Coffee_cost_benefit_analysis.pdf

Ethiopia AGP-AMDe Cost Benefit Analysis on Coffee Value Chain, October 2012 1

AGRICULTURAL GROWTH PROJECT—

AGRIBUSINESS AND MARKET DEVELOPMENT

(AGP-AMDe) PROJECT

COFFEE VALUE CHAIN: COST BENEFIT ANALYSIS

OF INTERVENTION

FINAL REPORT

This report was produced for review by the United States Agency for International Development (USAID). It was prepared by Optimal Solutions Group, LLC for USAID’s “Learning, Analysis, and

Evaluation Project (LEAP). Contract Number: AID-OAA-C-11-00169

Ethiopia AGP-AMDe Cost Benefit Analysis on Coffee Value Chain, October 2012

UNITED STATES AGENCY FOR INTERNATIONAL DEVELOPMENT

LEARNING EVALUATION ANALYSIS PROJECT

(AID-OAA-C-11-0169)

AGRICULTURAL GROWTH PROJECT—AGRIBUSINESS AND

MARKET DEVELOPMENT (AGP-AMDe) PROJECT COFFEE VALUE

CHAIN: COST BENEFIT-ANALYSIS OF INTERVENTIONS

FINAL REPORT

Prepared for:

The United States Agency for International Development (USAID)/Ethiopia

Prepared by:

Optimal Solutions Group, LLC University of Maryland Research Park, M Square 5825 University Research Court, Suite 2800 College Park, MD 20740 USA

Cambridge Resources International Inc.

60 Montgomery Street Cambridge, MA 02141 USA www.cri-world.com

Project Analysts:

Christopher Bennett, Senior Coffee Specialist Richard Barichello, Senior Cost-Benefit Analyst Leonard Leung, Cost-Benefit Analyst Habtamu Fuje, Local Consultant Dr. Glenn Jenkins, Cost-Benefit Manager Christabel Dadzie, Evaluation Analyst

October 24, 2012

Disclaimer

The authors’ views expressed in this publication do not necessarily reflect the views of the United States Agency for International Development or the United States Government.

http:www.cri-world.com http:www.OptimalSolutionsGroup.com

Contents

ACKNOWLEDGMENTS

LIST OF ACRONYMS

EXECUTIVE SUMMARY REPORT—COFFEE

Project Description

Interventions Description

Strategic Context and Rationale

Data Source

Results for Financial and Economic Analysis on High‐Yield Variety Coffee Cultivation

Results for Financial and Economic Analysis on a Single Nursery

Analysis on the Increased Traceability Interventions

Conclusions and Recommendations

High‐Yield Variety Coffee Cultivation

Increased Traceability Interventions

INTRODUCTION

INTERVENTION 1 — PRODUCTION AND MARKETING OF HYV SEEDLINGS

Part I — Financial Analysis of the Cultivation of High-Yield Variety

COFFEE-SECTOR INTERVENTIONS DESCRIPTION

Seedlings

Model Farm Description and Parameter Values

Financial Analysis Findings

Part I — Economic Analysis of the Cultivation of High-Yield Variety

Seedlings

Motivation

Downstream Externalities and Stakeholders

Economic Analysis Findings

Sensitivity Analysis Findings

Part II — Financial Analysis of a Seedling Nursery

Intervention Description

Nursery Model Description

Part II — Economic Analysis of an Integrated Program for the

Conversion of Ethiopian Coffee Production to High-Yield Varieties .. 36

Introduction

Section 1: Economic Analysis of a High‐Yield Variety Seedling Nursery

Section 2: Economic Appraisal of the Conversion Program for Coffee Cultivation

Conclusions and Recommendations

INTERVENTION 2 — IMPROVING COFFEE TRACEABILTIY

Contextual Background of the Intervention

Weaknesses of the Current Export Regime

Information Clouding in ECX Coffee Contracts

First‐In‐First‐Out Warehouse Storage and Bulking

Specification Mismatch at Delivery

Banning the Export of Tradable Coffee

Interventions Description

Analysis Findings

Conclusions and Recommendations

APPENDIX 1 — GENERAL RECOMMENDATIONS ON THE COFFEE VALUE

CHAIN

REFERENCES

ACKNOWLEDGMENTS

The LEAP team would like to thank the many individuals who have assisted the team in undertaking this study. The information for this study was obtained through the support of many stakeholders at all levels of the coffee value chain, and assistance from USAID/Ethiopia and ACDI-VOCA especially have greatly eased the data-collection efforts for the study team. Particularly, ACDI-VOCA’s data on farm budget, physical and labor input requirement by farming technology, and its nursery budget provide the backbone of the farm and nursery models constructed for the financial and economic analyses in this report. The Coffee Research Agency also provided invaluable yield estimates and commentary.

A major data-collection effort was carried out during field visits in Jimma and follow-up interviews with industry stakeholders in Addis Ababa. Especially helpful are the data provided by the Ethiopian Coffee Producers and Exporters Association regarding the investment and operating costs as well as the physical and labor input requirements of cherry-processing stations (both dry and wet methods); farm-level data provided by three smallholder farmers in villages around Jimma; coffee purchase data by a sebsabi (a coffee collector who buys directly from farmers) and a chaanyii (a supplier who processes and aggregates coffee for sales on the trading floor at the Ethiopian Commodity Exchange [ECX]). The Jimma regional office of the ECX, Coffee Exporters Association, Cooperative Unions and Cooperatives, individual large-scale growers also provided useful background information.

Despite acknowledging all the assistance provided, the study team claims full responsibility for the results and opinions expressed in this report.

LIST OF ACRONYMS

AGP-AMDe Agricultural Growth Program-Agribusiness and Market

Development

CLU Coffee Liquoring Unit

CBA Cost-benefit analysis

CBD Coffee-berry disease

CRI Cambridge Resources International

ECX Ethiopian Commodity Exchange

FEP Foreign exchange premium

GBE Green-bean equivalence

Ha Hectare

HYV High-yield variety

IP Identity preservation

JARC Jimma Agricultural Research Center kg Kilogram

NPV Net present value

OSG Optimal Solution Group

VAT Value-added tax

EXECUTIVE SUMMARY REPORT—COFFEE

Project Description The Agricultural Growth Program-Agribusiness and Market Development (AGP-AMDe) Project for Ethiopia is a component of the comprehensive Feed the Future (FtF) strategy developed by the USAID Mission for food-insecure developing countries. The main goals of AGP-AMDe include reducing poverty and hunger by improving the productivity and competitiveness of value chains that could give rural households greater opportunities for employment and increased income. For the duration of this project, USAID plans to target six specific commodity value chains: maize, wheat, coffee, sesame, chickpeas, and honey. The USAID Mission/Ethiopia will begin implementing the AGP-AMDe project in 83 woredas around Ethiopia in 2012. Of the 83 AGP-AMDe woredas, 16 are coffee-producing and are located in the regions of Oromiya and SNNP.

The total budget for the project spans a period of 5 years. At a 12 percent discount rate, it is equal to US$6.7 million in present value and will cover four components of the value chain: value chain competitiveness, access to financing, enabling environment, and innovation and investment. At the time of research, it was unclear how the budget would be divided between activities under each component.

Interventions Description Although the AGP-AMDe work plan proposes multiple interventions for the coffee sector, not all are suitable for evaluation via a cost-benefit analysis (CBA), because their benefits cannot be identified with reasonable precision. The interventions selected for analysis in this study tackle two specific bottlenecks in the coffee value chain: the low productivity at the smallholder level and the general lack of coffee traceability, which suppresses coffee prices in the international market.

The production and marketing of high-yield variety seedlings to smallholder farmers is the first intervention to be analyzed in this report. At the present, demand for high-yield variety (HYV) seedlings outstrips supply, partly due to the insufficient number of seedling nurseries and partly due to the capacity constraint of the Jimma Agricultural Research Center (JARC), the government agency responsible for seed development and the production of improved seeds and seedlings. The study integrates the analysis of Activities 1.1–2.5 of Intervention III, Component 1, in the AGP-AMDe work plan. Activities 2.4–2.5 promote the merits of HYV coffee cultivation among smallholder farmers, while Activities 1.1–2.3 (and likely some undefined activities under Component 4) support improved variety development and the training of individuals in the production of seedlings to boasts the capacity of nurseries. Accordingly, in this analysis, the financial and economic benefits of cultivating HYV coffee are considered first, followed by a financial and economic analysis of the nurseries. Finally, the results of the financial and economic outcomes of the nurseries and the cultivation of coffee from conversion to HYVs are integrated for a more complete value chain analysis. This analysis includes an estimate of the net benefits received by the various stakeholders along the value chain.

The second intervention considered in this study is improved traceability of export coffee. The lack of traceability is a problem that currently plagues the coffee-export sector and deprives all value chain participants of an export premium that Ethiopian coffees really deserve, Ethiopia being the origin of Arabica coffee with its many distinctly different, high-quality coffee varieties. The intervention is not well defined but rather is mentioned in broad strokes in the AGP-AMDe work plan. It falls loosely under Activities 1.1–1.4 and 3.1–3.9 of Intervention I, Component 1. However, this intervention and related issues were raised in different contexts during the study team’s interviews with many stakeholders in the industry. The study team finds lack of traceability to be the most pressing problem of the ones raised in the work plan and deserving of special attention.

Strategic Context and Rationale The two interventions—namely, the production and marketing of HYV seedlings and improved coffee traceability—embody different strategies for raising coffee farmers’ income. The former boosts coffee production at the farmstead and therefore the quantity available for sale. The higher throughput also generates downstream externalities for laborers who work at the farmstead and coffee-cherry processing stations as temporary employees and for many intermediary agents engaged in aggregating, processing, and trading coffee. As business transactions increase, the government then collects more direct and indirect tax revenues from the taxable parties in the value chain.

By contrast, the second intervention attempts to increase coffee prices by facilitating an information flow to exporters regarding the coffee’s origins. During field visits to Jimma, Oromiya, it was evident that coffee farmers are well informed of the prevailing coffee prices at the Ethiopian Commodity Exchange (ECX). If this intervention succeeds in raising export coffee prices, it is likely that such an increase will be transmitted up the value chain and reflected in the coffee-cherry prices for smallholder farmers.

Data Source Support from many stakeholders at all levels of the coffee value chain, and especially from USAID/Ethiopia and ACDI-VOCA, have greatly eased the study team’s data-collection efforts. In particular, ACDI-VOCA’s data on farm budgets, physical and labor input requirements by farming technology, and nursery budgets provide the backbone of the farm and nursery models constructed for these financial and economic analyses.

In areas where ACDI-VOCA data were missing or ambiguous, the study team relied on cross-checks of the data with information collected during field visits in Jimma and follow-up interviews with industry stakeholders in Addis Ababa. Especially helpful were the data provided by the Ethiopian Coffee Producers and Exporters Association on the investment and operating costs as well as the physical and labor input requirements of cherry-processing stations (dry and wet methods); farm-level data provided by three smallholder farmers in villages around Jimma; coffee purchase data from a sebsabi (a cofee collector who buys directly from farmers) and a chaanyii (a supplier who processes and aggregates coffee for sales on the trading floor at the ECX) of their dealings with coffee farmers. In addition, a number of additional data sources, such as macrolevel statistics, have been taken from the Central Statistical Agency (CSA) of Ethiopia.

Results for Financial and Economic Analysis on High-Yield Variety Coffee Cultivation Part I of the analysis on the seedling production and marketing intervention estimates the incremental financial and economic impacts of HYV coffee cultivation for the existing traditional farming methods and the model farming technology recommended by the JARC. At the present, the JARC requires that the recipients of its seeds or seedlings adopt its recommended agronomic practices. Although the recommended practices do indeed yield a higher incremental financial net present value (NPV) for farmers, they impose a heavy cash outlay burden in the initial phase of the cultivation cycle when trees are unproductive and intercropping cannot be continued for as long because it requires more intensive utilization of physical and labor inputs to tend the trees. The greater total cash outlays required by the model farm program in the first 3 years amounts to 20 percent of the total income earned over the same duration by an average Oromiyan household with 5.2 members.

Results of the financial analysis are summarized in table A. On a per-farmer basis (each farm holding 1 hectare [Ha] of coffee), the model farming system yields the highest financial NPV of US$20,060 over the 20-year cultivation period, as compared to US$11,936 and US$5,857 for traditional agronomic practices using HYVs and local varieties, respectively. If traditional practices continue, each farmer earns an incremental benefit of US$6,079, but this amount increases to US$14,202 if the farmer adopts JCRA’s recommendation.

Table A. Financial NPV of coffee cultivation by farming technology (12 percent discount rate, 20-year project life)

Traditional practices Model farm

Local variety (1)

HYV (2) HYV (3)

ETB

Financial NPV per household 102,854 209,600 352,248 Incremental financial NPV

Scenario (1) as the base case 106,746 249,394

US$

Financial NPV per household 5,857 11,936 20,060 Incremental financial NPV

Scenario (1) as the base case 6,079 14,202

Although the farm model with HYV cultivation yields the highest financial NPV, it requires substantially more labour and physical inputs. In monetary terms, the model farm imposes a significant cash outlay burden of around US$1,000, equivalent to 20 percent of the potential total income of a typical Oromiyan household. This perhaps explain for the low adoption or conversion rate to model farming. Ethiopia could reach more than twice as many farmers with financial and input support for the low input approach, while ensuring significant income distribution and export revenue gains in comparison with present prevailing traditional practices.

HYV coffee cultivation boosts coffee production and increases the demand for laborers during harvests and the throughputs for traders. The government collects additional tax revenues from the value chain participants liable to various forms of taxes. The government also collects more trade taxes, because the incremental coffee export generates foreign exchanges, allowing for more imports by the private sectors.

The incremental benefits of HYV cultivation for various stakeholders are listed in table B. The total incremental economic benefit generated per farmer obtaining HYV seedlings is US$8,088 and US$19,742, depending on the type of farming methods that the farmer practices.

Results for Financial and Economic Analysis on a Single Nursery Although the work plan mentions multiple possible partners in the establishment of nurseries, this analysis takes female household heads as the trainees in the production of seedlings. Upon the completion of their training, they will be employed by farmer cooperatives to produce seedlings that will be sold to farmers. At an output price of 1.25 ETB per seedling and an annual production level and turnover of 100,000 seedlings, the financial analysis indicates that a nursery makes a modest profit of US$14,541 in present value over a 10-year project life.

Table B. Incremental economic NPV generated by a household with 1 Ha of coffee holding by farming technology (12 percent discount rate, 20-year project life)

Stakeholders (US$) Coffee farmers (1) Laborers

(2) Traders

(3) Govt (4)

Economy

(US$) (5)

Traditional practices: HYV (1) 6,079 342 1,020 647 8,088 Model farm: HYV (2) 14,202 1,092 2,721 1,727 19,742

Difference (3) = (2) – (1) 8,123 750 1,701 1,079 11,654

The skill-enhancing training and the employment opportunity draw the female household heads out of their homesteads. They are likely to receive a surplus from their labor at nurseries, because their market opportunity is narrower. The nurseries are formal establishments, given the assumption of cooperative ownership. An implication of this intervention is that the nurseries and their female employees pay business and personal income taxes. In addition, member farmers of cooperatives periodically receive dividends from the nurseries. Table C reports the financial return of a nursery and externalities on a per-nursery basis. It has been estimated from the ACDI-VOCA nursery budget data that a nursery employs

3.84 females, who collectively earn a surplus of US$3,247 in present value. The government also collects US$4,654 of tax revenues from the nursery and its employees.

Table C. Present values of externalities per nursery (12 percent discount rate, 10-year project life)

US$

Financial NPV of nursery 14,541 Present value of labor externalities

Nursery laborers 2,162 Managerial/administrative staff 1,085

Present value of government tax revenues

4,654

Total economic NPV of nursery 22,443

The figures above are expressed on a per-nursery basis. The actual benefits that the interventions may bring depend on the nursery to be formed, itself a function of the intended coverage target. Of the 83 AGP-AMDe woredas, 16 produce coffee and have a collective total of 119,028 Ha of coffee holdings. A nursery can supply seedlings to 55.6 households, each with 1 Ha of coffee holding practicing traditional cultivation methods. If the coverage target is 1,000 Ha per year, then a total of 18 nurseries must be established. At the end of the 10-year project life, 100,000 Ha, or 8.4 percent of existing coffee holdings, will be converted to HYVs. The economic benefits of the HYV seedling production and marketing intervention generated at the nursery level for varying coverage target are reported in table D.

Table D. Present values of externality by the number of households covered (12 percent discount rate, 10-year project life) Seedlings required per year (million seedlings) (1)

Nursery required

(2)

Female employees (3)

Externality at nursery (US$) (6)

Per nursery 22,443

Farm households covered per year 1,000 1.8 18.0 69.1 403,975 1,500 2.7 27.0 103.7 605,962

2,000 3.6 36.0 138.3 807,949

The benefit of the intervention, however, extends beyond the nursery level, because coffee production increases, generating more coffee sales for farmers, more employment opportunities for casual laborers, more throughput and business for traders, and more tax revenues for the government. Combining the results from table B (the incremental benefit of traditional farming) and table D yields the total incremental economic NPVs of this intervention for varying coverage targets, which are reported in table E. Although economic NPVs from US$51.6 to US$103.2 million seem exaggerated, on average the 100,000 to 200,000 beneficiary farmers who obtain the HYV seedlings earn an incremental benefit of US$3,837 that is spread over 20 years of coffee cultivation. This table completes the analysis for the first intervention.

Table E. Stakeholder analysis of the production and marketing of HYV seedlings intervention by coverage target (12 percent discount rate, 10-year project life)

Stakeholders (thousand US$) Coverage target (Ha/year)

Coffee farmers Nurseries

(1) Coffee cultivation (2)

Nursery female employees (3)

Farm laborers

(4)

Traders (5)

Govt (6)

ENPV

(million

US$)

(7)

1,000 262 38,469 58 2,162 6,458 4,181 51.6 1,500 393 57,703 88 3,242 9,686 6,271 77.4

2,000 523 76,938 117 4,323 12,915 8,362 103.2

Analysis on the Increased Traceability Interventions The proposed interventions for increasing coffee traceability focus on refining the grading precision of the ECX, through which most of the coffee must be traded and exported by law. However, during field visits and discussions with many stakeholders at different levels of the coffee value chain, the study team found that traceability erosion occurs not because of imprecise ECX coffee cupping and grading standards but rather because of ECX treatment of coffee as a bulk commodity. Many aspects of ECX operations, from the information-clouding design of the coffee contracts available on the ECX trading floor to the way coffee shipments to ECX warehouses are stored in bulk, exacerbate the problem. In light of this information, the study team determined that the proposed interventions would be misplaced and ineffectual in combating traceability erosion.

Conclusions and Recommendations

High-Yield Variety Coffee Cultivation The main conclusion of this analysis is that converting the existing traditional varieties of coffee grown in Ethiopia to HYVs is entirely justified financially as well as economically. In addition, at the present level of seedling prices, it is financially rewarding to produce seedlings in nurseries that have annual production capacities of 100,000 seedlings.

One of the most important issues investigated in this analysis is why this process is moving so slowly in Ethiopia. A partial answer to this question can be found in the financial analysis of the model farm program. The cash-flow barrier of reducing the amount of intercropping during the first 3 years before coffee sales begin—a feature of the model farm approach—is likely a major factor, because the government will not issue the HYV seedlings to farmers unless they commit to utilizing the model farm design.

From this analysis, it appears that the best return to USAID investment in this area is a broad-based approach to marketing HYV seeds that informs farmers about estimated input-to-yield responses under various agro-ecologies. However, the farmers should not be required to use any single farm model.

Similar instances of comparative economies of HYV tree crops in other countries (e.g., coffee, cocoa, rubber, oil palm, and coconuts) indicate that two to four times as many farmers can be reached (and more successfully) for a given investment budget when additional restrictions are not imposed than when the investment is contingent on the adoption of highest-yielding or highest-input scenarios.

Voluntary certification of companies producing genuine HYVs should be promoted. Informing end users of the difference between certified and noncertified HYV sellers encourages better performance through market competition. Mandatory licensing schemes, on the other hand, invite corruption as the inefficient producers seek to subvert the coercive system to stay in business.

Increased Traceability Interventions The Ethiopia coffee export quality problem is not that the coffee is low-grade, per se, but rather that its quality is uncertain. This problem is exacerbated by the loss of spatial and business traceability at the

ECX. The ECX has apparently improved its performance over the 3 years of its existence, but its approach to traceability is out of step with the way that global coffee markets are moving.

The current interventions being proposed in the AGP-AMDe work plan have the objective of improving the traceability of the coffee sold at the ECX. Although they have the correct objective in mind, the interventions are likely to be ineffective. Any additional information regarding the improved quality of the coffee produced by a particular set of farmers is likely to be lost. The grading and storage methods of the ECX are not designed to provide a sufficiently refined quality- and source-classification system to allow prospective buyers to be able to offer quality premium prices due to uncertainty concerning the quality of shipment and the lack of knowledge of the sellers and their reputations.

During an interview, ECX officials mentioned plans to take on sesame, chickpeas, and other commodities. Currently, the regional ECX facilities are already undercapitalized. Without a corresponding increase in infrastructure investment, an ECX expansion will only further constrain its capacity to manage the coffee trade. Also, the idea of storing coffee in silos rather than bags could further exacerbate the loss of identity preservation and increase the risk of spoilage, together adding to the likelihood of price discounts. Both measures should be discouraged. Instead, management policies with respect to the bulking and first-in-first-out sales and delivery of coffee should be rethought to allow for traceability of the coffee supply by prospective buyers. For start, the ECX should consider allowing sellers’ identities to be revealed in their ECX contracts on an optional basis. The shipments should also be stored separately to preserve the coffee’s complete (spatial and business) identity, at extra storage or handling costs or an ECX commission fee. Sellers and buyers who find it advantageous to do so will subscribe to the services, because they will not only allow the sellers to earn premiums on high-quality coffee but also will help overcome information asymmetry between smallholder farmers and traders. The latter will acquire more bargaining power and thus can earn higher incomes from their coffee.

INTRODUCTION

USAID/Ethiopia’s Agricultural Growth Program-Agribusiness and Market Development (AGP-AMDe) project focuses on strengthening market linkages in six agricultural value chains: coffee, honey, wheat, maize, chickpeas, and sesame. Cambridge Resources International (CRI) has been commissioned by Optimal Solution Group (OSG), the primary project contractor, to conduct cost-benefit analyses (CBA) on AGP-AMDe’s interventions on the coffee and honey value chains. The purpose of this report is to present findings concerning interventions on components of the coffee value chain.

COFFEE-SECTOR INTERVENTIONS DESCRIPTION

For the AGP-AMDe project, USAID/Ethiopia has contracted ACDI-VOCA to design the interventions and to coordinate their implementations. ACDI-VOCA’s proposal of interventions on the coffee sector is an umbrella approach that encompasses activities ranging from sponsoring a cupping competition to providing training to different coffee value chain participants. The interventions are broadly divided into four components: value chain competitiveness, access to finance, enabling environment, and innovation and investment. Table 1A presents the budget allocated to each component. Components 1 and 4 account for 48 percent and 30 percent of the total budget, respectively. Table 1B summarizes the total number of hectares (Ha) dedicated to coffee cultivation and annual coffee production in AGP-AMDe-targeted areas.

Table 1A. AGP-AMDe coffee-sector budget by project components (thousands US$)

ACDI-VOCA work plan for the coffee sector

Start up Year 1 Year 2 Year 3 Year 4 Year 5

Total in

PV

Component 1 Value chain competitiveness 72.2 760.5 885.9 967.1 957.9 519.7 3,334.6 Component 2

Access to finance 16.5 173.8 202.5 221.1 218.9 118.8 762.2 Component 3 Enabling environment 16.5 173.8 202.5 221.1 218.9 118.8 762.2 Component 4 Innovation and investment 0.0 335.8 929.9 688.5 362.1 97.8 2,034.7

Total 105.1 1,443.9 2,220.8 2,097.7 1,757.9 855.1 6,893.7 Source: AGP-AMDe Project Work Plan (ACDI-VOCA 2012a).

Because the intervention design is currently at the formative stage, the cost breakdown for each intervention and activity under the first component is not finalized, and the specific interventions under the second to the fourth components are not yet defined.

Table 1B: Coffee area and production in AGP-AMDe-targeted regions

Region Zone (1) Number of woredas

(2) Coffee area (Ha)

(3) Coffee production

(MT) (4)

Oromiya Jimma 3 39,496 23,142

Illubabor 3 19,434 8,899 Oromiya subtotal 6 58,930 32,041

SNNP Gedeo 2 7,517 6,591 Bench Maji

1 9,057 5,434

Omo 2 6,021 4,785 Kaffa 2 29,002 13,802 Gurage 1 4,645 3,298 Besketo 1 1,374 549 Dawro 1 2,482 664

SNNP subtotal 10 60,098 35,123

Total 16 119,028 67,164 Source: AGP-AMDe Coffee Value Chain Intervention Woredas and FCUs (ACDI-VOCA 2012b).

Among the interventions that are defined in sufficient details in the work plan, not all are suitable candidates for cost-benefit analyses (CBA), because the benefits that ensue from the interventions are not identifiable or quantifiable with an acceptable degree of accuracy. This study focuses on two interventions for which CBA would be helpful to assist in decision making. The remainder of this section gives an overview of the selected interventions, as conveyed by ACDI-VOCA staff during in-person meetings and in the work plan submitted to USAID/Ethiopia. Table 2 illustrates how the two selected interventions relate to those in the ACDI-VOCA work plan. Interventions under components 2 to 4 are not analyzed, because the Cambridge Resources International (CRI) team has been unable to secure any information about the relevant interventions, aside from the budget allocated for the respective components, as shown in table 1.

Table 2. Classification of CBA interventions according to ACDI-VOCA work plan

ACDI-VOCA work plan for the coffee sector

Component Intervention Activity

Intervention 1 Part I: HYV seedlings cultivation C1 III 2.4–2.5 Part II: Seedling nursery C1 III 1.1–2.3

C4

Intervention 2: Increased coffee traceability C1 I 1.1–1.4

3.1–3.9 Source: AGP-AMDe Project Work Plan (ACDI-VOCA 2012a).

Increasing coffee production through the production and marketing of HYV and disease-resistant seedlings to smallholder coffee farmers is the first intervention to be analyzed in this study. It falls under Intervention III, Component 1, in the ACDI-VOCA work plan. ACDI-VOCA plans to first identify and train local partners in targeted woredas (districts) in the production of HYV coffee seedlings. Secondly, using seeds acquired from local seed providers, the trained local partners cultivate and sell coffee seedlings to farmer cooperatives, who in turn sell the seedlings to smallholder coffee farmers. Many local partners will receive technical and financial aid and training under this plan, including existing seedling nurseries, farmer cooperatives, commercial farms, and individual female-headed households. Similarly, the work plan lists multiple sources of coffee-seed providers.

For the purpose of this study, the local partners are taken to be female household heads, and the seed provider is taken to be the Jimma Agricultural Research Center (JARC). Training an unknown number of female household heads is mentioned in the work plan (Activity 2.3 of Intervention III) and is likely to empower the trainees, who would otherwise have limited opportunities outside the home. ACDI-VOCA and the JARC have recently signed a memorandum of understanding; although the memorandum’s specific content is not known, given the JARC’s mission and area of expertise, it is likely to promote the provision or distribution of improved seeds or seedlings to coffee farmers.

For this intervention to succeed in the long run, the nurseries must be financially sustainable in the absence of external funding. Achieving financial sustainability, however, requires sufficient demand from farmers willing to incur extra expenses for the use and cultivation of the improved seedlings. Given these considerations, the CBA for this intervention is divided into two parts. Part I estimates the financial and economic incremental benefits that cultivation of HYV seedlings brings to smallholder farmers and other stakeholders in the economy. This part of the analysis can serve as a stand-alone CBA on yield-improving interventions, and the results from its financial module help determine whether coffee farmers find the financial incentive to acquire the JARC’s coffee seeds or seedlings. In addition, results from the economic module help determine the net downstream economic externalities of seedling production.

Part II of the analysis focuses on the nursery level. Central to this analysis are the questions of whether the nurseries are financially sustainable and, if so, the magnitude of economic benefit generated by each nursery.

Increasing the traceability or branding of Ethiopian coffee is the second intervention evaluated in this report. The activities that fall under this intervention vary widely, including regional coffee quality profiling, providing training to cuppers, harmonizing cupping criteria with international standards, sponsoring a coffee-cupping competition, and organizing coffee-cupping trips to coffee-producing areas for female entrepreneurs. It is hoped that increased traceability and branding can increase coffee export prices and thus add value for all value chain participants. In terms of the methodology, the CBA for this intervention has been conducted using the farm model developed for the CBA for the nursery-sponsoring intervention. The success of the CBA hinges critically on whether the measures currently planned by ACDI-VOCA to improve traceability increase the price of coffee cherries at the farm gate.

INTERVENTION 1 — PRODUCTION AND MARKETING OF HYV

SEEDLINGS

It is estimated that each year, the demand for coffee seedlings from Ethiopian coffee farmers reaches five million. The AGP-AMDe project incorporates a series of interrelated measures to facilitate their distribution to ensure that farmers acquire the improved high-yield variety (HYV) seedlings, including the development of improved coffee varieties, the multiplication of improved seeds and seedlings, and the distribution of these verities to coffee farmers.

The Jimma Coffee Research Agency (JARC) is a government agency in charge of coffee research and the development of improved coffee varieties. A component of the AGP-AMDe project is to support the JARC and related parties in their efforts to expand the development of improved coffee varieties and their adoption by coffee farmers.1

Overcoming the bottleneck in the supply and distribution of HYVs of coffee could significantly raise smallholder coffee productivity and Ethiopian coffee production. HYVs are inherently higher-yielding and resistant to the destructive coffee-berry disease (CBD). In areas seasonally afflicted by CBD, adopting HYVs has the additional advantage of preventing substantial yield losses.

At the present, the JARC requires that the recipients of its improved coffee seeds or seedlings adopt recommended, yield-maximizing farm-management practices, such as intensifying the use of chemical fertilizers, more aggressively pruning coffee trees, and spacing them closer, and so forth.

Maximizing coffee yield, however, does not necessarily imply optimizing the economic outcome. The JARC’s recommended model farming system entails high labor and high physical input costs. The costs are incurred early in the production cycle, and months will lapse before revenues from the sale of coffee cherries are realized. Limited access to credit does little to finance the cultivation operation of many cash-strapped farmers. In addition, those who adopt the JARC’s recommendations must forego part of the income generated from intercropping with cash and subsistence crops, as coffee trees are planted at higher densities. As a consequence, they have less time for intercropping before the coffee canopies shade out the intercrops.

The difference between yield maximization and optimization of economic returns is understood by large-scale and smallholder coffee growers. The majority smallholder Ethiopian coffee farmers who have managed to get access to HYVs use little to no chemical fertilizer and in general have adopted no other new cultivation practices. Even some large-scale growers with access to HYVs report using only a small amount of fertilizer, and then only during the initial planting of coffee seedlings.

The supply bottleneck for the propagation of HYVs thus seems to arise just as much from the conditions that restrict farmers from implementing the recommended farming practices as from the limited capacity of the JARC to produce HYV seeds or seedlings.

The financial module in Part I is conducted at the farm-household level to investigate the financial incentive for coffee farmers to acquire HYV seeds from the JARC or seedlings from its licensed seedling nurseries rather than to recycle or to use local seeds. Building upon the financial module, the economic module of the analysis estimates the incremental net economic benefit generated by a household that cultivates HYV coffee trees. Dividing the incremental net economic benefit by the number of HYV trees per hectare (Ha) of coffee holding gives the levelized net economic benefit of a HYV seedling. Part II of the analysis relies on the levelized benefit per seedling to examine the financial and economic merits of sponsoring the establishment of a HYV coffee-seedling nursery and the training of its producers.

1 Recently a memorandum of understanding to this effect was signed. The CRI has not been able to acquire the details but believes the memorandum should relate to Component 4 (innovation and investment) of the work plan.

Part I — Financial Analysis of the Cultivation of High-Yield Variety Seedlings

Model Farm Description and Parameter Values This analysis considers the cultivation of coffee trees in three separate scenarios: (1) the continuation of traditional agronomic practices using local variety seedlings; (2) the continuation of traditional practices using HYV seedlings; and (3) the adaptation of model farming practices using HYV seedlings. A separate farm model is constructed for each of the three scenarios. Based on the data collected from multiple sources, the farm model specifies the labor and physical input requirements and forecasts the revenue generated and costs incurred. The overall objective is to compute the net cash flow over the project life and the corresponding financial net present value (NPV) that a household may receive from coffee cultivation. Note that the first scenario describes the status quo. The differences in financial NPVs between scenarios 1 and 2, and between scenarios 1 and 3, represent the incremental financial benefit of the HYV seedling cultivated with traditional practices and model farming practices, respectively.

Although coffee trees can live as long as 30 years, this analysis considers a time frame of 20 years. For coffee or other commodities whose prices fluctuate widely from year to year, the analysis findings may lose precision as the time frame extends beyond a few years. However, given that 3 years will lapse before the newly planted coffee seedlings bear cherries, and that stumping (cutting the tree stem to knee height) is required 7 years after the trees become productive, a relatively short time frame would fail to capture the full financial return of coffee cultivation.

It is assumed that a household has a plot of coffee holding that measures 1 Ha. In addition, to facilitate a fair comparison between scenarios, it is assumed that the cultivation is on previously uncultivated land that requires land clearing and development, followed by the planting of coffee seedlings. Under the traditional farming system, fewer coffee trees are planted. Numerically, the average tree density in the traditional system is 1,800 seedlings per Ha, as opposed to 2,350 seedlings under the model farming technique.

Seedlings of both varieties take 3 years to grow. The wider spacing between trees in the traditional system allows for 2 years of intercropping, which generates some income to partially finance household consumption during the interim. Under the model farming system, however, intercropping is limited to the first year. In the fourth year, the young trees bear fruit. The harvest quantity varies by the seedling variety and farming technologies. For scenario 1, the annual yield is expected to be 4 quintal (400 kilograms [kg]) of green-bean equivalence (GBE) or 2.4 tons of fresh cherries per Ha; for scenario 2, the expected annual yield is 7 quintal (700 kg) of GBE or 4.2 tons of fresh cherries per Ha; for scenario 3, the expected annual yield is 12 quintal (1.2 tons) of GBE or 7.2 tons of fresh cherries per Ha. The trees remain productive from years 4 to 10. At the end of the 10th year, farmers cut the tree stems to knee height, a process known as stumping, so that new stems will develop from suckers to replace the original stems, whose productivity begins to fall past the 10th year. At the 14th year, the new stems bear fruit and thus begin the second productive phase. At the 20th year, the trees are stumped again, and the project is considered to end at this point. Table 1 summarizes the yield and production cycle of different farming technologies over the 20-year period.

Table 1. Yield estimate and coffee cultivation cycle over 20 years by farming technology

Traditional practices Model farm

Local variety

(1)

High yield variety

(2)

High yield variety

(3) Coffee holding area and yield

Coffee holding area per household Ha 1 1 1 Tree density trees/Ha 1800 1800 2350

Yield per coffee tree Green bean equivalence (GBE) kg of GBE per tree 0.22 0.39 0.51 Red cherry kg of red cherry per tree 1.33 2.33 3.06

Yield per hectare

Green bean equivalence (GBE) quintal/Ha per year 4.00 7.00 12.00

Red cherry quintal/Ha per year 24.00 42.00 72.00

Production Cycle Length of coffee cultivation year 20 20 20 Planting year 1st 1st 1st First harvest year 4th 4th 4th

Stumping year 10th 10th 10th Source: Own derivation based on farm-budget data provided by ACDI-VOCA (2012c).

Besides seedling density, the two farming systems also vary by the intensity of labor and physical input utilization that is required. In both cases, a set of hand farm tools and drying materials are required. The model farming system uses 2 kg of organic compost/Ha for fertilizer, as compared to 1 kg for traditional methods. In addition, the model farming system requires 2,350 HYV seedlings, whereas the traditional method requires only 1,800 seedlings of either variety. Despite these quantitative differences, the monetary values of physical inputs are largely identical for the two farming systems.

However, the same cannot be said about labor inputs. To achieve an annual yield of 12.0 quintal GBE/Ha, the model farming system is substantially more labor intensive. For any phase of cultivation, the amount of labor input required is multiple times that required for the traditional system. During the harvest seasons, additional labor is required to selectively strip ripened red cherries off the branches. There is a general consensus that the average quantity of harvested cherries by an able-bodied adult is 30 kg/day. A breakdown of labor input by activity for each scenario is reported in table 2.

Over the 20-year coffee-cultivation period, the amount of required labor inputs fluctuates depending on the stages of tree development. For most of the time, a household that continues traditional farm practices has a sufficient labor supply to perform its own manual activities. Hired labor is only required intermittently to make up for the temporary labor shortfall during harvests or land clearing and development. However, the model farming system is so labor intensive that households adopting it are likely to recruit external laborers regularly. To estimate the extra costs this method entails, it is helpful to first consider a household’s own labor supply, which is a function of the household size.

Table 2. Labor requirements for different farming technologies (person day/Ha/year) Traditional practices Model farm

Local variety (1) HYV (2) HYV (3) One-off

Land development 24.9 24.9 32.5 Holing, refilling, pegging 32.6 32.6 42.5 Planting seedlings 9.6 9.6 12.5 Subtotal 67.0 67.0 87.5

Unproductive (young or stumped) tree maintenance Weeding and cultivation 0.0 0.0 50.0 Hoeing and ridge maintenance 5.0 5.0 25.0 Mulching 12.5 12.5 25.0 Subtotal 17.5 17.5 100.0

Intercropping Intercropping operations 25.0 25.0 25.0 Duration* 2 years 2 years 1 year

Productive tree maintenance Shade-tree regulation 2.5 2.5 5.0 Weeding and removing lichen 2.5 2.5 16.0 Desuckering, pruning, stumping 7.5 7.5 32.0 Subtotal 12.5 12.5 53.0

Harvest preparation Drying bed construction 25.0 25.0 37.5 Small storage construction** 12.5 12.5 25.0 Subtotal 37.5 37.5 62.5

Harvest 30 kg daily 30 kg daily 30 kg daily

* Wider spacing allows for 1 more year of intercropping.

** This occurs once during the first harvest in the fourth year.

Source: Own derivation based on farm-budget data provided by ACDI-VOCA (2012c).

Statistics on the average household size for Jimma are not available, but the average household for the Oromiya region is estimated to be 5.2 persons. Table 3 reports the number of people in the average Oromiyan household along with their ages and gender composition.

For the same labor duration, the productivity and thus the amount of work completed by individual household members will vary owing to the difference in physical strength. To capture this difference, the labor of female adults, children, and seniors has been converted to male adult labor equivalence. By default, the labor equivalence of male adult labor is 1; those of female adults, children, and seniors are assumed to be 0.7, 0.3, and 0.3, respectively. Based on this information, it is estimated that a household can allocate up to 91 person days of labor to coffee farming each year (table 4, column 3).

Table 3. Average size, age, and gender composition of an Oromiyan household Oromiya

Average household size 5.2

Age composition

15–64 46.8%

0–14 50.0%

65+ 3.2%

Gender composition

Male 50.3%

Female 49.7%

Source: 2007 Household Income, Consumption, and Expenditure Survey, Table 3.2 (CSA 2007).

Table 4. Own labour supply of an Oromiya Household Household composition - Oromiya

Family member Person day equivalence

Day on coffee farm per year

Person day equivalence per year

Adult male (15-64) 1.22 1.0 40.0 49.0 Adult female (15-64) 1.21 0.7 40.0 33.9 Children (0-14) 2.60 0.3 10.0 7.8 Senior (65+) 0.17 0.3 10.0 0.5

Total 5.20 91.1 Source: Own calculation based on 2007 Household Income, Consumption, and Expenditure Survey, Table 3.2 (CSA 2007).

In any year when the total labor requirement exceeds 91 days, the household hires external labor. During field-visit interviews, stakeholders reported that the daily wage rate for unskilled laborers is approximately 30 ETB. In addition, coffee farmers provide meals to laborers, which cost 10 ETB/laborer/day. The total daily cost of hired labor is thus approximately 40 ETB per person. While tending the coffee trees, coffee farmers must leave unattended other agricultural, income-generating activities, which means there is also an opportunity cost to their labor on the coffee farm. The imputed labor cost/day is assumed to be 24 ETB. The farm model accounts for both labor types (own household supply and hired) and assigns costs accordingly.

During a field visit to Jimma, the CRI team acquired from a sebsabi (collector) and a chaanyii (supplier) data about their purchases of coffee from farmers to study the output price of coffee cherries. The sebsabi’s data covers the past 2 years, while those of the chaanyii run from November 2011 to May 2012.

The price of cherries ranges from 14 ETB/kg in late 2011 to 9 ETB/kg in recent months. In the model, the assumed cherry price is 14 ETB/kg.

Financial Analysis Findings Using the aforementioned information, the farm model makes forecasts about labor and physical inputs required for the three scenarios, the associated costs incurred, and the revenue generated over the 20-year cultivation period. The financial NPVs summarize the overall profitability of the ventures. The results for each of these ventures include the revenues and expenditures for their coffee production as well as their intercropping activities.

At a 12 percent discount rate, traditional farming that uses local-variety seedlings yields a financial NPV of 102,854 ETB (US$5,857). If HYV seedlings are planted instead, traditional farming yields a financial NPV of 209,600 ETB (US$11,936). Lastly, the model farming system yields a financial NPV of 352,248 ETB (US$20,060). Even when farmers continue traditional farming practices, the incremental financial benefit of cultivating HYV seedlings is 106,746 ETB (US$6,079). The incremental financial benefit of switching from traditional practices using local-variety seedlings to the model farming system is even higher, at 249,394 ETB (US$14,202).

Table 5 summarizes the financial NPV of each coffee-producing system, inclusive of the benefits and costs of intercropping. Sensitivity analyses are conducted on several risk variables, including percentage deviation in yield estimate and annual percentage change in real daily wages of hired labor and coffee-cherry price. The presentation of these sensitivity analysis results occurs later in the economic module of this analysis so the incremental benefits of other stakeholders can be juxtaposed alongside farmers’ financial benefit for comparison.

Table 5. Financial NPV of coffee cultivation by farming technology (12 percent discount rate, 20-year project life)

Traditional practices Model farm

Local variety (1)

HYV (2) HYV (3)

ETB

Financial NPV per household 102,854 209,600 352,248 Incremental financial NPV

Scenario (1) as the base case 106,746 249,394 Adopting model farming practices (3) – (2) 142,648

US$

Financial NPV per household 5,857 11,936 20,060 Incremental financial NPV

Scenario (1) as the base case 6,079 14,202

Adopting model…