/ha in 1996 to nearly 700 kg/ha by 2005. With support from GTZ, the company is now attempting to increase yields further through its YIELD program. 6 Note that yield figures based on household surveys consistently generate higher mean and median yield estimates than those indicated by cotton outgrower companies. We estimate that the yields in Table 4 overstate actual yields by 20 to 30 percent, but we have no reason to believe that the relative patterns in the table are biased. 11 Data from Cargill may be relatively reliable in the sense that the company reports high repayment rates, so that yields calculated as total purchases divided by total hectares financed may be more accurate than for Dunavant, which gets lower repayment rates. These data show mean yields of 974 kg/ha in 2004, 695 kg/ha in 2005 (a drought year), and 795 kg/ha in 2006. These yields are consistent with previous understandings that yields in Eastern province (the only province in which Cargill operates) averaged about 900 kg/ha. Such yields are substantially better than those in neighbouring Mozambique or Tanzania (each with mean yields closer to 500 kg/ha), and comparable to Zimbabwe when Cottco was still operating its outgrower scheme. Regression results (Table 5) provide some insights on yield drivers. As found in similar research on cotton in Mozambique (Benfica et al, 2006), and more generally on agriculture throughout Africa (Huffman, 1980; Yang, 1997), there appear to be no returns to education in cotton cultivation. Households with more family labor achieve higher yields, as shown by the pattern of coefficients on household size and number of children. Use of basins/zero tillage and inorganic fertilizer have significant positive effects on yield, as does the use of manure; yet none of these practices are applied by more than 5 percent of farmers. Having your own animal traction teams and rotating cotton with maize are common practices that have important and significant positive effects on yield. The importance of animal traction in this regression is consistent with both PRA results and simpler regression analysis using a different data set (see Appendix B). Given the widespread use of inorganic fertilizer on maize, but not on cotton,7 the positive effect of rotation with maize probably reflects continuing positive impacts from previous fertilizer applications; this dynamic between the two crops was frequently referred to by Cargill field personnel during the PRA. 3.3 Cotton Ginning The structure of cotton ginning in Zambia changed substantially from 2004 through 2006. Eight ginning companies operated in Zambia during the 2005/06 growing season (Table 6) 8. Of these, three began operations during that season (Great Lakes, Alliance, and Birchand) and a third began in 2004/05 (CCC). Between them, these four companies purchased more than 30,000 MTs of seed cotton in 2005/06. Great Lakes entered Southern Province in mid- 2005 and competed directly with Dunavant, even hiring a number of Dunavant Distributors for its input distribution program. Alliance also competed directly with Dunavant in Central Province, while CCC operated in areas of Eastern Province where both Cargill and Dunavant have strong field operations. Continental ginneries, which had operated in the country for several years, opened a new gin in Eastern Province (the main producing province) during 2005/06 while continuing to operate its gin in Southern Province. 7 Only 3 percent of cotton fields received inorganic fertilizer in our sample. 8 We drop Mukuba from the count due to their tiny purchases. 12 Table 5: Linear regression results on cotton yield in Zambia Coefficient Std. Error P value Constant 6.08 0.190 0.000 * Demographics HH is headed by a female (0,1) 0.12 0.083 0.156 Years of education of household (hh) head 0.00 0.008 0.876 # of children in hh -0.03 0.019 0.117 Total size of hh 0.02 0.012 0.103 * Field level practices Farmer plowed cotton field (0,1) 0.03 0.081 0.746 Farmer used basins/zero tillage (0,1) 0.40 0.154 0.009 *** Farmer used ridging (0,1) 0.03 0.086 0.714 Farmer used tractor to prepare field (0,1) 0.13 0.180 0.467 Farmer has own animal traction teams (0,1) 0.15 0.072 0.036 ** Farmer planted before the rains (0,1) 0.01 0.060 0.909 Weeks from planting to first weeding 0.01 0.024 0.710 Total number of weedings 0.02 0.028 0.540 Farmer used manure on field 0.27 0.149 0.076 * Kg of basal fertilizer applied to field 0.01 0.002 0.016 ** Previous crop was maize (0,1) 0.18 0.077 0.022 ** Previous crop was groundnut (0,1) 0.00 0.103 0.992 Previous crop was cotton (0,1) 0.13 0.105 0.229 Field was previously in fallow (0,1) -0.28 0.350 0.431 Farmer left residue on field (0,1) 0.02 0.056 0.759 Animals fed in field (0,1) -0.06 0.109 0.582 Other Agricultural/Agro-ecological Factors Log total maize production 0.05 0.015 0.000 *** Log value of productive assets 0.02 0.004 0.000 *** Log total cultivated ha -0.25 0.048 0.000 *** Zone 2 (Lower Rainfall) (0,1) -0.03 0.086 0.707 Zone 3 (Higher Rainfall) (0,1) 0.19 0.129 0.134 Zone 4 (Highest Rainfall) (0,1) 0.19 0.649 0.765 Central Province (0,1) 0.00 0.123 0.973 Eastern Province (0,1) 0.11 0.111 0.309 Lusaka Province (0,1) 0.16 0.169 0.345 R-Squared 0.139 Adj. R-squared 0.096 N 611 Dependent variable = log cotton yield (kg/ha) * sig at 0.10; ** sig at 0.05; *** sig at 0.01 Source: Author's calculations from CSO/FSRP 2004 Supplemental Survey 13 Table 6: Ginning Companies Operating in Zambia as of 2005/06 Company Ownership Gin Location Capacity Seed Cotton Throughput (MT/ season) 03/04 04/05 05/06 Dunavant Multi-national Lusaka, Lusaka Province 10,000 Kabwe, Central Province 22,000 Mumbwa, Central Province 25,000 Gwembe, Southern Province 19,000 Petauke, Eastern Province 17,000 Lundazi, Eastern Province ? Katete, Eastern Province 22,000 Sub-total > 115,000 112,5001 131,3001 112,0001 Cargill Multi-national Three gins in Chipata, Eastern Prov. 60,000 48,9761 44,1961 42,0231 Great Multi-national One gin in Sinazongwe, Southern Prov. 10,0002 0 0 10,0002 Lakes (Plexus) Alliance Multi-national Lusaka Province (planned) ? 0 0 8,0002 Cotton (Alliance Cotton) Continental Local Sinda, Eastern Prov. 15,0002 5,0001 7,0001 8,0001 Kalomo, Southern Prov. 10,0002 Mulungushi Zambian/Chinese Kabwe, Central Province 10,000 5,8201 8,3141 5,1401 Chipata- Chinese Chipata, Eastern Province 15,000 0 ? 12,0002 China Cotton Ginnery (CCC) Mukuba Local Ndola, Copperbelt Province 500 43 113 24 Birchand Tanzanian Tanzania 0 0 0 ? Oil Mills Total > 215,500 1 2 Notes: ?=data not available. Self-reported; Rough estimates from discussions with stakeholders Source: Zambian Cotton Sector Review, Ministry of Agriculture, Food and Fisheries 2000; FSRP Ginners Survey, 2003; additional FSRP interviews 2006 An array of information suggests that these second-tier competitors to Dunavant and Cargill are poised to increase their purchases over the coming years. First, CCC is currently installing a second gin with at least 15,000 MT capacity in Petauke district of Eastern province. Second, following the visit of President Hu of China, an accord was signed for the construction of 3-5 new ginneries in the country with Chinese financing. Indications are that most of these gins will be installed in Eastern Province. Third, Great Lakes and Alliance are 14 both local affiliates of multi-national cotton companies that have been expanding operations in southern and eastern Africa in recent years. Finally, Continental more than doubled its ginning capacity in 2005/06, but used barely one-third of this higher capacity. The entrance of so many new and aggressive buyers in the Zambian cotton market has major implications for sector governance, which we will address in Chapter 6. Dunavant and Cargill are both very large multinational cotton trading companies. Dunavant Enterprises trades over 800,000 MT of cotton lint per year worldwide, and claims to be the largest privately owned cotton merchandiser in the world. In addition to Zambia, it owns cotton gins in Mozambique (in Tete province, directly across the border from Eastern Province in Zambia) and Uganda. Cargill Cotton purchased Clark`s Cotton`s operations throughout Southern Africa in May 2006: three gins in Zambia, three in South Africa, and a majority interest (with ADMARC) of two in Malawi. According to Afgri management (the parent company of Clark Cotton), key reasons for the sale were that the cotton enterprise did not deliver sufficient return on capital, and Clark Cotton did not have the expertise or the critical mass to effectively market its cotton lint in the international market, whereas Cargill did (Business Day, 2006). Cargill also operates in Zimbabwe and Tanzania. Great Lakes is a joint venture between Plexus Cotton Limited and Africa Resources Holdings Limited, with cotton gins in Uganda, Malawi, and Zimbabwe. Plexus itself owns a gin in Mozambique and in recent years has emerged as one of the largest ginners in that country. Mulungushi Textiles is a joint venture between the Government of Zambia and the Government of China. This unusual arrangement in an otherwise entirely privatized industry has caused unease among competing private companies, some of whom suggest that the arrangement might confer competitive advantages to Mulungushi, especially in the area of working and investment capital, that these other firms do not have. There is, however, currently no concrete evidence of these and other possible advantages conferred on Mulungushi. Continental Ginners and Mukuba Textiles are both locally owned firms, while Chipata Cotton Ginners appears to be primarily financed with Chinese capital. The operating practices of these firms, especially regarding input supply on credit, will be discussed in more detail in Chapter 4. Here we note that Dunavant and Cargill both run input distribution, extension, and seed multiplication programs that are recognized as serious efforts to build capacity and productivity among their farmers. Both companies also made major successful efforts to eliminate polypropylene contamination in cotton. Great Lakes emphasizes productivity and quality in its promotional materials, and appears likely to expand its pre-financing activities during the 2007/08 season. Many more questions exist regarding the design, coverage, and consistency of the input credit and extension programs of the other companies in the sector. 3.4 Independent Cotton Traders Independent cotton traders, individuals trading cotton who do not own and are not employed by a ginning company, played a major role in the credit default crisis of the late 1990s, but after 2000, had largely disappeared. After launching its Distributor System for input delivery and credit recovery in 1999, Dunavant (Lonrho at the time the system was launched) created strong incentives for distributors to remain loyal to the company. Clark Cotton (which has 15 since been taken over by Cargill Cotton) also began paying stronger attention to its relations with farmers, providing an input package that resulted in steady rises in productivity, and also maintaining detailed credit repayment data for each farmer--those who failed to repay debts were summarily removed from the list of those eligible for input credit. Finally, the very high international cotton prices that prevailed for several years after the reform of the sector, making cotton trading a potentially attractive business, had fallen to historically low levels by late 1999 and, after a brief recovery, reached even lower levels by late 2001. For a company to remain a major player in the cotton business at these prices required a long-term commitment and, increasingly, global reach in marketing. The largely locally owned firms that supported independent cotton traders (Amaka, which left the sector in 2002, and Continental and Mulungushi, which have remained) found it very difficult to compete under these circumstances. Nonetheless, Zambia clearly has a set of actors with extensive experience in the cotton trade and no investment in cotton ginning. Dunavant`s distributors, which at one point may have numbered 2,000, are independent entrepreneurs experienced in recruiting farmers, delivering inputs, and mobilizing crops. Whether these actors are called agents (Mulungushi and Chipata Cotton Ginners) or contact farmers (Continental), other companies run similar programs. In all cases, the cotton ginner relates to farmers primarily or even exclusively through its distributors or agents, who make the final determination as to the credit worthiness of farmers. During the 2006 harvest season, events surrounding the macroeconomic environment (discussed in section 2.a) led to credit default becoming a major issue in the sector for the first time since 2000. It appears that Dunavant suffered much more heavily from default than did Cargill; the latter claims repayment rates of 92 percent during 2006 (low by their standards), while Dunavant claims only 60 to 70 percent and openly acknowledges that Cargill achieves higher repayment rates even during normal years. 3.5 Cotton Spinning/Weaving/Apparel Industries Zambia`s spinning industry appears to absorb a small and declining share of the country`s lint production. The last available data indicate that the country`s four operating spinning mills processed less than 10,000 MT of lint in 2002, or less than one quarter of lint production in the country (Table 7). Export value of yarn fell from about USD $40million in 1997/98 to USD $21million in 2001/2002, and remained at about that level through 2005. As cotton production has increased by about 70 percent since 2002, the spinning industry`s share seems likely to have declined. During the 2001/02 season, Dunavant indicates that it sold nearly 20 percent of its lint in the local market, exporting the rest. The cotton ginners and Swarp (a spinner) estimated in 2002 that 90 percent of Swarp`s lint needs are met by purchases from Dunavant and Clark (now Cargill); the balance appears to come from smaller ginners. Mukuba Textiles and Mulungushi Textiles both have gins within their premises and purchase seed cotton for processing. Starflex, Excel, Mulungushi and Kafue all experienced serious financial problems in the early 2000s which led to temporary and sometimes prolonged shut downs (ZCSR, 2000; RATES 2003). The other smaller spinners indicate that they periodically import to meet their lint needs when they are unable to reach agreement on price with local ginners. 16 Table 7: Spinning Mill Capacity and Throughput in Zambia, 2002 Textile Mill Location Capacity Throughput, 2002 (MT) (MT)1 Swarp Ndola, Copperbelt Province 14,000 6.400 Mukuba Ndola, Copperbelt Province 1,900 1,200 Starflex Ndola, Copperbelt Province 1,200 Not operational Excel Ndola, Copperbelt Province 1,650 500 Mulungushi Kabwe, Central Province 3,000 1,500 Kafue Kafue 3,000 Not operational Others Mostly Copperbelt Province 1,000 N/A Total 25,750 1 Sources: Data on Swarp from phone interview with that company and RATES (2003). Other data based on estimates by Swarp, Ministry of Agriculture, Food and Fisheries 2000, and RATES (2003). Despite the problems that these value-added sectors have faced, their combined size is not trivial when compared to cotton lint: total exports of yarn, woven fabric, and apparel totalled USD $23.5 million in 2002 (over USD $21 million from yarn), compared to USD $30 million in lint exports. More updated information on the spinning, weaving, and apparel manufacture industries in the country would thus appear to be warranted. 3.6 Oil and Cake Ginners provide very little information regarding their sales of oil and cake. Key informants indicate that most seed is sold to South Africa, with some remaining in the country and the rest frequently going to Botswana, due to that country`s large livestock sector. CCC processes about one-third of its seeds in its own oil processing plant, and exports the rest. Estur estimates a net average price of USD $90/kg, based on prices in South Africa and assumed sales of 75 percent of seed. Local ginners claim prices of USD $60-90/kg. 3.7 Independent Input Dealers Private input dealers in Zambia have grown up primarily serving maize farmers, selling fertilizer and maize seed. For example, during the 2003/04 cropping season, 35 percent of farmers used fertilizer, over half of these (18 percent of all farming households) purchased the fertilizer from a private input dealer, and over 80 percent of all fertilizer transactions (through private dealers or other programs) were for maize. Similarly, 35 percent of all households purchased seed from a private input dealer or seed company, and 59 percent of all these market transactions were for maize seed.9 9 All data from the 2004 MACO/FSRP Supplemental Survey. Differences in data collection for the two inputs prevent us from presenting identically structured analyses. 17 Figure 6: Spatial pattern of market purchases of seed and fertilizer, and growing of cotton Source: 2004 CSO/FSRP Supplemental Survey The spatial pattern of market purchases of seed and fertilizer compared to that of growing cotton shows that private input markets are relatively less developed where cotton is most commonly grown ­ in Eastern Province (Figure 6). As a result, private input dealers play very little direct role in providing cotton inputs to farmers. Nearly all cotton inputs in Zambia are delivered to farmers through the cotton ginning companies or through ginners` agents who receive the inputs from the cotton companies. The cotton companies negotiate for inputs in bulk from local and international companies. With regard to seed, all companies interviewed reported that they grow their own seed through contract farmers and the seed is certified by the Seed Control and Certification Institute (SCCI), the government`s certification unit under the Ministry of Agriculture and Cooperatives (MACO). 4. Current Institutional Arrangements and Performance For eight years after reform in 1994, Zambia`s cotton sector operated as a concentrated, market-based system with almost no government involvement, even on a regulatory basis. Extra-market coordination, whether across ginning firms or between ginners, organized farmers, and other stakeholders, was minimal. Since that time, government has developed a more noticeable presence in the sector, and efforts at sector-wide coordination have increased markedly. Most recently, the dominance of the top two ginning companies has become less pronounced, and the new competitors may have a greater ability to remain in the market than did those who affected the sector so strongly in the late 1990s. In this chapter we review the current organization and performance of the sector across a number of tasks: extra-market coordination, input credit, research and extension, quality control, pricing of seed cotton, and competition among companies. 18 4.1 Sector Coordination Through the 2005 marketing season, extra-market coordination within Zambia`s cotton sector focused most intensively on vertical coordination between ginners and smallholder farmers, and to a lesser extent between ginners and spinners. Efforts at horizontal coordination among, for example, ginning companies, were intermittent, as were sector-wide initiatives involving multiple players from all levels in the system. Starting in 2005, two developments increased the level of effort put into sector-wide coordination. First, ZNFU finalized the creation of the Cotton Association of Zambia to represent farmer interests in the sector, providing the Ginners` Association with an organized private sector body with whom to dialogue on key issues. Second, efforts at revision of the Cotton Act became a focus of intense collaboration across stakeholders, with CAZ and the ginners playing the predominant roles. If the proposed revisions to the Act are accepted and the Act is put into practice, then sector-wide coordination efforts will take a major step forward through the Cotton Board. By early 2007, however, the Act had not yet been passed, and a new horizontal coordination effort was beginning: attempts to form a Zambian Cotton Pre-Financiers` Association. Likely members are Dunavant, Cargill, and Great Lakes; these three have invited CCC also to participate, though it is not yet clear whether this company will do so. It is also not clear whether these companies see the new association as a complement to, or a substitute for, the Cotton Board and the Zambia Cotton Ginners` Association. The Cotton Act proposes the formation of a Cotton Board as a statutory body with public and private membership and no mandate to participate as a buyer or seller in the cotton market. The genesis of the Board dates to at least 2000, when the Cotton Development Trust and private stakeholders started developing a regulatory framework for the sector, driven in large measure by a desire to avoid a repeat of the credit default crisis that nearly destroyed the sector from 1997 through 1999. Perhaps as a result of this starting point, the original version of the Cotton Act (which would create the Cotton Board) granted very broad policing powers to the Board. It used vague language to specify the conditions under which these powers could be exercised, and attempted to insulate decisions of the Board from judicial review. It also transferred powers and responsibilities reasonably within the mandate of the Ministry of Agriculture to an agency another step away from political accountability. A 2004 assessment of the Act (Tschirley and Zulu, 2004) suggested that such an approach was at odds with the fact that the sector survived the crisis of the 1990s due in large measure to the institutional innovations and improved management that emerged from competition between the two major players. The assessment further suggested that the Act focus instead on developing legal bases and operational approaches to improve information on borrowers` credit history, on promoting collective action to improve cotton quality and productivity, and on improving the monitoring of sector performance beyond credit repayment. Partly in response to this assessment, revisions to the proposed Act during 2005 allowed judicial review of Board decisions while not substantially altering the Board`s policing powers, and provided a starting point for the creation of a credit bureau for the sector. However, the revisions also altered the balance between the public and private sectors in the Board, to the extent that its new composition threatened to make the Board a largely public sector body rather than a balanced multi-stakeholder body. 19 Stakeholders and research groups raised strong concerns about these and other provisions of the Act, and were taken off-guard when the President signed the Act in December 2005. When these concerns were again expressed, the President directed that the Act not be put into effect until further consultations could take place. This directive was followed by a highly participatory stakeholder process involving the Zambian Cotton Ginners` Association, The Cotton Association of Zambia (CAZ, representing farmers), the Ministry of Agriculture and Cooperatives (MACO), and other interested stakeholders. In a series of three working meetings the Act was fully reviewed and specific revisions were agreed to, including a rebalancing of the membership of the Board to avoid public sector domination. Following elections in late 2006, stakeholders are now awaiting the opportunity to present these proposed revisions to the new government. A final indicator of the government`s engagement with the cotton sector was the role of the Minister of Agriculture during the 2006 marketing season. The rapid appreciation of the Kwacha (see section 2.a) put tremendous pressure on the prices that ginning companies could pay to farmers. When Dunavant and others announced a price of ZKW850/kg, down from 1,220/kg the previous year, the Minister of Agriculture announced in June a suspension of cotton marketing, apparently to allow farmers and ginners to negotiate a higher price. Because the Minister had no legal authority to suspend private marketing of a crop, and judging that the announced price was the best that they could pay, ginners proceeded with the purchase of the crop. However, the Minister`s announcement increased what had already been a very tense situation, and stressed what had been a strong relationship between the Ginners` Association and the Cotton Association of Zambia, representing farmers. Though government in Zambia has long influenced maize marketing both directly through purchases and indirectly through public statements, 2006 was the first time since at least the late 1990s that it had attempted to do the same in cotton. 4.2 Input Credit and Extension The activities of input distribution, credit recovery, and farmer extension have typically been combined within the same field operation units among outgrower companies. We therefore review them together here. The section focuses first on a government initiative to complement existing private outgrower schemes ­ the Cotton Outgrower Credit Fund ­ then briefly discusses the input credit and extension systems run by Dunavant and Cargill. The Cotton Outgrower Credit Fund is part of a broader government effort, launched in 2002, to support export crop production. Stated objectives are to increase production by increasing the availability of inputs on credit, and to reduce pirate buying in which firms purposely buy cotton from farmers who have been supported by other companies. The Fund started with an allocation of about USD $250,000 from government during the 2002/03 growing season. Two years of credit recovery and an additional allocation from government for the 2003/04 season increased the Fund to USD $340,000 for the 2004/05 growing season, and turned it effectively into a revolving fund. No additional government funds were received for the 2004/05 or 2005/06 seasons. Distribution of resources from the Fund favors smaller players in a relative sense: while the 2004/05 allocations to the two largest companies (Dunavant and Clark) could finance only 1-2 percent of their previous year`s area, allocations to smaller companies were substantially larger relative to area (Figure 7). Disbursements to two ginners that had almost no production during the previous year clearly intended to allow 20 them to become at least recognizable players in financing of farmers. The total area financed by the program remains small, at about 3 percent of the previous year`s harvested area. For the smaller players, however, the financing has a substantial impact on their ability to work with farmers. Table 8 shows allocations and recoveries by company for the three years through 2005/06. To date the scheme has avoided the error of centralizing input procurement and distribution to farmers within itself ­ a key factor in the demise of post-reform input distribution efforts in Uganda and Tanzania. By channeling credit to private cotton companies already working with farmers and allowing the companies full freedom in using it, the Fund essentially becomes a means to increase resources in the system and reduce borrowing costs for the companies. Yet the fund suffers from at least one major problem, anticipated from the beginning and now apparent after four years of operation. One expressed objective of the fund was to reduce the amount of pirate buying, promotion by cotton buyers of credit default among farmers, in the sector. To accomplish this objective, one eligibility criterion for the fund needed to be that a participating firm maintain open records of credit disbursements to and recoveries from farmers; if such records were not made available to Fund management, or if questions regarding the information were not adequately addressed, the company`s allocation would need to be eliminated or substantially reduced the following year (Tschirley and Zulu, 2004). This has not been done, and serious questions about the activities of some Fund beneficiaries have to date not been addressed. With default during 2006 back to levels not seen since 1999 or 2000, this issue deserves priority attention from Fund management. Figure 7: Credit allocations to cotton companies from Cotton Outgrower Credit Scheme relative to previous year's cotton area, allocations for 2004/05 21 Table 8: Credit allocations and repayment rates under Cotton Outgrower Credit Fund, 2003/4 to 2005/06 Cropping Year Companies 2003/04 2004/05 2005/06 000,000 ZMK Cargill 295 282 220 Continental 300 298 285 Dunavant 605 503 0 Mukuba 40 90 0 Mulungushi 100 208 395 Stuhardt 20 78 71 Retained by CDT 80 110 279 Total Funds Allocated 1,440 1,569 1,251 Interest 108 118 94 To be recovered 1,548 1,687 1,344 Actual Recoveries 1,532 1,589 878 Recovery rate 0.99 0.94 0.65 Dunavant: Immediately following the credit default crisis of the late 1990s, Lonrho led a period of significant private institutional innovation in 1999, which at the time was being sold to Dunavant. Starting with the 1999/2000 growing season, the company began to create its Distributor System to replace its original system for delivering inputs on credit to farmers and recovering the credit. Tschirley and Zulu (2004) provide a detailed review of the system; Poulton et al. (2004) and Tschirley et al (2007) also touch on it. Key elements of the system include: The elimination of extension agents as company employees, instead relying on formal written contracts with independent Distributors. These Distributors are responsible for identifying farmers to whom they wish to provide cotton inputs 10, receiving the inputs on credit from Dunavant, delivering these inputs to their selected farmers along with technical advice, and ensuring the sale of the farmers` crop to Dunavant to recover the input credit. The Distributor`s remuneration is directly tied to the amount of credit recovered on an increasing scale. Distributors have had a good deal of flexibility regarding how many and which farmers to support; this observation is consistent with Dunavant`s view of Distributors as small businessmen rather than company employees. Dunavant has dropped any Distributor who did not reach minimum credit repayment rates; these cut-offs now stand at 80 percent. 10 The company refers to farmers under a Distributor as farmer groups. In fact, the farmers are groups in only the loosest sense, being organized explicitly for cotton production and without a structure to allow them to be active as a group in other commercial activities. 22 The company invested heavily in training of Distributors in credit management and cotton production and harvesting. The credit management course has been conducted once a year, focusing on improving the Distributor`s ability to pick good farmers and keep them. The production training has been conducted in three phases: before planting, focusing on production techniques; just after planting, aimed at the best use of chemicals and other products to control weeds and pests; and just before harvest, focusing on harvesting and storage, with a strong emphasis on avoiding contamination and properly grading the cotton. Tschirley and Zulu suggest that, from the company`s perspective, the Distributor system greatly diminishes the amount of information that the company needs to manage to ensure adequate credit recovery. The company develops strong relationships with a limited number of Distributors and creates incentives for them to recover as much credit for the company as possible. Thus, the company attempts to substitute the Distributors` local knowledge, social capital, and financial incentives (specified in the written contract) for its own databases and enforcement mechanisms. Dunavant reports that credit recovery rose from 67 percent in 1998/99 (the year prior to the system) to 80 percent in 1999/00, 88 percent in 2000/01, and 93 percent in 2001/02. Rates likely remained around these levels until 2006, when Dunavant reports they may have fallen again below 70 percent, due to the exchange rate appreciation crisis discussed above. As part of its continuing effort to improve productivity at the farm level, Dunavant in 2005/06 launched its YIELD Program (Yield Improvement through Empowerment, Learning, and Discipline) with funding from the German Development Agency (GTZ). The effort is also part of the Cotton Made in Africa project, a German retailer-sponsored effort to mainstream sustainable cotton production practices in Africa 11. The program is based on experience in Uganda`s Agricultural Productivity Enhancement Program (APEP), funded by USAID. The program is based on demonstration plots run by Lead Farmers who apply the 5-finger approach to field management: Early and proper land preparation Planting with first rains Correct plant population (seeding rates, gap filling, thinning) Timely weeding Effective pest management A key decision by Dunavant was to base the demonstration plots on its standard input package featuring treated seed, foliar spray for micro-nutrients only, and pesticides for six sprays per season; herbicide and macro fertilizers were used on some plots during the 2004/05 pilot phase, but eliminated for the launch year in 2005/06. Dunavant classifies farmers based entirely on their management practices, with little or no variation in external input use across the groups. In this classification, traditional farmers achieve mean yields 11 The project was organized by Foundation for Sustainable Agriculture and Forestry (FSAF), created by Otto, a large German retailer with a long history of attention to environmental and social sustainability. Current countries of action are Benin, Burkina Faso, and Zambia. See http://www.fsaf.de/index.php?en-projects, http://www.otto.com/Sustainability.nachhaltigkeit0.0.html?&L=1, and http://www.cottonmadeinafrica.com/index.php?en-home. Dunavant is the only ginner that is an official Project Partner in CMiA, and recently jointed the FSAF Board of Trustees. 23 of 600 kg/ha, better farmers average 1,200 kg/ha, and committed farmers average 1,800 kg/ha. Monitoring data suggest that Lead Farmers, who in principle properly applied the 5- finger approach, achieved mean yields of 1,413 kg/ha, with about 20 percent meeting or exceeding Dunavant`s level of 1,800 kg/ha for committed farmers. Cooperating Farmers averaged 788 kg/ha, compared to 538 kg/ha for other farmers. The latter group was comprised of randomly selected farmers who may have attended FFDs but were not recruited by LFs. We will return to these figures when we review our own crop budget results in Chapter 5. Dunavant advocates spraying on a calendar basis, complemented by scouting. In practice, the recommendation amounts to six calendar sprays plus additional sprays as indicated by scouting results. Farmers are educated about harmful and beneficial insects, and are trained to use a simple pegboard to assess scouting results and decide whether they merit an additional spray. Dunavant`s own monitoring suggests that the number and timing of sprays by Lead Farmers did not differ significantly from that of Cooperating Farmers, suggesting that scouting techniques during 2005/06 had little, if any, effect on spraying practice. Cargill: Cargill Cotton, and Clark before it, has relied on a more traditional system for input distribution, recovery, and extension, featuring heavy monitoring throughout the chain. Figure 8 compares the structure of Dunavant`s Distributor system, its YIELD program model, and Cargill`s approach. The figure shows that Dunavant`s Distributor system relied much less on salaried employees than did Cargill; even the new YIELD program uses less salaried personnel than Cargill. Cargill also signs written contracts and maintains input delivery and cotton sale data for every farmer they support, unlike Dunavant. Data on input receipts, sales, and credit repayment on every farmer are maintained in the company`s data centre. Each year, every farmer applying for input credit is screened for loan default during previous years and is rejected if any default record is shown. Because Clark Cotton seldom participated in sector wide meetings in the past, it has been difficult to evaluate their performance. With the change of ownership to Cargill, there are indications that the company may now become more engaged in such activities. Current management claims credit recovery rates consistently above 90 percent, including in 2005/06, when Dunavant indicate that recovery fell below 70 percent. Differential credit repayment performance by Cargill and Dunavant this year deserves close examination to assess the relative effectiveness of the two companies` approaches. For ten years Cargill has provided farmers with Folifert, which provides NPK in addition to micronutrients. They consider this instrumental, along with tighter organization of its extension system, for reaching a mean yield above 900 kg/ha similar to the yields they achieve in Eastern Province. The basic input package provided by most outgrower firms until 2006/07 included a micronutrient fertilizer, an optional aphicide, an insecticide for bollworm control, and seed. Additionally, each firm sells ultra-low volume (ULV) or knapsack sprayers. Assessing the technical quality of input packages provided by outgrower firms is beyond the scope of this paper; price differences in input packages can reflect differences in type, quality, and quantity of inputs and differences in the price charged for the same input, so these comparisons should be considered only a starting point for possible further analysis. 24 Mulungushi`s basic package includes the chemicals mentioned above along with F135 seed, and cost ZKW103,000 per ha for farmers in 2005/06. Continental provides a choice of F135 and Chureza seed in its package, at a total package cost per hectare of ZKW239,000. Cargill is the only company in the country that encourages herbicide use; it indicates substantial uptake in lowland areas where weed growth can overwhelm cotton. It also appears to allow farmer some flexibility in choosing among specific inputs, reflecting a range of prices and quality. The package without herbicide and with fuzzy seed ranged in 2005 in price from ZKW226,000 to ZKW256,000, while the package that adds herbicide and replaces fuzzy seed with delinted seed ranges from ZKW296,000 to ZKW384,000. 25 Figure 8: Structure of Dunavant and Cargill field operations Dunavant Distributor System Dunavant YIELD Program Cargill Shed Shed Dunavant Employees: District Manager Cargill Employees: Supervise Managers (?) Managers (?) Supervise ACs (1/District, 6 MOs, AMOs total) Area Dunavant Employees: Marketing Cargill Employees: Supervise Coordinator train Site Coord's Officer (2/DM, and train AMOs (18) 12 total) Incentivized farmers: Cargill Employees: Work Site Ast. Mktg Officer Distributor (~ 2,000) train and supervise directly w/ CFs and other Coordinators (10-12/MO, 132 Lead Farmers farmers (289, ~ 15/AC) total) Incentivized farmers: Incentivized Farmers: Meet Lead Farmers recruitCFs, run Contact Farmer weekly w/ AMOs, deal (2890, 10/SC) Demo Plots, hold (4/AMO, ~ 520 directly with farmers Farmer Field Days total) Independent Cooperating Independent farmers: Independent Farmers Farmer (~ Farmers (~ Farmers (~ attend FFD, apply "5- 120/CF, ~ 60,000 80/Distributor, ~ 42,000, finger" mgmt tech's total) 160,000 total 15/LF) Company Employees Incentivized farmers 26 In response to the profitability crisis of 2005/06, Dunavant and Cargill both introduced less expensive input packages that included treated seed, a single type of pesticide, and Soluba for Dunavant, Folifert for Cargill. The cost per hectare was reduced to ZKW151,000 for Dunavant and ZKW110,000 for Cargill. Table 9 provides information on the specific inputs that Continental, Mulungushi, and Cargill make available to farmers, the prices they charged for the 2005/06 season, and retail prices charged during September 2006 for the same chemicals by input companies. Given the differing time period, the prices are not strictly comparable. In general, however, the data suggest that prices charged for some inputs by outgrower companies may be well above the retail prices charged by input companies. Given the concentrated structure of the cotton sector and the very limited choice that most farmers have among companies, this issue deserves further attention. 4.3 Varietal Development & Agricultural Research Prior to the formation of CDT in 1999, all cotton research was done by the MACO Research Branch. The ministry`s researchers conducted trials with government funding and were responsible for releasing varieties. All these responsibilities along with Ministry researchers were transferred to CDT after its inception. Chureza, released in 1988, and F135 released 1992, were both developed under MACO and remain the primary seed stock used in the country. Chureza is best adapted to dry areas and predominates in Southern and Eastern provinces, while F135 is mostly used in Central and Western provinces. CDT released one variety in 2005/06 called CDT 2. Farmer field trials have given yields between 1.5 and 2MT/ha. The staple is of medium length and strong; Cargill indicates that it has better micronaire and longer staple length than Chureza, along with a slightly higher ginning outturn ratio (42 percent). Cargill introduced it into two districts of Eastern Province during 2006/07. Dunavant got it as pre-basic seed which it is now multiplying. Varieties in the pipeline include ALbar9314 which CDT has pre-released. CDT indicates that varieties called CA336, CA347, and MF-20kr are at an advanced stage in testing: CA336 and CA347 are in the third year of multi-location testing while MF-20kr is in the second year. Cargill and Dunavant both indicate that they are very happy both with the existing Chureza and F-135 varieties, and with the new CDT2 variety. Both commonly indicate that the yield potential of all these varieties is 2-3 MT/ha, providing great scope for yield improvement in the country without continual generation of new varieties. Cargill explicitly warns against having too many varieties in use, given the difficulty of enforcing zoning agreements to maintain seed purity. 27 Table 9: Cotton production inputs provided by outgrower companies in 2005/06 growing season, prices charged, and prices at retail charged by input companies Cotton Companies Input Companies Continental Mulungushi Cargill CropChem Omnia Cropserve Input Name Price Charged Price Charged Price Charged Price Charged Price Charged Price Charged Input Type ZKW Unit ZKW Unit ZKW Unit ZKW Unit ZKW Unit ZKW Unit Wuxal 44,000 2l Fertilizers Foliar mix 40,000 1l 15,000 1l 44,000 1l 18,500 1l Soluba 17,000 1 kg 13,000 1 kg - - 10,000 1 kg 11,080 1 kg 8,500 1 kg Decistab 116,000 60 tblts Delta-x 116,000 400 ml Marshal 116,000 400 ml 750 Aphicides 60,000 ml Cypermethrine 90,000 1l 50,000 1l 28,000 1l Karate 90,000 1l 50,000 1l 55,000 1l Pesticides Diamethroate 45,000 1l 28,000 1l Boll pack 40,000 1l Fenvelarate 50,000 1l 40,000 1l Monocrotophos 40,000 1l 60,000 1l Novacetam 16,000 100 ml Agro-prid 16,000 100 ml 200 Acetemiprid 10,000 ml 120,000 1l 60,000 1l Weed mix 50,000 10 l 5,500 1l Herbicides Cotto gard 108,000 2l 52,400 1l CA223 (Fuzzy) 50,000 30 kg Seed CA223 (Delinted) 70,000 15 kg Chureza (Fuzzy) 40,000 20 kg 50,000 30 kg Chureza (Delinted) 70,000 15 kg 20 F135 (Fuzzy) 40,000 20 kg 35,000 kg F135 (Delinted) Sprayer 300,000 250,000 230,000 28 4.4 Seed Multiplication and Importation Dunavant and Cargill both rely on commercial seed producers for their seed supply, ginning the production in separate batches to maintain purity. Seed is then certified by officials from the Seed Control and Certification Institute, and most is treated prior to distribution farmers; all farmers in the PRA indicated that they use treated seed. All other ginners (with the possible exception of Great Lakes) distribute their seed untreated. Cargill imports CA223 variety from CIRAD in South Africa, using it on 22 percent of its area (Chureza covers the other 78 percent), but began replacing this in 2006/07 with CDT 2 and intends to continue that process. Three public sector bodies have some dealing with testing of products manufactured or imported into the country: the Phytosanitary Unit (PU) of the MACO, the Environmental Council of Zambia (ECZ), and the Zambia Bureau of Standards (ZABS). The PU is mostly engaged with assessing the suitability of agricultural products, the ECZ is mostly visible when there is an environmental concern, and ZABS has the mandate to monitor and set up standards of an almost unlimited range of products. The lack of a clear law as to who should test inputs such as fertilizer, herbicides or insecticides has created a situation where firms import these inputs and sell them with no central review and approval. At times firms conduct their own tests usually for commercial effect (CDT 2002). 4.5 Quality Control Dunavant and Cargill both use the official grading system of A, B, and C for seed cotton, though Dunavant added A+. This company estimates that 60 percent of the seed cotton arriving during the 2001 harvest was graded either A or A+. Cargill re-grades every bag at the gin, and both companies maintain strict separation of grades for ginning. Prices paid by Cargill in 2006 were ZKW850/kg for Grade A, 830 for Grade B, and 810 for Grade C. They indicate that most cotton was purchased as Grade A, though the grade can change at the gin. Dunavant and Cargill used very different approaches in achieving what is probably the major success in the cotton sector in recent years: control of polypropylene contamination. Until 1999, most cotton in Zambia was bagged at the farm level using woven polypropylene bags. Fibers from these bags then entered the seed cotton and remained in the cotton lint. Since the polypropylene fibers will not accept dyes, lint contaminated in this way received substantial discounts among buyers. Dunavant addressed this problem primarily by installing cleaning stations in each gin, slow moving conveyor belts at which women are seated, finding and manually removing polypropylene fibers. The company also informed farmers that they would not accept cotton arriving at buying stations in anything other than plastic bags, but have not been able fully to eliminate that practice, despite providing plastic bags in rural areas. As a result, Dunavant continues to employ between 36 and 64 women at each gin, during three shifts per day and six months per year, to clean all cotton entering the gin. At current minimum wages, this practice adds about USD $0.014 to each pound of lint that the 29 company processes.12 Compared to a typical premium of USD $0.06/pound over Index A that Zambia now receives, largely due to the control of polypropylene (and which would quickly become a discount if such contamination returned), the cleaning stations appear to be a good investment by Dunavant. Cargill does not use cleaning stations, relying instead on the strength of its highly organized field operation. The company provides all farmers with plastic bags for cotton picking. Contact farmers and AMOs write farmer contract numbers (the farmer`s national identity number) on every bag of cotton that the company buys; this allows Cargill to trace polypropylene contamination, or opportunistic behavior such as putting foreign matter in the middle of bags of seed cotton, back to individual farmers. Cargill personnel indicate that, in such cases, they have returned the bag and made an example of the offending farmer. Cargill and Dunavant both indicate that CCC previously supplied farmers with polypropylene bags for cotton picking, but ended the practice under pressure from both companies. The virtual elimination of polypropylene contamination from Zambian cotton has increased the premium its top grades receive from international buyers from USD $0.01/lb of lint in the mid-1990s to at least USD $0.06 in 2006/07 (Estur, 2006). This is the largest revealed improvement in quality in SSA during this time and places Zambia second only to Zimbabwe in the premium its cotton lint receives13. 4.6 Pricing There has been no government mandated price, nor any pricing guidance of any kind from government, since liberalization in 1994. Dunavant has typically acted as a price leader, announcing a minimum pre-planting price to farmers, which may be adjusted upwards at the start of the buying season. Cargill typically follows Dunavant`s pricing, while smaller ginners frequently pay higher prices than Dunavant. As competition among private firms began to emerge in 1997, price competition became a key tool in attracting buyers, and also contributed to the serious credit repayment problems which began at that time. A lack of transparency in price setting was stated by some as contributing to misunderstandings between farmers and outgrowers firms, and thus to the repayment crisis (Govereh et al. 2000). There remains a great deal of variability in the level of input credit support offered to smallholders by the various ginners; these differences may allow the companies offering less or no support to use price to attract sellers who may have received input support from another company. Pricing became the focus of intense conflict in 2006, driven by the appreciation of the Kwacha. A key source of the conflict in the sector was that Dunavant announced a pre- planting price in late 2005 of ZKW1,200/kg, with the stipulation that it depended on an exchange rate not below ZKW4,200/USD. With the fall of the dollar to ZKW3,200/kg, Dunavant reduced its price to ZKW850/kg, and others followed their lead. To our 12 36 women work at small gins, and 64 at large gins. Assuming capacity of 15,000 and 25,000 MT of seed cotton, respectively, a GOR of 41 percent, and the official minimum wage of ZKW560,000/month, costs are USD14.05/MT lint at small gins and IUSD14.98 at large gins. 13 With declining premia in Zimbabwe linked to the changing structure of seed cotton marketing in that country, Zambia may now produce Africa`s best quality cotton (see Poulton and Hanyani-Mlambo, 2007). 30 knowledge, this was the first time that Dunavant had not met its pre-planting price; this decision also led to the first instance of government attempting to influence prices directly, and the first time that farmers attempted to negotiate prices with ginners in an organized way (through CAZ). The latter have insisted that they will discuss but not negotiate prices. Dunavant does indicate that it needs to re-build confidence in (its) pre-planting price. They again announced a pre-planting price for 2006/07, of ZKW1,050/kg, and intend to meet it. Despite the concentrated structure of the sector, Zambian companies have paid nominal prices comparable to those in Tanzania, where more companies compete more intensely for the cotton crop (Figure 9). In Chapter 5 we will examine whether these prices translate into attractive earnings for farmers. Figure 10 shows Zambia producer prices in ZKW/kg of seed cotton and USD/kg lint, compared to Index A. The gap between lint-equivalent producer prices and Index A widened slightly in 2006, driven by the very sharp reduction in the Kwacha price in response to the appreciation crisis. Figure 9: Prices paid to cotton farmers in Tanzania and Zambia, 1995 ­ 2003 (USD/kg) Figure 10:Zambian producer prices and Index A, 1995 - 2006 31 5. Cost Competitiveness, Returns to Farmers, and Sustainability 5.1 Processing and Marketing Cost Table 10 provides a comparison of ginning costs in Zambia compared to seven other countries that were part of a multi-country study of cotton sector reform in Africa (Tschirley, Poulton and Labaste, 2009). Because ginners in Zambia did not provide full cost information, Zambia`s budget came from a combination of ginner provided information, secondary data on items such as electricy costs, and international benchmarks for energy usage, packaging, and other items. As such, these estimates represent a best guess on the situation in Zambia. Results suggest that Zambian ginners are substantially more efficient than most West African ginners, but slightly less efficient than those in most neighboring countries of East and Southern Africa. The negative effect of the overvaluation of the kwacha is clearly seen in the reduction of estimated cost from US$0.137/kg lint to US$0.108/kg if we use the mean exchange rate over the past decade rather than the 2007 rate. Table 10:Estimated ginning costs in Zambia and seven other countries of SSA (US cents per kg of lint cotton) Burkinaa Malib Cameroonc Mozamb.d Zambiae Zimbab.f Tanzaniag Ugandag Local National National Local Concen- Concen- Competi- Type of system Monop Monop Monop monop trated trated tive Hybrid Exchange rate to US$ (2006) 505 505 505 23.5 3,600 Variable 1,200 1,800 Type of gins saw saw saw saw saw saw/roller roller roller Avg unit ginning capacity 45,000T 40,000T 31,000T 13,500T 20,000T 25,000T 6,300T 5,000 T % capacity utilized 100% 65% 100% 20% 80% 64% 80% 20% Fixed costs/kg of lint 5.84 7.99 4.03 17.15 5.86 3.30 1.84 12.29 Depreciation 3.31 4.59 3.06 7.81 3.13 1.9 0.65 6.02 Salaries 1.18 1.08 0.77 9.29 2.60 1.35 1.19 6.27 Other 1.35 2.32 0.2 0.05 0.14 0.05 0 0 Variable costs/kg of lint 9.99 15.39 9.39 6.51 7.61 4.76 6.31 7.66 Energy 2.5 4.4 3.07 2.36 0.5 0.04 0.94 3.04 Packaging 3.49 3.45 3.49 3.91 3.5 2.17 4.17 3.05 Other (incl maintenance) 4 7.54 2.84 0.24 3.61 2.56 1.2 1.58 Total cost ... at assumed capacity 15.83 23.38 13.42 23.66 13.47 8.06 8.15 19.95 ... at 100% capacity 15.83 20.58 13.42 9.94 12.30 6.87 7.78 10.12 ... at assumed cap. & mean 1995-2006 exch.ange rate 13.62 17.71 11.55 NA 10.78 NA NA NA Source: Gergely (2009) Note: Both Zambia and Uganda use some secondhand ginning equipment. a. SOFITEX actual accounts for 2003/04; b. CMDT budget for 2006/07; c. SODECOTON actual account for 2004/05; d. Estimate for 2005 calendar year; e. Estimates by Gerald Estur (consultant) for 2005/06; f. Estimates for 2005/06 (Poulton and Hanyani-Mlambo 2007); g. Estimates based on 2006/07 costs but 2004/05 capacity utilization (Poulton and Maro 2007). 32 5.2 Cost Competitiveness at Farm Level Table 11 presents summary information from the crop budgets developed through Participatory Rapid Appraisal exercises in six villages spread over Chipata and Katete districts of Eastern Province during March 2007. In each village, 15 to 40 farmers attended the sessions, though typically about 10 provided nearly all the information. Detailed budgets are provided in Appendix A. The shares of each group in the number of cotton farmers, cotton area, and cotton production are based on a complete enumeration of all cotton farmers in each village, and assignment of each to one of the groups. Shares of area and production are then based on data generated during the interviews. Because village selection was not random, and because the number of villages is small, these shares should not be taken as representative of Eastern Province as a whole. Five points emerge from the table. First, the area-weighted average yield that emerged from the PRA exercise is comparable to that claimed by Cargill and Dunavant in Eastern Province: 836 kg/ha compared to the claimed 900 kg/ha. Notably, yields based on household surveys are substantially higher. Second, about 15 percent of farmers in these villages are equipped with animal traction and related equipment ­ all of group 1, and a minority of group 2. These figures are somewhat lower than household survey estimates, which suggest that 20-25 percent of cotton farmers in Eastern Province own such equipment. This low level of ownership (even in household surveys) is in part a reflection of serious reductions in animal herds over the past decade due to reduced veterinary service support and disease; farmers indicate now that herd sizes are beginning to rise again. Third, better equipped farmers (group 1) and those able to hire labor whenever needed (primarily groups 1 and 2) are able to achieve higher yields due to greater timeliness in field operations. This pattern is consistent with analysis of household data, which suggests that households with animal traction enjoy yields more than 300 kg/ha higher than those without animal traction (Appendix B). Fourth, yield and area planted are both highest for households with most animal traction equipment; groups 3 and 4, both of whom have only handhoes, are distinguished by group 3 hiring more labor and conducting its field activities in a more timely manner, resulting in somewhat higher yields. Finally, total cost of production (assuming an opportunity cost of family labor of ZKW5,000/day, equal to the typical daily wage rate in rural areas) is lowest for the best equipped farmers, and rises substantially for the least equipped households with lowest yields. 33 Table 11: Summary crop budget indicators by farmer type, mean of crop seasons 2004/05 ­ 2006/07 Group 1 2 3 4 Area weighted mean Share of cotton farmers 9% 15% 42% 34% Share of cotton area 27% 20% 42% 11% Share of cotton prod'n 36% 26% 32% 6% Area in cotton (ha) 4.5 2.0 1.5 0.5 1.51 Cotton Yield (kg/ha) 1,200 1,050 600 450 836 Animals, plough, Equipment ridger, sprayer, ox-cart, Sprayer, handhoes Handhoes Handhoes handhoes Labor All hired Mostly hired Mostly family All family Revenue Indicators ZKW USD ZKW USD ZKW USD ZKW USD ZKW USD Gross revenue 1,248,000 312 1,092,000 273 624,000 156 468,000 117 868,920 217 Input cost/gross revenue 0.15 0.17 0.19 0.21 0.18 Gross margin (excl. fam. labor) 426,522 102 349,871 83 466,113 115 328,634 81 417,052 101 Returns/day family labor 11,284 2.69 11,214 2.66 3,917 0.97 3,029 0.75 7,268 1.75 Net margin (after fam. labor) 237,522 53 193,871 43 -128,887 -37 -213,866 -58 25,247 1 Total cost/kg 842 0.22 855 0.22 1,255 0.32 1,515 0.39 1,092 0.28 Gross margin from all cotton 1,919,350 457 699,742 166 699,169 173 164,317 41 969,899 234 34 5.3 Returns to farmers and poverty alleviation considerations Table 10 can also be examined for insights regarding the impacts of cotton cultivation on poverty. We complement this discussion with Figures 11 and 12. Four comments are relevant. First, farmers in all villages overwhelmingly indicated that only the poorest households did not grow cotton. Second, during 2004/05 and 2006/07, Groups 1 and 2 earned 2.5-3 times the going daily wage rate in rural areas from their cotton growing activities, while groups 3 and 4 earned below this wage rate during all three years (Figure 11). Third, during all three years, earnings from cotton were typically less than 10 percent of the extreme poverty line for a family of six for all but Group 1 (Figure 12; average household size in Zambia is 6). Finally, returns to cotton were far lower in local currency terms in 2005/06 than during the other two years in our analysis. During that year, none of the groups earned above the going daily wage rate, and only Group 1 barely earned enough from its cotton activities to keep a family of six above the poverty line. Because the sudden and large appreciation of the Kwacha in 2005/06 makes it a very unusual year, we consider 2004/05 and 2006/07 to be more representative of what farmers typically earn from their cotton crop. Establishing causation between cotton cultivation and total household income levels is exceptionally difficult. Appendix C and Figure 13 present results from an econometric analysis of the effect of cotton cultivation on net farm incomes, the probability of earning off- farm income, and total household incomes. The results are shown by quintile of cotton area, and depict the estimated association between cotton cultivation at each of these levels and net farm income, and probability of having off-farm income, and net total household income from all sources (including the value of retained food production; see Appendix C for full results)14. The regressions control for household education, household assets, total land area, whether the household is headed by a female, and household size and composition (number of children) in an attempt to isolate the effects of cotton. Asterisks above or below each bar indicate whether the result was statistically significant at p=0.10 or better. Three patterns stand out. First, households with small areas in cotton appear to earn the highest net cropping incomes, controlling for all these other factors; the top two quintiles (top 40 percent in terms of cotton area) appear not to gain any net cropping income from cotton. Second, cotton cultivation appears to be associated with a reduced probability of a household earning off- farm income, except among those devoting the least area to cotton. Third, and consistent with the pattern on net farm income, total household income (again adjusted for the other independent variables) may decline with cotton area, though none of these results are statistically significant. This pattern echoes previous findings in Mozambique (Benfica et al, 2006; Tschirley and Benfica, 2001; Tschirley and Weber, 1996) showing that cotton farmers systematically gave up off-farm income and typically did not enjoy any net gain in overall income. The specific pattern across cotton area quintiles is also remarkably similar to that found more recently in Mozambique (Boughton, et al ­ paper for this project). 14 The regressions are single stage OLS for net farm income and total household income, and a logit analysis for having off-farm income. The single stage OLS approaches do not control for possible selection bias, and will be replaced with two-stage Heckman or double hurdle approaches in the next version of the paper. Households were included in the analysis only if they resided in a district that had at least two cotton farmers. This reduced the number of districts from 70 to 24, and the sample size from 5419 to 2275. 35 Figure 11: Returns per day of labor by farmer group and year, compared to average rural wage rate Figure 12: Total net earnings from cotton by farmer group and year as share of extreme poverty line for family of six 36 Figure 13: Summary regression results from analysis of association between area planted to cotton and household income in Zambia Yet this analysis also suffers from serious weaknesses. Results in this paper, research in Mozambique, and broader research throughout Africa suggest negligible or negative returns to education in agriculture, including cotton. Parallel research suggests that level of education is a key determinant of access to remunerative off-farm income. It may be, then, that relatively uneducated household heads have few remunerative off-farm income opportunities, and that cotton provides the kind of reliable cash income that they would not be able to earn otherwise. Panel data sets in Zambia and Mozambique, which allow more effective control of unobserved variables, may now allow this issue to be explored more rigorously. In the absence of results from such analysis, the following line of reasoning, based on observations from the field and examination of available data, suggests that cotton has meaningful positive effects in reducing poverty in rural Zambia. If cotton cultivation were not available to farmers, most would attempt to produce more maize. This claim is strongly supported by observations in the field during the PRA exercises: many farmers who initially decided not to plant cotton this year, due to very low prices the previous year, attempted instead to plant maize, but when they were unable to obtain fertilizer at subsidized prices, they returned to the cotton companies to obtain planting seed and other inputs. Tobacco and coffee, other important smallholder cash crops in Zambia, are produced in agro-ecological zones not suited to cotton. Other potential cash crops, such as groundnut, have small markets and no organized system for input provision or output marketing. Cash crops that do provide a more organized output market and which can be produced in the same areas as cotton, such as paprika, have not shown nearly the production growth that cotton has, and do not enjoy as ready a world market as cotton. 37 Because Zambia is a land-locked country and a relatively high cost maize producer, market prices for maize would be likely to fall substantially if most cotton area were dedicated to maize. Maize is much less drought tolerant than cotton, meaning that in drought years, farmers would be more likely to earn negative cash returns from maize than from cotton. Table 12 provides an indication of the difficulty smallholders face in relying on maize as their principal cash crop. Yields for both cotton and maize are means from the Central Statistical Office`s Post-Harvest Survey. Input costs for maize come from previous crop budget work (Haggblade and Tembo, 2004), while those for cotton come from our PRA work. For maize, we focus on 2006, when the Food Reserve Agency (FRA) bought a substantial quantity of maize at a very high price: ZKW760/kg (ZKW38,000/50-kg bag), or about USD $181/MT. During the same year, key informants in Eastern province indicated that farmers unable to sell to the FRA received prices of about ZKW500/kg (ZKW25,000/50- kg bag). We examine both of these scenarios, and compare them to cotton in 2005/06, when prices were exceptionally low in Kwacha terms, and 2006/07, when they partially recovered. Results show that average cotton returns during both years were below those for maize growers who were able to sell to FRA, but were competitive with returns earned by those selling to the open market. With very high maize plantings in 2006, a large crop expected for 2007, and questions about FRA`s ability to buy quantities similar to last year, domestic open market prices might fall below the ZKW500/kg level, making cotton more attractive than maize. Table 12: Indicative crop budgets for maize and cotton in Zambia Maize, 2006 Cotton Selling on Selling to FRA market 2006 2007 Yield (kg/ha) 1,348 1,348 874 874 Price (ZKS/kg) 760 500 850 1050 Gross revenue (ZKW) 1,024,100 673,750 742,475 917,175 Input cost (ZKW) 200,000 200,000 200,000 130,000 Gross margin (ZKW) 824,100 473,750 542,475 787,175 Labor days 90 90 110 110 Returns per day (ZKW) 9,157 5,264 4,932 7,156 Exchange rate 3,200 3,200 3,200 4,200 Returns per day (USD) 2.86 1.64 1.54 1.70 Source: CSP/PHS for yields; FRA, key informants, and cotton companies for prices; Haggblade and Tembo for maize input costs and maize labor days; PRA for cotton input costs and days of labor; maize labor days adjusted down from Haggblade and Tembo to reflect cotton`s recognized greater labor intensity. Though previous analysis suggested that maize delivered substantially higher returns than cotton (Haggblade and Tembo, 2004, quoted in Tschirley and Zulu, 2005), the results reported here are more consistent with the rapid increase in number of cotton farmers and area planted to maize over the past seven years, and with the fact that essentially all of these cotton growers also grew maize. 38 5.4 Sector Sustainability Zambia`s cotton sector faces two key challenges to its long-term sustainability: managing and adjusting to the real exchange rate effects of copper exports, and sector coordination challenges in light of the recent ­ and perhaps long-term ­ changes in sector structure. A third possible sustainability challenge ­ soil fertility depletion due to lack of use of macro fertilizers on cotton ­ is likely not pressing, due to cotton`s rotation with maize and the widespread use of such fertilizers on maize. The startling appreciation of the Kwacha from late 2005 through mid-2006 was simply an accentuated version of the appreciation that had been taking place since mid-2002. In fact, Calí and te Welde show long-term appreciation of the Kwacha since the late 1980s. The recent decline of the Kwacha still leaves its real level 35 percent stronger than the fairly steady rates seen from 1996 through 2002. Fynn and Haggblade predicted the strongly negative impact of the recent sharp appreciation on cotton and other export sectors, and events during the harvest in 2006 and the later planting bore out their forecasts; Dunavant reduced its support to the sector, it and Cargill both provided less expensive input packages to farmers for 2006/07, and plantings fell by as much as 50 percent. Calí and te Welde further show that Chile, a country nearly as dependent on copper exports as Zambia, has managed its copper revenues in such a way that its exchange rate shows very few rapid movements and almost no correlation with the copper price; Zambia`s real exchange rate, in contrast, is strongly correlated with the copper price, especially over the past two years, and is also much more volatile. Fynn and Haggblade (2006) summarize the challenge for Zambia: "Governments who have successfully managed similar foreign exchange windfalls to the advantage of their agricultural producers have used the windfall earnings to promote, rather than impede, economic diversification. Their main tools have been active management to avoid excessive exchange rate volatility, sterilization of foreign exchange earnings to avoid currency appreciation, strict controls on government spending in order to combat inflation, and significant public investment in agricultural technology and infrastructure. To date, the Zambian government has adopted none of these measures." Previous sections in this paper highlighted the substantial structural changes at the ginning level in Zambia over the past two years, and suggested that they might represent a long-term shift in the level of direct competition between firms in the sector. Experience in Uganda and Tanzania shows that heavy competition for seed cotton among ginners undermines input credit provision and cotton lint quality. More recently, Zimbabwe has moved from a concentrated sector to one with much more competition, and has also seen input credit provision and quality decline. While Zambia had not reached the productivity levels seen in Zimbabwe, it was moving in that direction, and had already nearly matched Zimbabwe in the premia its lint receives in world markets. The rise of CAZ since 2005, its active collaboration with the ginners and MACO in revising the Cotton Act, and its continued constructive engagement with ginners in not allowing disagreements on pricing to divert attention from productivity enhancement, is also a positive sign for the sector. Finally, Zambia joined ICAC in 2006, potentially contributing to greater engagement and knowledge within the sector regarding the world market in which it operates. Yet all these gains can be undone with unrestrained competition among ginning companies. Poulton and Hanyani-Mlambo suggest that, If Zimbabwe policy makers wish the national 39 cotton industry to maintain its historic high yielding and high quality` profile, then some degree of regulation will be necessary ... In Zambia, stakeholders worked together to support revisions to the Cotton Act that would create a stakeholder-led Cotton Board. Despite repeated public statements by Dunavant, Cargill, and CAZ (and private statements by CCC) that they support the revised Act and wish to see it passed, no forward movement has occurred since the elections in September 2006; the Act currently sits in MACO awaiting endorsement from the Minister and forwarding to Parliament. Presently, the three main multi- nationals in Zambia`s market (Dunavant, Cargill, and Great Lakes), possibly joined by CCC, are discussing creation of a Cotton Pre-Financers` Association. Whether these companies see such an association as a complement or substitute for the Cotton Board is not clear at the present time. Recent experience across Africa suggests that, however it occurs, more overt coordination among ginners, and between ginners and farmers, will be required in future years if the sector is to maintain its impressive gains in lint quality and build on the progress it has made to date in farm level productivity. Assuming the sector can consolidate a workable approach to horizontal coordination, management of the Cotton Outgrower Credit Fund would be enhanced with more rigorous criteria for eligibility. More generally, government needs to avoid actions and statements that increase uncertainty within the cotton sector. Such behaviour has been a persistent characteristic of government action in the maize sector, and has resulted in more price volatility than would otherwise have prevailed. Events surrounding the Kwacha appreciation, especially the Minister`s attempt to suspend cotton marketing in June of 2006, suggest that some of the same governmental patterns seen with regards to maize may begin to emerge in cotton. It is imperative that these mistakes be avoided; creation of the Cotton Board may help in this regard by investing a recognized multi-stakeholder body, with both public and private representation, with the authority to speak for the sector`s interests. At the present time, soil fertility and seed quality do not appear to be hindrances to sustainable increases in cotton production. Historically in Zambia, maize has received substantial application of external nutrients in the form of urea and basal compound fertilizers. Cotton in Zambia is rotated with maize and, because it is deep-rooting, is able to benefit from any nutrients that may have leached below the maize root zone. Both Dunavant and Cargill are pleased with the performance of Chureza and F-135, noting that the country has not yet come close to exhausting their yield potential of 2-3 MT. CDT, while criticized by many sector stakeholders, has developed a variety (CDT 2) that appears to have won acceptance among ginners, and has 2-3 other varieties in the pipeline. 40 6. Lessons Learned Several lessons emerge from Zambia`s post-reform experience. First, a concentrated sector has inherent advantages over more competitive sectors in the areas of input distribution, credit repayment, and quality improvement. Unlike Tanzania and Uganda, whose post-reform structures were highly competitive, Zambia`s input distribution system never collapsed, and in fact has consistently improved since reform. Furthermore, government played almost no role in this expansion of input credit and improvement in quality, because the two largest private firms were able to ensure sufficient credit recovery to make it financially worthwhile for them to mount the input credit programs. More competitive sectors such as in Tanzania and Uganda have required much more government action to attempt to resolve these problems. Second, concentrated, market-based systems may be subject to periodic structural instability. Zambia`s early duopoly after liberalization was shaken after 3-4 years by substantial credit default instigated by new entrants who provided little if any input credit. Institutional innovation by both Lonrho/Dunavant and Clark/Cargill allowed these companies to re- establish their supply chains, reduce credit default to manageable levels, and substantially grow the overall crop. Over the past two years, more companies have again entered and credit default has returned as a serious problem, at least for Dunavant. When the established companies are multi-national cotton trading firms, as they are in Zambia, they are likely to have substantial advantages over smaller local firms in terms of access to low-cost off-shore capital markets and in their ability to maximize revenue and minimize trading costs through their worldwide trading networks. A key empirical question is whether these advantages are large enough to allow these firms to continue providing input credit when smaller firms provide little credit and promote side-selling by farmers. Finally, government can easily get it wrong in these sectors, as shown by the sudden Kwacha appreciation, the Minister`s attempt to suspend cotton marketing, and the inability or unwillingness of CDT to impose more stringent eligibility criteria for the COCF. MACO`s participation in the highly collaborative review of the Cotton Act is a much better model for government action. It deals explicitly with some of the crucial regulatory issues raised above, and provides hope that government and private stakeholders together can learn how to collaborate in addressing the sector`s key challenges. One key challenge for concentrated sectors such as Zambia`s, therefore, is to develop a flexible and commercially supportive regulatory regime that understands the strengths and weaknesses of the concentrated model. Specifically: Concentrated sectors need limited barriers to entry (licensing rules that specify strict capabilities and conduct of firms wishing to participate in the sector) to defend the ability of firms within the sector to coordinate on input supply, extension, quality control, and perhaps other matters. Concentrated sectors must retain some contestability to provide incumbents with an incentive to maintain attractive seed cotton prices. As in the case of local monopolies, it is important for those in charge of policy for the sector to form a clear idea of the 41 types of company that they wish to allow into the sector, so as to be able to formulate rules accordingly. Given these sectors` tendency to slide towards unrestrained competition and credit default crises, a strong commitment to raising farmer productivity and improving quality within the chain should be given high priority in the selection criteria Given the problems of relying entirely on the threat of entry to discipline incumbent firms within concentrated sectors, it may also be desirable to develop price setting mechanisms that are more formalized than the price leadership that has prevailed in concentrated systems so far. As piloted in many West African cotton sectors, farmer organizations have a potentially very important role to play within such mechanisms. However, this role needs to be informed by a solid understanding of world markets to avoid the serious problems of sector sustainability that now afflict many West African sectors. The second key challenge is to mount a sustained and coordinated approach to substantially raising productivity at the farm level. Despite the country`s relatively good performance on input credit provision, yield growth has been slow, yields remain well below those in West- and Central Africa, and average returns to farmers appear to be no higher than in Tanzania, where input use and yields are lower. Two findings of the farm productivity analysis in chapter five are especially relevant. First, we found that animal traction is strongly correlated with cotton yields, not just cotton area planted. The more general point of this finding is that timely and adequate cultural practices could generate much higher yields at current levels of input use. However, the second key finding is that a large plurality of farmers (group 4) are unable to prioritize cultural practices on their cotton fields because of limited labor, lack of animal traction, lack of cash to hire labor, and the absolute necessity of ensuring an adequate maize harvest. Dunavant`s focus on the five non-negotiables (or five fingers) thus makes a great deal of sense, but under present circumstances, it appears that only a minority of farmers will be able consistently to apply them. Forty percent of farmers in many West- and Central African countries commonly achieve yields of 1,200 kg/ha or more, and animal traction is a key reason they are able to do this (fertilizer use is also high, but soil fertility is lower than in Zambia). The cotton systems in those countries helped farmers build up these assets over a long period of time, and did so in part with outside resources that saw cotton as a vehicle for rural development, not just a single cash crop. Dunavant`s YIELD program, financed by donor funds, may be a starting point for an industry-wide approach to attract the resources needed to effect this transformation. Greater attention, however, would need to be paid to building the farm assets of households. 42 7. References Agricultural Consultative Forum (2002). Findings of the Joint MAC/ZNFU Study on Agriculture Sector Competitiveness and Impact of the COMESA Free Trade Area: Phase I Final Report. ACF/ZNFU/Royal Netherlands Embassy. Lusaka. Baffes, John (2002). Tanzania`s Cotton Sector: Constraints and Challenges in a Global Environment. Mimeo. The World Bank. Washington, D.C. Boughton, Duncan, David Tschirley, Afonso Osorio, Higino De Marrule, and Ballard Zulu (2003). Cotton Sector Policies and Performance in Sub-Saharan Africa: Lessons Behind the Numbers in Mozambique and Zambia. Paper prepared for the IAAE Triennial Conference, 2003. Durban. Business Day (2006). Cargill Sews up Afgri`s Cotton Gins. 10 April 2006. http://www.businessday.co.za/articles/companies.aspx?ID=BD4A183738 Calí, Massimiliano and Dirk Willem te Velde (2007). Is Zambia contracting Dutch Disease?. ODI Working Paper 279. London. Export Board of Zambia (2001). 2001 Exporter Audit Report. Gergely, Nicolas (2009). Cost efficiency of companies, overall sector competitiveness, and macro impact in Tschirley, Poulton and Labaste, eds. Organization and Performance of Cotton Sectors in Africa: Learning from Reform Experience. Forthcoming from World Bank. Gibbon, Peter (1999). Free competition without sustainable development? Tanzanian cotton sector liberalization, 1994/95 to 1997/98. Journal of Development Studies 36 (1), 128- 150. Goodland, Andrew and Anne Gordon (2000). Production credit for African small-holders: conditions for private provision. Savings and Development, Vol. XXIV, No. 1. Govereh Jones, Jayne T.S. Tschirley David, Donovan Cynthia Donovan, Nijhoff J.J., Weber Michael, and Zulu Ballard (2000), Improving Smallholder & Agribusiness Opportunities in Zambia`s Cotton Sector: Key Challenges & Options, Zambia Food Security Research Project Paper Number 1. Haggblade, Steven and Gelson Tembo (2003). Conservation Farming in Zambia. Mimeo. Michigan State University Food Security Research Project/IFPRI. Lusaka. Huffman, W. E. (1980): Farm and Off-farm Work Decisions: The Role of Human Capital, The Review of Economics and Statistics 62 (1): 14-23. ICAC (2004). ICAC Press Release, January 2, 2004". Lundbæk, Jeppe (2002). Privatization of the Cotton Sub-sector in Uganda: Market Failures And Institutional Mechanisms to Overcome These. Master Thesis, The Royal Veterinary and Agricultural University, Department of Economics and Natural Resources, Copenhagen. 43 Ministry of Agriculture Food & Fisheries (2000), Zambian Cotton Sector Review, Draft Report, Prepared by Societe Francaise de Realisation, d`Etudes et de Conseil. Nijhoff, J.J., Gelson Tembo, James D. Shaffer, T.S. Jayne, and Julius Shawa (2003). How Will the Proposed Crop Marketing Authority Affect Food Market Performance in Zambia? An ex Ante Assessment to Inform Government Deliberation. Policy Synthesis No. 8. Food Security Research Project. Lusaka. Patel J. Ramesh and Mtonga C.Q (2002). State of the Textile and Clothing Industries in Zambia and Policies to Revive These Industries for the Promotion of Export Led Growth, Paper prepared for a Ministry of Commerce and Industry Workshop on the Textile and Clothing Industry in Zambia. Poulton, Colin, Peter Gibbon , Benjamine Hanyani-Mlambo Jonathan Kydd, Wilbald Maro , Marianne Nylandsted Larsen, Afonso Osorio , David Tschirley and Ballard Zulu (2003). Competition and Coordination in Liberalized African Cotton Market Systems. World Development, Vol. 32, No. 3. RATES (2003). Cotton-Textile-Apparel Value Chain Report, Zambia. Presented to the RATES Center. Nairobi. Tschirley, David, and M.T. Weber (1994). Food Security Strategies under Extremely Adverse Conditions: The Determinants of Household Income and Consumption in Rural Mozambique. World Development, 22, 2: 159-173. Tschirley, David and Ballard Zulu (2003). Zambian Cotton in a Regional Context: Performance under Liberalization and Future Challenges. Policy Synthesis No. 7. Food Security Research Project. Lusaka. Tschirley, David, Ballard Zulu and James Shaffer (2004). Cotton In Zambia: An Assessment Of Its Organization, Performance, Current Policy Initiatives, And Challenges For The Future. Working Paper No. 10. Food Security Research Project, Lusaka. Tschirley, David and Ballard Zulu (2004). Presentation on Recent Policy Initiatives in Zambia`s Cotton Sector. Zambia Cotton Stakeholders` Workshop, sponsored by the Agricultural Consultative Forum and the Food Security Research Project. 1 April 2004. Lusaka. Tschirley, David, Colin Poulton and Patrick Labaste, eds (2009). Organization and Performance of Cotton Sectors in Africa: Learning from Reform Experience. Forthcoming from World Bank. Yang, D.T. (1997): Education and Off-Farm Work, Economic Development and Cultural Change 45 (3): 613 -632. Zulu, Ballard and David Tschirley (2004). An Assessment Of Current Policy Initiatives In Zambia`s Cotton Sector. Policy Synthesis No. 9. Food Security Research Project. Lusaka. 44 Annex 1: Data and Statistics Table A1. Detailed crop budgets by farmer type and year Groups 1 2 3 4 Technical Itinerary Total cotton area (2-limas) 9 4 3 1 Share of area in cotton 0.5 0.33 0.33 0.33 Use of manure No No No No Cleaning, 1 plowing, 1 ridging Cleaning, followed by single Cleaning, plow/ridge by hand hoe Cleaning, plow/ridge by hand hoe with own ox plow ridging/plowing with hired ox team with family labor with family labor Land Preparation Planting method On ridges On ridges On ridges On ridges Planting seed (kg; all treated de-linted) 7.5 7.5 7.5 7.5 Thinning Hired Family Family None Gap Filling Hired Family Family None Number of weedings 3 3 3 2 Type of labor for weeding Hired Hired Family Family Number of sprayings (pest + 6 6 3 2 Soluba/foliar) Type of labor for spraying Hired Hired Family Family Harvest labor Hired Hired Hired Family Transport labor Hired Hired Hired Family Input Cost Manure 0 0 0 0 Seed 20,500 20,500 20,500 20,500 Pesticides+Soluba or Foliar Mix 44,750 44,750 22,375 14,917 Sub-total 65,250 65,250 42,875 35,417 Ownership of equip., animals Animals for AT Yes No No No Plough Yes No No No Ridger Yes No No No Handhoe 5 4 3 2 Knapsack sprayer Yes Yes No No ULV sprayer No No No No 45 Table A1. Detailed crop budgets by farmer type and year Groups 1 2 3 4 Ox-cart Yes No No No Amort'ion of equip, animals Two oxen (50% used off-farm) 14,086 0 0 0 Plough 6,299 0 0 0 Ridger 11,061 0 0 0 Handhoe 4,500 5,346 5,346 10,692 Knapsack sprayer 2,224 3,302 0 0 ULV sprayer 0 0 0 0 Ox-cart (50% used off-farm) 11,738 0 0 0 Sub-total 35,822 8,648 5,346 10,692 Hired Services AT team for plowing/ridging 0 60,000 0 0 Spraying (rented sprayer only) 0 0 7,500 5,000 Ox, cart for transp field to home 0 14,000 8,000 6,000 Sub-total 0 74,000 15,500 11,000 Family Hired Hired Family Hired Hired Family Hired Hired Family Hired Hired Labor requirement (man-days) days days cost days days cost days days cost days days cost Cutting 0 7 30,000 0 7 30,000 3.5 3.5 15,000 7 0 Plowing 6 0 0 0 0 0 0 0 0 0 0 Ridging 3 0 0 0 0 0 0 0 0 0 0 Combined Plow/ridge 0 0 0 0.6 3 0 0 0 0 Handhoe land preparation 0 0 0 0 0 0 10 10 45,000 20 0 Planting 0.8 4 17,000 4 0 0 4 0 0 4 0 Thinning 0.4 2 8,500 2 0 0 2 0 0 0 0 Gap filling 0.2 1 4,000 1 0 0 0 0 0 0 0 Weeding 1 1.8 9 45,000 1.8 9 45,000 9 0 0 9 0 Weeding 2 1.8 9 45,000 1.8 9 45,000 9 0 0 9 0 Weeding 3 1.8 9 45,000 1.8 9 45,000 9 0 0 0 0 Soluba/Foliar mix (applied with insecticides) Spray 1 0.1 0.5 12,000 0.1 0.5 2,500 0.5 0 0 0.5 0 Spray 2 0.1 0.5 12,000 0.1 0.5 2,500 0.5 0 0 0.5 0 Spray 3 0.1 0.5 12,000 0.1 0.5 2,500 0.5 0 0 0 0 Spray 4 0.1 0.5 12,000 0.1 0.5 2,500 0 0 0 0 0 46 Table A1. Detailed crop budgets by farmer type and year Groups 1 2 3 4 Spray 5 0.1 0.5 12,000 0.1 0.5 2,500 0 0 0 0 0 Spray 6 0.1 0.5 12,000 0.1 0.5 2,500 0 0 0 0 0 Harvest 2 10 50,000 2 10 50,000 15 0 0 11.25 0 Transport home 0.5 0 0 0 1 0 Sub-total 18.9 47 286,500 15.6 44 200,000 59.5 10 0 54.25 0 RETURN SCENARIOS 2006/07 Price= 1,050 1,050 1,050 1,050 Input cost = 65,250 65,250 42,875 35,417 cotton farm budget 1a 1b 2 3 Total production/2-lima 600 525 300 225 Price seed cotton (ZK/kg) 1,050 1,050 1,050 1,050 Gross income 630,000 551,250 315,000 236,250 Input cost 65,250 65,250 42,875 35,417 Hired services (AT + pump) 0 74,000 15,500 11,000 Hired labor 286,500 200,000 0 0 Amortization of equip, animals 35,822 8,648 5,346 10,692 Total cash cost 387,572 347,898 63,721 57,109 Net cash income (excluding family labor) 242,428 203,352 251,279 179,141 Returns per day of family labor (ZKW) 12,827 13,035 4,223 3,302 Returns per day of family labor (USD) 3.05 3.10 1.01 0.79 Implicit value family labor 5,000 5,000 5,000 5,000 Total value family labor 94,500 78,000 297,500 271,250 Net income (including value family labor) 147,928 125,352 -46,221 -92,109 Full cost of seed cotton (ZKW/kg) 803 811 1,204 1,459 Full cost of seed cotton (USD) 0.19 0.19 0.29 0.35 Typical area cultivated in cotton (2- lima) 9 4 3 1 Total cash earnings (ZKW) 2,181,850 813,408 753,837 179,141 Exchange rate 4,200 4,200 4,200 4,200 47 Table A1. Detailed crop budgets by farmer type and year Groups 1 2 3 4 Total cash earnings (USD) 519 194 179 43 2005/06 Price= 850 850 850 850 Input cost = 100,000 100,000 65,709 54,278 cotton farm budget 1a 1b 2 3 Total production/2-lima 600 525 300 225 Price seed cotton (ZK/kg) 850 850 850 850 Gross income 510,000 446,250 255,000 191,250 Input cost 100,000 100,000 65,709 54,278 Hired services (AT + pump) 0 74,000 15,500 11,000 Hired labor 286,500 200,000 0 0 Amortization of equip, animals 35,822 8,648 5,346 10,692 Total cash cost 422,322 382,648 86,555 75,970 Net cash income (excluding family labor) 87,678 63,602 168,445 115,280 Returns per day of family labor (ZKW) 4,639 4,077 2,831 2,125 Returns per day of family labor (USD) 1.45 1.27 0.88 0.66 Implicit value family labor 5,000 5,000 5,000 5,000 Total value family labor 94,500 78,000 297,500 271,250 Net income (including value family labor) -6,822 -14,398 -129,055 -155,970 Full cost of seed cotton (ZKW/kg) 861 877 1,280 1,543 Full cost of seed cotton (USD) 0.27 0.27 0.40 0.48 Typical area cultivated in cotton (2- lima) 9 4 3 1 Total cash earnings (ZKW) 789,100 254,408 505,336 115,280 48 Table A1. Detailed crop budgets by farmer type and year Groups 1 2 3 4 Exchange rate 3,200 3,200 3,200 3,200 Total cash earnings (USD) 247 80 158 36 2004/05 Price= 1,220 1,220 1,220 1,220 Input cost = 100,000 100,000 65,709 54,278 cotton farm budget 1a 1b 2 3 Total production/2-lima 600 525 300 225 Price seed cotton (ZK/kg) 1,220 1,220 1,220 1,220 Gross income 732,000 640,500 366,000 274,500 Input cost 100,000 100,000 65,709 54,278 Hired services (AT + pump) 0 74,000 15,500 11,000 Hired labor 286,500 200,000 0 0 Amortization of equip, animals 35,822 8,648 5,346 10,692 Total cash cost 422,322 382,648 86,555 75,970 Net cash income (excluding family labor) 309,678 257,852 279,445 198,530 Returns per day of family labor (ZKW) 16,385 16,529 4,697 3,660 Returns per day of family labor (USD) 3.56 3.59 1.02 0.80 Implicit value family labor 5,000 5,000 5,000 5,000 Total value family labor 94,500 78,000 297,500 271,250 Net income (including value family labor) 215,178 179,852 -18,055 -72,720 Full cost of seed cotton (ZKW/kg) 861 877 1,280 1,543 Full cost of seed cotton (USD) 0.19 0.19 0.28 0.34 Typical area cultivated in cotton (2- lima) 9 4 3 1 Total cash earnings (ZKW) 2,787,100 1,031,408 838,336 198,530 Exchange rate 4,600 4,600 4,600 4,600 49 Table A1. Detailed crop budgets by farmer type and year Groups 1 2 3 4 Total cash earnings (USD) 606 224 182 43 Notes: 1. We assume that no hired labor is unsupervised, and that supervision/joint work by family = 20 percent of the hired labor days 2. The difference between 1a and 1b is that the hhs in group 1b did not own oxen, they only hired a ridger to do ridging as single operation 3. Hired labor for sprays - ZK2000 per 1 knapsack sprayer and there were 6 knapsacks per acre hence having a total charge of ZK12000 4. There was mixed labor for group 1 while the other groups 2 and 3 used only family labor for weedings 5. Groups 2 and 3 do not have sprayers, they just rent 50 Table A2. Summary crop budgets by farmer type and year Group 1 2 3 4 Area weighted mean Share of cotton farmers 9% 15% 42% 34% Share of cotton area 27% 20% 42% 11% Share of cotton prod'n 36% 26% 32% 6% Area in cotton (ha) 4.5 2.0 1.5 0.5 Cotton Yield (kg/ha) 1,200 1,050 600 450 836 Animals, plough, ridger, Equipment Sprayer, handhoes Handhoes Handhoes sprayer, ox-cart, handhoes Labor All hired Mostly hired Mostly family All family Revenue Indicators ZKW USD ZKW USD ZKW USD ZKW USD ZKW USD Gross revenue 2006/07 630,000 150 551,250 131 315,000 75 236,250 56 438,638 104 2005/06 510,000 159 446,250 139 255,000 80 191,250 60 355,088 111 2004/05 732,000 159 640,500 139 366,000 80 274,500 60 509,655 111 Input cost/gross revenue 2006/07 0.10 0.12 0.14 0.15 0.13 2005/06 0.20 0.22 0.26 0.28 0.24 2004/05 0.14 0.16 0.18 0.20 0.17 Gross margin (excl. fam. lab) 2006/07 242,428 58 203,352 48 251,279 60 179,141 43 231,369 55 2005/06 87,678 27 63,602 20 168,445 53 115,280 36 119,821 37 2004/05 309,678 67 257,852 56 279,445 61 198,530 43 274,389 60 Returns/day family labor 2006/07 12,827 3.05 13,035 3.10 4,223 1.01 3,302 0.79 8,207 1.95 2005/06 4,639 1.45 4,077 1.27 2,831 0.88 2,125 0.66 3,491 1.09 2004/05 16,385 3.56 16,529 3.59 4,697 1.02 3,660 0.80 10,105 2.20 Net margin (after fam. labor) 2006/07 147,928 35 125,352 30 -46,221 -11 -92,109 -22 35,466 8 2005/06 -6,822 -2 -14,398 -4 -129,055 -40 -155,970 -49 -76,081 -24 2004/05 215,178 47 179,852 39 -18,055 -4 -72,720 -16 78,486 17 Total Cost/kg 2006/07 803 0.19 811 0.19 1,204 0.29 1,459 0.35 1,045 0.25 2005/06 861 0.27 877 0.27 1,280 0.40 1,543 0.48 1,115 0.35 2004/05 861 0.19 877 0.19 1,280 0.28 1,543 0.34 1,115 0.24 Gross margin from all cotton 2006/07 2,181,850 519 813,408 194 753,837 179 179,141 43 1,088,098 259 2005/06 789,100 247 254,408 80 505,336 158 115,280 36 488,860 153 2004/05 2,787,100 606 1,031,408 224 838,336 182 198,530 43 1,332,738 290 51 Annex 2: Regressions - Yield Regression to estimate area-adjusted yield advantage of households owning animal traction Source: Author`s calculations from CSO 2003/04 Post-Harvest Survey data Model Summary Adjusted R Std. Error of Model R R Square Square the Estimate 1 .285(a) .081 .081 534.46467 a Predictors: (Constant), prov8ownat, hect, prov3ownat, prov1ownat, prov3, prov8, prov1 Coefficients(a) Unstandardized Standardized Coefficients Coefficients t Sig. B Std. Error Beta B Std. Error (Constant) 1097.198 19.310 56.821 .000 Hectares of cotton -126.219 2.244 -.191 -56.255 .000 Central Prov -18.803 19.949 -.013 -.943 .346 Eastern Prov -79.954 19.378 -.066 -4.126 .000 Southern Prov -236.277 20.713 -.130 -11.407 .000 Central Prov* own AT 43.480 8.484 .021 5.125 .000 Eastern Prov* own AT 322.354 5.251 .214 61.389 .000 Southern Prov* own AT 13.266 11.319 .005 1.172 .241 a Dependent Variable: yield 52 Annex 3: Regressions ­ Income Net Farm Income (OLS) Total Household Income (OLS) Prob. of off-farm income (logit) Variable Coefficient Std. Error P-value Coefficient Std. Error P-value Coefficient Std. Error P-value (Constant) 12.91 0.066 0.000 *** 13.151 0.082 0.000 *** -1.33 0.13 0.000 *** loghatotal 0.93 0.022 0.000 *** 0.726 0.028 0.000 *** -0.07 0.03 0.007 *** Logit: linear, not log logasset 0.02 0.002 0.000 *** 0.029 0.002 0.000 *** 0.00 0.00 0.026 ** Logit: linear, not log femhead 0.01 0.040 0.859 -0.071 0.050 0.160 0.20 0.12 0.090 * logedhhh 0.01 0.005 0.003 *** 0.038 0.006 0.000 *** 0.11 0.01 0.000 *** Logit: linear, not log lognchild -0.02 0.006 0.001 ** -0.015 0.007 0.034 ** 0.02 0.03 0.588 Logit: linear, not log loghhsize 0.25 0.037 0.000 *** 0.277 0.046 0.000 *** 0.04 0.02 0.067 * Logit: linear, not log qnt1 0.26 0.089 0.004 *** 0.045 0.112 0.688 -0.10 0.26 0.708 qnt2 0.20 0.061 0.001 *** 0.011 0.076 0.885 -0.57 0.20 0.004 *** qnt3 0.10 0.063 0.099 * -0.059 0.079 0.452 -0.34 0.20 0.085 * qnt4 0.01 0.085 0.903 -0.154 0.106 0.145 -0.55 0.26 0.033 ** qnt5 0.06 0.064 0.317 -0.002 0.080 0.977 -0.47 0.20 0.016 ** central -0.06 0.041 0.164 0.201 0.051 0.000 *** 0.80 0.12 0.000 *** lusaka -0.08 0.067 0.235 0.783 0.084 0.000 *** 1.75 0.23 0.000 *** southern -0.10 0.039 0.009 *** 0.092 0.049 0.060 * 0.62 0.11 0.000 *** Adj. R-square 0.638 0.479 Cox & Snell R-square 0.122 N 2275 2275 2360 Dep. Variable log net farm income Log net household income 0,1 has off-farm income Source: Author`s calculations from 2004 CSO/FSRP Supplemental Survey 53 Africa Region Working Paper Series Series # Title Date Author ARWPS 1 Progress in Public Expenditure Management in January 1999 C. Kostopoulos Africa: Evidence from World Bank Surveys ARWPS 2 Toward Inclusive and Sustainable Development March 1999 Markus Kostner in the Democratic Republic of the Congo ARWPS 3 Business Taxation in a Low-Revenue Economy: June 1999 Ritva Reinikka A Study on Uganda in Comparison with Duanjie Chen Neighboring Countries ARWPS 4 Pensions and Social Security in Sub-Saharan October 1999 Luca Barbone Africa: Issues and Options Luis-A. Sanchez B. ARWPS 5 Forest Taxes, Government Revenues and the January 2000 Luca Barbone Sustainable Exploitation of Tropical Forests Juan Zalduendo ARWPS 6 The Cost of Doing Business: Firms` Experience June 2000 Jacob Svensson with Corruption in Uganda ARWPS 7 On the Recent Trade Performance of Sub- August 2000 Francis Ng and Saharan African Countries: Cause for Hope or Alexander J. Yeats More of the Same ARWPS 8 Foreign Direct Investment in Africa: Old Tales November Miria Pigato and New Evidence 2000 ARWPS 9 The Macro Implications of HIV/AIDS in South November Channing Arndt Africa: A Preliminary Assessment 2000 Jeffrey D. Lewis ARWPS 10 Revisiting Growth and Convergence: Is Africa December C. G. Tsangarides Catching Up? 2000 ARWPS 11 Spending on Safety Nets for the Poor: How January 2001 William J. Smith Much, for How Many? The Case of Malawi ARWPS 12 Tourism in Africa February 2001 Iain T. Christie D. E. Crompton ARWPS 13 Conflict Diamonds February 2001 Louis Goreux ARWPS 14 Reform and Opportunity: The Changing Role March 2001 Jeffrey D. Lewis and Patterns of Trade in South Africa and SADC ARWPS 15 The Foreign Direct Investment Environment in March 2001 Miria Pigato Africa ARWPS 16 Choice of Exchange Rate Regimes for April 2001 Fahrettin Yagci Developing Countries ARWPS 18 Rural Infrastructure in Africa: Policy Directions June 2001 Robert Fishbein ARWPS 19 Changes in Poverty in Madagascar: 1993-1999 July 2001 S. Paternostro J. Razafindravonona David Stifel 54 Africa Region Working Paper Series Series # Title Date Author ARWPS 20 Information and Communication Technology, August 2001 Miria Pigato Poverty, and Development in sub-Sahara Africa and South Asia ARWPS 21 Handling Hierarchy in Decentralized Settings: September Navin Girishankar A. Governance Underpinnings of School 2001 Alemayehu Performance in Tikur Inchini, West Shewa Zone, Yusuf Ahmad Oromia Region ARWPS 22 Child Malnutrition in Ethiopia: Can Maternal October 2001 Luc Christiaensen Knowledge Augment The Role of Income? Harold Alderman ARWPS 23 Child Soldiers: Preventing, Demobilizing and November Beth Verhey Reintegrating 2001 ARWPS 24 The Budget and Medium-Term Expenditure December David L. Bevan Framework in Uganda 2001 ARWPS 25 Design and Implementation of Financial January 2002 Guenter Heidenhof Management Systems: An African Perspective H. Grandvoinnet Daryoush Kianpour B. Rezaian ARWPS 26 What Can Africa Expect From Its Traditional February 2002 Francis Ng Exports? Alexander Yeats ARWPS 27 Free Trade Agreements and the SADC February 2002 Jeffrey D. Lewis Economies Sherman Robinson Karen Thierfelder ARWPS 28 Medium Term Expenditure Frameworks: From February 2002 P. Le Houerou Concept to Practice. Preliminary Lessons from Robert Taliercio Africa ARWPS 29 The Changing Distribution of Public Education February 2002 Samer Al-Samarrai Expenditure in Malawi Hassan Zaman ARWPS 30 Post-Conflict Recovery in Africa: An Agenda for April 2002 Serge Michailof the Africa Region Markus Kostner Xavier Devictor ARWPS 31 Efficiency of Public Expenditure Distribution May 2002 Xiao Ye and Beyond: A report on Ghana`s 2000 Public S. Canagaraja Expenditure Tracking Survey in the Sectors of Primary Health and Education ARWPS 34 Putting Welfare on the Map in Madagascar August 2002 Johan A. Mistiaen Berk Soler T. Razafimanantena J. Razafindravonona ARWPS 35 A Review of the Rural Firewood Market Strategy August 2002 Gerald Foley in West Africa P. Kerkhof, D. Madougou 55 Africa Region Working Paper Series Series # Title Date Author ARWPS 36 Patterns of Governance in Africa September Brian D. Levy 2002 ARWPS 37 Obstacles and Opportunities for Senegal`s September Stephen Golub International Competitiveness: Case Studies of 2002 Ahmadou Aly Mbaye the Peanut Oil, Fishing and Textile Industries ARWPS 38 A Macroeconomic Framework for Poverty October 2002 S. Devarajan Reduction Strategy Papers : With an Application Delfin S. Go to Zambia ARWPS 39 The Impact of Cash Budgets on Poverty November Hinh T. Dinh Reduction in Zambia: A Case Study of the 2002 Abebe Adugna Conflict between Well Intentioned Bernard Myers Macroeconomic Policy and Service Delivery to the Poor ARWPS 40 Decentralization in Africa: A Stocktaking Survey November Stephen N. Ndegwa 2002 ARWPS 41 An Industry Level Analysis of Manufacturing December Professor A. Mbaye Productivity in Senegal 2002 ARWPS 42 Tanzania`s Cotton Sector: Constraints and December John Baffes Challenges in a Global Environment 2002 ARWPS 43 Analyzing Financial and Private Sector Linkages January 2003 Abayomi Alawode in Africa ARWPS 44 Modernizing Africa`s Agro-Food System: February 2003 Steven Jaffee Analytical Framework and Implications for Ron Kopicki Operations Patrick Labaste Iain Christie ARWPS 45 Public Expenditure Performance in Rwanda March 2003 Hippolyte Fofack C. Obidegwu Robert Ngong ARWPS 46 Senegal Tourism Sector Study March 2003 Elizabeth Crompton Iain T. Christie ARWPS 47 Reforming the Cotton Sector in SSA March 2003 Louis Goreux John Macrae ARWPS 48 HIV/AIDS, Human Capital, and Economic April 2003 Channing Arndt Growth Prospects for Mozambique ARWPS 49 Rural and Micro Finance Regulation in Ghana: June 2003 William F. Steel Implications for Development and Performance David O. Andah of the Industry ARWPS 50 Microfinance Regulation in Benin: Implications June 2003 K. Ouattara of the PARMEC LAW for Development and Performance of the Industry ARWPS 51 Microfinance Regulation in Tanzania: June 2003 Bikki Randhawa Implications for Development and Performance Joselito Gallardo of the Industry 56 Africa Region Working Paper Series Series # Title Date Author ARWPS 52 Regional Integration in Central Africa: Key June 2003 Ali Zafar Issues Keiko Kubota ARWPS 53 Evaluating Banking Supervision in Africa June 2003 Abayomi Alawode ARWPS 54 Microfinance Institutions` Response in Conflict June 2003 Marilyn S. Manalo Environments: Eritrea- Savings and Micro Credit Program; West Bank and Gaza ­ Palestine for Credit and Development; Haiti ­ Micro Credit National, S.A. AWPS 55 Malawi`s Tobacco Sector: Standing on One June 2003 Steven Jaffee Strong leg is Better than on None AWPS 56 Tanzania`s Coffee Sector: Constraints and June 2003 John Baffes Challenges in a Global Environment AWPS 57 The New Southern AfricanCustoms Union June 2003 Robert Kirk Agreement Matthew Stern AWPS 58a How Far Did Africa`s First Generation Trade June 2003 Lawrence Hinkle Reforms Go? An Intermediate Methodology for A. Herrou-Aragon Comparative Analysis of Trade Policies Keiko Kubota AWPS 58b How Far Did Africa`s First Generation Trade June 2003 Lawrence Hinkle Reforms Go? An Intermediate Methodology for A. Herrou-Aragon Comparative Analysis of Trade Policies Keiko Kubota AWPS 59 Rwanda: The Search for Post-Conflict Socio- October 2003 C. Obidegwu Economic Change, 1995-2001 AWPS 60 Linking Farmers to Markets: Exporting Malian October 2003 Morgane Danielou Mangoes to Europe Patrick Labaste J-M. Voisard AWPS 61 Evolution of Poverty and Welfare in Ghana in October 2003 S. Canagarajah the 1990s: Achievements and Challenges Claus C. Pörtner AWPS 62 Reforming The Cotton Sector in Sub-Saharan November Louis Goreux Africa: SECOND EDITION 2003 AWPS 63 (E) Republic of Madagascar: Tourism Sector Study November Iain T. Christie 2003 D. E. Crompton AWPS 63 (F) République de Madagascar: Etude du Secteur November Iain T. Christie Tourisme 2003 D. E. Crompton AWPS 64 Migrant Labor Remittances in Africa: Reducing Novembre Cerstin Sander Obstacles to Development Contributions 2003 Samuel M. Maimbo AWPS 65 Government Revenues and Expenditures in January 2004 Francisco G. Carneiro Guinea-Bissau: Casualty and Cointegration Joao R. Faria Boubacar S. Barry 57 Africa Region Working Paper Series Series # Title Date Author AWPS 66 How will we know Development Results when June 2004 Jody Zall Kusek we see them? Building a Results-Based Ray C. Rist Monitoring and Evaluation System to Give us the Elizabeth M. White Answer AWPS 67 An Analysis of the Trade Regime in Senegal June 2004 Alberto Herrou- (2001) and UEMOA`s Common External Trade Arago Policies Keiko Kubota AWPS 68 Bottom-Up Administrative Reform: Designing June 2004 Talib Esmail Indicators for a Local Governance Scorecard in Nick Manning Nigeria Jana Orac Galia Schechter AWPS 69 Tanzania`s Tea Sector: Constraints and June 2004 John Baffes Challenges AWPS 70 Tanzania`s Cashew Sector: Constraints and June 2004 Donald Mitchell Challenges in a Global Environment AWPS 71 An Analysis of Chile`s Trade Regime in 1998 July 2004 Francesca Castellani and 2001: A Good Practice Trade Policy A. Herrou-Arago Benchmark Lawrence E. Hinkle AWPS 72 Regional Trade Integration inEast Africa: Trade August 2004 Lucio Castro and Revenue Impacts of the Planned East African Christiane Kraus Community Customs Union Manuel de la Rocha AWPS 73 Post-Conflict Peace Building in Africa: The August 2004 Chukwuma Challenges of Socio-Economic Recovery and Obidegwu Development AWPS 74 An Analysis of the Trade Regime in Bolivia August 2004 Francesca Castellani in2001: A Trade Policy Benchmark for low Alberto Herrou- Income Countries Aragon Lawrence E. Hinkle AWPS 75 Remittances to Comoros- Volumes, Trends, October 2004 Vincent da Cruz Impact and Implications Wolfgang Fendler Adam Schwartzman AWPS 76 Salient Features of Trade Performance in Eastern October 2004 Fahrettin Yagci and Southern Africa Enrique Aldaz- Carroll AWPS 77 Implementing Performance-Based Aid in Africa November Alan Gelb 2004 Brian Ngo Xiao Ye AWPS 78 Poverty Reduction Strategy Papers: Do they December Rene Bonnel matter for children and Young people made 2004 Miriam Temin vulnerable by HIV/AIDS? Faith Tempest AWPS 79 Experience in Scaling up Support to Local December Jean Delion Response in Multi-Country Aids Programs (map) 2004 Pia Peeters in Africa Ann Klofkorn Bloome 58 Africa Region Working Paper Series Series # Title Date Author AWPS 80 What makes FDI work? A Panel Analysis of the February 2005 Kevin N. Lumbila Growth Effect of FDI in Africa AWPS 81 Earnings Differences between Men and Women February 2005 Kene Ezemenari in Rwanda Rui Wu AWPS 82 The Medium-Term Expenditure Framework: The April 2005 Chukwuma Challenge of Budget Integration in SSA Obidegwu countries AWPS 83 Rules of Origin and SADC: The Case for change June 2005 Paul Brenton in the Mid Term Review of the Trade Protocol Frank Flatters Paul Kalenga AWPS 84 Sexual Minorities, Violence and AIDS in Africa July 2005 Chukwuemeka Anyamele Ronald Lwabaayi Tuu-Van Nguyen, and Hans Binswanger AWPS 85 Poverty Reducing Potential of Smallholder July 2005 Paul B. Siegel Agriculture in Zambia: Opportunities and Jeffrey Alwang Constraints AWPS 86 Infrastructure, Productivity and Urban Dynamics July 2005 Zeljko Bogetic in Côte d`Ivoire An empirical analysis and policy Issa Sanogo implications AWPS 87 Poverty in Mozambique: Unraveling Changes August 2005 Louise Fox and Determinants Elena Bardasi, Katleen V. Broeck AWPS 88 Operational Challenges: Community Home August 2005 N. Mohammad Based Care (CHBC) forPLWHA in Multi- Juliet Gikonyo Country HIV/AIDS Programs (MAP) forSub- Saharan Africa AWPS 90 Kenya: Exports Prospects and Problems September Francis Ng 2005 Alexander Yeats AWPS 91 Uganda: How Good a Trade Policy Benchmark September Lawrence E. Hinkle for Sub-Saharan-Africa 2005 Albero H. Aragon Ranga Krishnamani Elke Kreuzwieser AWPS 92 Community Driven Development in South October 2005 David Everatt Lulu Africa, 1990-2004 Gwagwa AWPS 93 The Rise of Ghana``s Pineapple Industry from November Morgane Danielou Successful take off to Sustainable Expansion 2005 Christophe Ravry AWPS 94 South Africa: Sources and Constraints of Long- December Johannes Fedderke Term Growth, 1970-2000 2005 AWPS 95 South Africa``s Export Performance: December Lawrence Edwards Determinants of Export supply 2005 Phil Alves AWPS 96 Industry Concentration in South African December Gábor Szalontai Manufacturing: Trends and Consequences, 1972- 2005 Johannes Fedderke 96 59 Africa Region Working Paper Series Series # Title Date Author AWPS 97 The Urban Transition in Sub-Saharan Africa: December Christine Kessides Implications for Economic Growth and Poverty 2005 Reduction AWPS 98 Measuring Intergovernmental Fiscal Performance May 2006 Navin Girishankar in South Africa David DeGroot Issues in Municipal Grant Monitoring T.V. Pillay AWPS 99 Nutrition and Its determinants in Southern July 2006 Jesper Kuhl Ethiopia - Findings from the Child Growth Luc Christiaensen Promotion Baseline Survey AWPS 100 The Impact of Morbidity and Mortality on September Zara Sarzin Municipal Human Resources and Service 2006 Delivery AWPS 101 Rice Markets in Madagascar in Disarray: September Bart Minten Policy Options for Increased Efficiency and Price 2006 Paul Dorosh Stabilization Marie-Hélène Dabat, Olivier Jenn-Treyer, John Magnay and Ziva Razafintsalama AWPS 102 Riz et Pauvrete a Madagascar Septembre Bart Minten 2006 AWPS 103 ECOWAS- Fiscal Revenue Implications of the April 2007 Simplice G. Zouhon- Prospective Economic Partnership Agreement Bi with the EU Lynge Nielsen AWPS 104(a) Development of the Cities of Mali June 2007 Catherine Farvacque- Challenges and Priorities V. Alicia Casalis Mahine Diop Christian Eghoff AWPS 104(b) Developpement des villes Maliennes June 2007 Catherine Farvacque- Enjeux et Priorites V. Alicia Casalis Mahine Diop Christian Eghoff AWPS 105 Assessing Labor Market Conditions In June 2007 David Stifel Madagascar, 2001-2005 Faly H. Rakotomanana Elena Celada AWPS 106 An Evaluation of the Welfare Impact of Higher June 2007 Noro Andriamihaja Energy Prices in Madagascar Giovanni Vecchi AWPS 107 The Impact of The Real Exchange Rate on November Mireille Linjouom Manufacturing Exports in Benin 2007 AWPS 108 Building Sector concerns into Macroeconomic December Antonio Estache Financial Programming: Lessons from Senegal 2007 Rafael Munoz and Uganda 60 Africa Region Working Paper Series Series # Title Date Author AWPS 109 An Accelerating Sustainable, Efficient and December Hans P. Binswanger Equitable Land Reform: Case Study of the 2007 Roland Henderson Qedusizi/Besters Cluster Project Zweli Mbhele Kay Muir-Leresche AWPS 110 Development of the Cites of Ghana January 2008 Catherine Farvacque- ­ Challenges, Priorities and Tools Vitkovic Madhu Raghunath Christian Eghoff Charles Boakye AWPS 111 Growth, Inequality and Poverty in Madagascar, April 2008 Nicolas Amendola 2001-2005 Giovanni Vecchi AWPS 112 Labor Markets, the Non-Farm Economy and April 2008 David Stifel Household Livelihood Strategies in Rural Madagascar AWPS 113 Profile of Zambia`s Smallholders: Where and June 2008 Paul B. Siegel Who are the Potential Beneficiaries of Agricultural Commercialization? AWPS 114 Promoting Sustainable Pro-Poor Growth in June 2008 Michael Morris Rwandan Agriculture: What are the Policy Liz Drake Options? Kene Ezemenary Xinshen Diao AWPS 115 The Rwanda Industrial and Mining Survey June 2008 Tilahun Temesgen (RIMS), 2005 Survey Report and Major Findings Kene Ezemenari Louis Munyakazi Emmanuel Gatera AWPS 116 Taking Stock of Community Initiatives in the June 2008 Jean Delion Fight against HIV/AIDS in Africa: Experience, Elizabeth Ninan Issues, and Challenges AWPS 117 Travaux publics à Haute Intensité de Main d` August 2008 Nirina H. Oeuvre (HIMO) pour la Protection Sociale à Andrianjaka Madagascar : Problèmes et Options de Politique Annamaria Milazzo AWPS 118 Madagascar : De Jure labor Regulations and August 2008 Gaelle Pierre Actual Investment Climate Constraints AWPS 119 Tax Compliance Costs for Businesses in South August 2008 Jacqueline Coolidge Africa, Provincial Analysis Domagoj Ilic Gregory Kisunko AWPS 120 Umbrella Restructuring of a Multicountry October 2008 Nadeem Mohammad Program (Horizontal APL) Restructuring the Norbert Mugwagwa Multicountry HIV>AIDS Program (MAP) in Africa AWPS 121 Comparative Analysis of Organization and October 2008 Gérald Estur Performance of African Cotton Sectors AWPS 122 The Cotton Sector of Zimbabwe February 2009 Colin Poulton Benjamine Hanyani- Mlambo 61 Africa Region Working Paper Series Series # Title Date Author AWPS 123 The Cotton Sector of Uganda March 2009 John Baffes AWPS 124 The Cotton Sector of Zambia March 2009 David Tschirley/ Stephen Kabwe 62