Chickpeas_Cost_benefit_analysis.pdf

i

This report was produced for review by the United States Agency for International Development (USAID).

It was prepared by Optimal Solutions Group, LLC and Cambridge Resources International for USAID’s “Learning, Evaluation, and Analysis Project (LEAP). Contract Number: AID-OAA-C-11-00169

COST-BENEFIT ANALYSIS OF A

CHICKPEAS VALUE CHAINS IN ETHIOPIA

AGRICULTURAL MARKET DEVELOPMENT -

AMDe PROJECT

FINAL REPORT

i

UNITED STATES AGENCY FOR INTERNATIONAL DEVELOPMENT

LEARNING, EVALUATION, AND ANALYSIS PROJECT

(AID-OAA-C-11-00169)

COST BENEFIT ANALYSES OF AMDe INTERVENTIONS

IN SELECTED VALUE CHAINS: CHICKPEAS

FINAL REPORT

Prepared for:

U.S. Agency for International Development/Ethiopia

Prepared by:

Optimal Solutions Group, LLC Cambridge Resources International Inc.

University of Maryland Research Park, M Square

60 Montgomery St Cambridge, MA, 02141

5825 University Research Court, Suite 2800 USA College Park, MD 20740 USA www.cri-world.com www.OptimalSolutionsGroup.com

With assistance from USAID/ E3 Analysts

Project Analysts:

Jerrod Mason, USAID/ E3 Juan Buttari, USAID/ E3 Glenn P. Jenkins, Cost-Benefit Manager

January 13, 2013

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.

ii

Table of Contents Project Description

Chickpeas: Characteristics and Potential

Yields and Challenges

Value Chain Actors and Dynamics

Interventions and Spreadsheet Logic

Model and Quantitative Results

Financial Flows With and Without Interventions

Sensitivity Analysis Findings

Implications of the Benefit Cost Results

Bibliography

Project Description

AMDe’s interest in chickpeas is based on the pea’s importance as a source of food security and its export potential. To help build long-term economic stability for smallholder chickpea farmers, AMDe seeks to increase the competitiveness of the Ethiopian chickpeas industry through such activities as: (1) addressing impediments to the availability of improved seeds, (2) increasing productivity through rhizobium inoculation and use of mineral fertilizers needed to enhance biological nitrogen fixation and subsequent crop growth, and (3) supporting the upgrading and expansion of post-harvest handling and storage, including capacity-building for better warehouse management practices. Likewise, AMDe will endeavor to strengthen the management capacities of cooperatives and unions.

To achieve its objectives, AMDe is committed to exploring innovative investment opportunities to increase overall performance of the chickpea industry. Accordingly, the project will also seek to improve the enabling environment for chickpeas production, marketing, and distribution. The total budget evaluated for improvements in Ethiopia’s chickpeas value chain is 15 million birr (equivalent to $857,928) over five years.

Specific interventions being contemplated under the project include:

Introduction and multiplication of improved seeds, especially of the Kabuli type (see below);

Rhizobium inoculation and mineral fertilizer application;

Expansion of cleaning, handling, and storage facilities;

Variety and fertility trials;

Facilitation of access to credit for the purchase of seed and fertilizer; and

Enhancing the capacity of cooperatives to provide training in harvest and post-harvest handling techniques;

Introduction to mechanized systems for tillage, planting and harvesting;

Linkage to ECX marketing, warehousing and receipting systems.

This chapter reports on the cost-benefit estimates attaching to the first two interventions.

Chickpeas: Characteristics and Potential

While presently cultivated as a secondary crop, chickpeas have export potential and can also contribute to increasing food security among Ethiopian farmers. Chickpeas are a rich source of nutrion – zinc, folate, proteins, phosphorus, high in dietary fibers, low in fat which, is mostly polyunsaturated-- and there is evidence that they can assist in lowering cholesterol in the bloodstream.

Chickpea cultivation minimizes or eliminates the need for supplemental nitrogen, which is also beneficial for rotational crops that follow. The plant requires relatively low levels of inputs and management, is very drought-resistant, and is often grown immediately after the harvest of cereal crops. The chickpea plan’ts deep-roots are able to exploit the substantial moisture reserves that remain in the vertisols – black clay soils –found throughout the project area.

Chickpea prices have trended upward in real terms in recent years. World prices for chickpeas, which are a key determinant of domestic prices, increased at roughly 8% annually in real terms between 1994 and 2005. This was followed by larger food-price increases after 2007.

Moreover, chickpea profitability has increased over the past two decades when compared to other agricultural commodities. Accordingly, Ethiopian farmers have responded to these rising prices by increasing annual production by 7.6% from 1995 to 2008. Chickpeas are currently the second-largest pulse crop in the country after fava beans. About 1.5 million farmers produce this crop. In 2011, production of 330,000 tons occurred on 220,000 hectares, making Ethiopia the largest chickpea producer in Africa and one of the largest in the world.

There are two main types of chickpea: Desi, which has a smaller and darker seed, and a rough coat; and Kabuli, which has a larger and lighter-colored seed with a smoother coat. The Desi chickpea, Ethiopia’s traditional variety, is widely grown in the country and is mainly used to satisfy household food needs and broader domestic markets. The production of Kabuli types is a much more recent development, as new varieties have been released.

Table 1: Kabuli and Desi: A Comparison for Debre Zeit, Ethiopia Item Chickpea Enterprise

Desi Grain Production Kabuli Grain Production

Kabuli Seed Production

Yield (quintal per hectare)

18 18 18

Price of grain/seed (per quintal)

166 240 281

Gross Margins (birr per hectare)

2988 4320 5058

Incremental Benefits (birr per hectare)

1332 2070

Source: Richard Jones, Patrick Audi, Bekele Shiferaw and Eastonce Gwata, “Production and Marketing of Kabuli Chickpea Seeds in Ethiopia:

Experiences from Ada District,” Adapted from International Crops Research Institute for the Semi-Arid Tropics May 2006.

Despite the rapid growth in chickpeas prices, production worldwide is roughly trendless over a decade. In contrast, from 2004 to 2011 Ethiopian chickpea export volumes grew at 24% per year, while revenues grew 38% per year.

By 2011 chickpeas were the country’s third largest export crop after coffee and sesame. Many observers argue that prospects for expansion of production and exports are good due to the steady and substantial growth in international demand. In 2011, exports of chickpeas represented about 18% of the year’s production, with the Desi variety accounting for 82% of Ethiopia’s chickpeas exports and Kabuli for the remainder. Nonetheless, while Ethiopia is an important chickpea producer, its chickpeas are produced mainly for domestic household consumption, with Ethiopian women playing the key role in determining the quantity sold in the market for cash.

While there is an export market for both types of chickpea, the market for the larger, lighter Kabuli variety is expanding more rapidly. Moreover, Kabuli chickpeas fetch significantly higher prices in the export markets. In 2006 Kabuli prices were 45% higher -- in 2011 they were 15% higher.

Furthermore, it has been argued that Kabuli may offer relative financial advantages even in the domestic market. Table 1 compares historical prices and marginal returns to Desi and Kabuli chickpeas (Jones and others, May 2006). The results shed light on the potential of the Kabuli variety for producers.

The analysis on which the table is based pointed to very high marginal rates of return to Kabuli as compared to Desi. As the authors indicated, farmers were better off producing Kabuli grain.

According to IFPRI (2010), Ethiopia could expand its foreign market presence through increased production levels. If achieved, increased production could double the country’s annual exports. In fact, IFPRI is argued that even under conservative assumptions, doubling production will result in at least doubling export earnings from US$90 million to US$180 million.

Table 2: Premium Grain Chickpea Ethiopian Varieties for High End Markets Not Widely Promoted or Adopted by Ethiopian Farmers

Key Quality Specifications

Variety Grain Size (mm)

100 seed weight in grams

Color Released by Adoption Status

Premium International

7-9 30-45 White or Cream

Arerti 6 25.7 Cream white Debrezeit Agricultural

Research Center

Limited to Shewa and some parts of Amhara region

Shasho 6-7 29.9 Cream white

Ejere 8-9 37.4 Cream white Farmers don’t know the varieties or have just recently heard of them

Chefe 6 27.7-39 Cream white

Teji 8-9 38.1 Cream white

Akuri 6-7 37-39 White Sirinka Agricultural Research Center

Recently released (2011) and not yet promoted and adopted by farmers

Kasech 6-7 32-39 White

Yelibe 6-7 32-39 Yellowish Source: ATA, Final Chickpea Roadmap, May 2012

In addition, there is consensus that the international market for chickpeas will continue to expand given the growing demand in the region and elsewhere for sources of protein and nutrition. Further, chickpeas can be used for a broad variety of dishes. Pakistan, the United Arab Emirates, and Sudan are major importers or chickpeas. India and Australia are potential export markets.

In sum, the current export market for Ethiopia is underdeveloped. However, to expand Ethiopia’s share in the international market for chickpeas, the country will have to overcome several challenges. For instance, exporters reject a significant proportion of produce due to low quality standards and farmers claim that they are not familiar with export quality requirements. Thus, if such constraints are addressed, there is potential to increase chickpeas yields and to expand exports.

Yields and Challenges

The International Food Policy Research Institute (IFPRI) has stated that while current average chickpea yields are 1.2 metric tons per hectare, Ethiopia could achieve yields of 2.9 tons per hectare if cultivation is accompanied by the appropriate inputs. In fact, in Lume, Oromia, yield is estimated at 3.5 MT per hectare and this is taken by the Ethiopian Agricultural Transformation Agency (ATA) as evidence of untapped potential in Ethiopia (ATA, May 2012). Productivity gains from improvements in planting techniques could double overall pulse production to two million tons over a period of five years. This gain in productivity, as IFPRI points out, would increase smallholder income by 40% to 70% per hectare and ensure greater food security through meeting domestic pulse demand.

The low level of productivity of chickpeas is attributed to the following main factors:

1. Limited availability of improved seed and limited familiarity with the variety of existing pulse types – the use of certified improved seeds by farmers is quite low;

2. Low input usage, especially mineral fertilizers capable of increasing yields in field trials by 10 to 80% – however, the importance of mineral fertilizers for chickpeas is a topic on which there are differing views (more below); and

3. Limited usage of modern agronomic practices ( e.g., timing of plowing, rhizobium inoculation, insecticide applications, crop rotation, and weeding and harvesting ).

Two points are worth highlighting. First, during field visits, the analysis team met farmers who argued that, at least for Desi, chemical fertilizer was not that important, given that the plant could thrive in relatively tough conditions. Even if this is accepted, fertilizer could still be important for the cultivation of Kabuli chickpea. Second, consistent with IFPRI’s analysis, the farmers agreed that substantial yield increases were possible by just introducing modern basic practices, which require very little investment and which are easy to demonstrate.

Reinforcing the point on the returns to the introduction of modern practices, the ATA indicates that current extension services provided to farmers are inadequate in that they:

a) Are designed for pulses in general and fail to address specific essential components for chickpea farming

b) Neglect important agronomic factors such as optimum tillage required, critical weeding time, initial fertilizer application, rhizobium inoculation, optimum plant population, seed quality requirements, and optimum planting date

c) Lack required knowledge for scouting pests and skills in pesticide application

Moreover, continues the ATA, chickpea is primarily harvested through uprooting which decreases soil fertility benefits provided by the plants, and threshing is conducted using the traditional on-floor beating of the straw ‘audima’. The latter practice has negative impacts on the final product leading to high levels of impurity, with the presence of foreign matter, and broken and shriveled seeds.

In this light, AMDe can help by facilitating inputs, including training, that help the country achieve its potential in chickpeas yields. Two of those inputs are improved seeds – which should be provided along with training on their use – and better post-harvest facilities and management to reduce losses and the deterioration in quality noted by exporters.

As relates to improved seeds, the ATA has noted that available quantities of improved seed do not match demand because of insufficient production of improved chickpea seed. They estimated that only 40% of improved seeds requirements were met in 2011 and that the prospects for 2012 were even lower at 32% (ATA, 2012). Moreover, as Table 2 shows, many Ethiopian varieties yielding premium grain quality for high end markets are not widely promoted and adopted by farmers.

A practice for the development and distribution of improved seeds is that the research centers under the Ethiopian Institute of Agricultural Research provide breeder seeds to enterprises, for example the Ethiopian Seed Enterprise, which multiply the seeds in their own farms to produce what is termed pre-basic or basic/foundation seed. The foundation seed is afterwards sold to agricultural extension departments for distribution in the different regions. However, this system seems to be insufficient to make enough improved Kabuli and Desi types available to smallholder producers.

In connection with post-harvest handling of crops, it has already been noted how the lack of access to adequate storage causes a high proportion of loss of pulse production, as various pests such as insects, rodents and birds can decimate pulses that are inadequately protected in storage.

Value Chain Actors and Dynamics

As seen on graph 1, the value chain for chickpeas is complex and has a number of channels. Fifty seven percent of small farmers’ output is collected by rural assemblers who, in turn, sell 54% of the stock to wholesalers. Likewise, small farmers directly sell 19% of the produce to the wholesalers. This means that, not counting sales from union cooperatives to wholesalers, the latter account for 73% of small farmers’ sales. Sales by farmers to union cooperatives are 16% of their crop.

As shown in the graph, the network for the production and distribution of chickpeas involves farmers, assemblers, retailers, wholesalers, exporters, processors and cooperatives. The assemblers collect the grain from smallholders and deliver the grain to wholesalers. While they generally are independent agents, who risk their own financial resources, they can also act as agents for wholesalers on a commission or fixed-fee basis.

Figure 1: Marketing Flows for Chickpeas in Ethiopia

Source: Bekele Shiferaw* and Hailemariam Teklewold, “Structure and functioning of chickpea markets in Ethiopia: Evidence based on analyses of value chains linking smallholders and markets”, 2007.

Wholesalers operate at woreda and urban levels. They make high use of cellular phones to carry out buying and selling activities through their contacts in different markets, as well as accessing ECX price data.

Generally, warehouse facilities sell products at more wholesale levels than rural markets. Nonetheless, the practice of storing chickpeas to benefit from inter-seasonal price movements is rare and is attributed to liquidity constraints and high storage risks.

Furthermore, there are relatively few large scale and medium level mills that process chickpeas. Most process Desi; although more recently built facilities are also processing Kabuli chickpeas.

Reportedly, there are no exporters specializing only on chickpea; rather, they export a variety of goods and respond to changes in domestic and international prices when choosing which goods to export. In recent years, there has been growing interest on the part of exporters to increase the volume of Kabuli exports and until recently the exporters complained about a shortage of that type of chickpea.

Additionally, income paid to farmers, as well as farmers’ share of final prices, vary depending on the distribution channel. In general, the simpler the distribution channel (that is, the fewer the number of intermediate steps between production and sale to final consumer), the lower the price to the farmer, but the higher share of the final price becomes. For example, Shiferaw and Teklewold (2007) report that when sales take place in the primary market – the local rural market that involves a farmer and rural retailer channel – the farmer’s share is 97% of the final price. In this simple channel, the farmer sold chickpeas at 261 birr per 100 kilograms, and the rural retailer sold at 269.5 birr per 100 kilograms. In contrast, in a woreda wholesaler-supermarket channel, the farmer sold at 310.5 birr per 100 kilograms and the supermarket sold at 679 birr with the farmer’s share of the final price being 45.5%.Farmers can receiving a higher dividend if a simpler channel of sales is instilled. The marketing costs rise as the number of intermediate transactions increase between sales at farm gate and sales to final consumer. Nonetheless, from the farmer’s financial perspective, farmers be better off if they are able to increase the proportion of chickpeas marketed through more complex channels.

Table 3 shows seasonal variation in prices for Desi and Kabuli chickpeas. The differences are notable, especially in primary markets. Peak prices occur during the low supply months (off-production) of July and August, and the lowest prices take place during the high supply (production) months of January to May. These differences might be indicative of price smoothing possibilities associated to more accessible and better managed warehouse facilities.

Table 3: Seasonal Variation in Chickpea Prices (Birr per 100 Kg) by Type, Market and Peak Off- Peak Price Period

Crop/Market Peak Price Period Off-Peak Price Period Absolute Difference Kabuli (all ) 443 364 79 Primary 414 294 120 Secondary 403 398 5 Tertiary 553 535 18 Desi (all) 271 216 55 Primary 265 192 73 Secondary 233 225 8 Tertiary 313 297 16 Source: Bekele Shiferaw and Hailemariam Teklewold, “Structure and Functioning of Chickpea Markets in Ethiopia: Evidence Based on Analysis of Value Chains Linking Smallholders and Markets,” (2007).

Interventions and Spreadsheet Logic

Two interventions are analyzed in this section: (1) Improving the accessibility to improved seeds of farmers in the Amhara and Oromia regions of Ethiopia; (2) the effects of better post-harvest handling and warehousing services in regions of Productive Ethiopia. The potential of such interventions was modeled and assessed in spreadsheets that may also be used to simulate the impact on others regions as/if the range of the interventions widens.

In the spreadsheets, the cost-benefit effects of improved seeds are analyzed region by region by first presenting a baseline without-project scenario, then by introducing improved seeds and tracing out effects on income and cost flows without any improvement in post-harvest methods/facilities. Aggregating these results provides a glimpse of how higher productivity could be attained in Ethiopia by simplly intervening with improved seeds in Amhara and Oromia. Then, as a separate step, the analysis assumes a gradual impact of access to more adequate post-harvest methods or facilities.

Finally, the analysis simulates the flows for rural retailers, assemblers, wholesalers, processors, supermarkets, cooperative unions, and exporters for scenarios under both interventions. Spreadsheet values are in constant prices – so changes are real changes and do not incorporate inflation.1 Some of the assumptions are:

a) The amount of land used for chickpeas grows in the initial years to reflect the application of land-extensive cultivation. Such amount of land reaches a plateau at year 5 after which no further additional land is assumed to be dedicated to chickpeas.

b) With the seeds intervention, the proportion of land using improved seeds rises from zero in year one (2012) to 5% in 2016 after which it remains constant.

c) With the improved seeds, yields increase by approximately 50% starting in 2013.

d) Post-harvest losses without improved handling services or facilities amount to 18% of the chickpeas produced. With better handling losses decrease from 18% to 13% over 5 years.

e) Improved post-harvest handling is supposed to cover 5% of the regions at the end of 5 years.

f) Revenue under the improved seeds intervention results from the combination of revenues from the sale of Kabuli and Desi sesame as the proportion of cultivated land used by Kabuli seed increases.

1 Doing otherwise would involve guessing inflation rates for a long time horizons. While such alternative approach would be useful for analyses of nominal cash flows, it is not useful for the present analysis that fundamentally attempts to assess whether the project promises to be financially profitable and whether the economic impact of the project justifies the economic costs. Nonetheless, the formulas used in the spreadsheet contain a specific term for inflation. Accordingly users of the spreadsheets can simply insert in the inflation term their best guess as to what inflation may be for given years or the full time horizon of 20 years used in the analysis. For the present analysis the value inserted in the inflation term was zero.

Model and Quantitative Results

The model traces out the impact on actors along four different channels:

1. Rural retailer

2. Assembler supermarket channel

3. Assembler exporter

4. Union exporter channel

As indicated by the model’s bottom line numbers—the net present value and internal rate of return, both interventions have a positive net economic impact. Moreover, the sensitivity analysis implies that the results are relatively robust to changes in the assumptions. For example, if yield increases are only half of the base case assumption of 50%, the project would still be valid. Likewise, if the regional coverage of the improved seeds is much less than that assumed or if the price paid for Kabuli sesame falls by close to 10%, the project still falls in the acceptance range. The same holds for less optimistic outcomes in regard to the post-harvest crop handling intervention.

Evaluating who gains and who loses due to the project, the analysis shows net gains for all major value chain actors under both interventions. Although this analysis has not been able to undertake an evaluation of gender-specific impacts, the high involvement of women at the producer level suggests that the interventions are likely to have strong favorable impacts on women.

Financial Flows With and Without Interventions

Table 4 summarizes net financial flows for a more “productive” Ethiopia under the improved seeds and the no-project baselines scenarios. With the improved seeds intervention net income is not only positive but increases faster than in the no-project scenario – an annual growth rate of about 3% compared to less than 1%. Accordingly, incremental net income (i.e., the difference in net income under both scenarios) is positive and grows at 7% yearly.

Table 4: Financial Flows, Yearly Net Income in Productive Ethiopia. Improved Seeds and No- Project Scenarios Compared (thousand birr) 2012 2013 2014 2015 2016 to 2031 Improved Seeds 75,745 80,290 85,107 85,107 85,107 No Project 75,745 86,896 96,304 100,499 108,888 Incremental Net Income - 6,607 11,197 15,392 23,781

In the above, chickpeas farmers in “productive” Ethiopia stand to gain from the facilitation of improved seeds. While cultivation costs rise when improved seeds are used, revenues rise faster, hence resulting in the increased net incomes.

Table 5: Financial Flows, Yearly Net Income in Productive Ethiopia. Post-Harvest Handling Intervention and No-Project Scenarios Compared (thousand birr) 2012 2013 2014 2015 2016 to 2031 Post-Harvest Handling Intervention

75,745 80,290 85,107 85,107 85,107

No Project 75,745 80,701 86,850 89,029 96,002 Incremental Net Income - 411 1,743 3,922 10,896

As Table 5 shows, interventions designed to reduce post-harvest losses through better handling also have positive potential. However, their net benefits are lower than those targeted at increasing availability of improved seeds. With the PHL intervention, financial net income would grow at 2% annually. This results in a 20% marginal increase in annual net incomes by 2016.

Table 6 shows the financial flows of other value chain actors (i.e., aside from farmers) in key distribution channels under the improved seeds intervention compared to no-project situations. All these actors can potentially benefit from the intervention relative to no-project scenarios.

Table 6: Financial Flows, Yearly Net Income of Other Value Chain Parties. With-Improved Seeds Intervention and No-Project Scenarios Compared (thousand birr) Incremental Net Income 2012 2013 2014 2015 2016 to 2031

Assemblers - 131.4 278.6 417.9 696.4 Woreda Wholesalers - 84.1 178.3 267.4 445.7 Processors - 135.0 286.3 429.4 715.7 Supermarkets - 662.6 1,404.8 2,107.2 3,512.0 Grain Exporters - 0.1 0.3 0.4 0.6 Co-ops - 12.0 25.4 38.1 63.4 Rural Retailers - 8.3 17.7 26.5 44.1

In terms of absolute levels of net benefits, supermarkets stand to gain the most from the improved seed intervention; other big gainers are processors and assemblers. In contrast, grain exporters gain the least among these actors, with rural retailers and co-ops following.

Table 7 shows the net financial benefit to other value chain actors from the PHL-focused intervention.

While levels are much lower than those experienced with the improved seeds intervention, again supermarkets, processors, and assemblers experience the bigger gains in net income with rural retailers and coops the lowest.2

2 At the time of this work the precise nature of the post-harvest handling intervention was not defined as far as the analysts could determine. Naturally, the precise definition of the intervention is likely to alter the basic arithmetic of relative winners.

Table 7: Financial Flows, Yearly Net Income of Other Value Chain Parties. Post-Harvest Handling Intervention and No-Project Scenarios Compared (thousand birr) Incremental Net Income 2012 2013 2014 2015 2016 to 2031

Assemblers - 9.2 39.1 88.0 244.4 Woreda Wholesalers - 5.9 25.0 56.3 156.4 Processors - 9.5 40.2 90.4 251.1 Supermarket - 46.5 197.2 443.6 1,232.2 Grain Exporter - 2.1 8.9 20.0 55.5 Coops - 0.8 3.6 8.0 22.3 Rural Retailer - 0.6 2.5 5.6 15.5

Tables 8 and 9 present the final project NPVs and IRRs for both the financial and economic analyses.

The financial analysis indicates that the improved seeds intervention is potentially very beneficial – the present value of total net benefits is around $6.6 million and the internal rate of return is 39%. The net present value of net benefits per beneficiary amounts to $150.

Table 8: Financial Analysis, Benefit-Cost Criteria by Intervention

Improved Seeds Improved Post Harvest Handling Total Net Present Value Birr 116 million ($6.6 million) Birr 36.9 million ($2.2 million)

Total Net Present Value per Beneficiary

Birr 2,649 ($151)

Birr 842 ($50)

Internal Rate of Return 39% 33%

The improved post-harvest handling intervention can be very profitable, with about $2.2 million in total net present value, an internal rate of return of 33%, and a net present value per beneficiary of approximately $50. While both the improved seeds and the post-harvest handling interventions make financial sense, the improved seeds interventions is the more attractive of the two.

Table 9: Economic Analysis, Benefit-Cost Criteria by Intervention

Improved Seeds Improved Post Harvest Handling Total Net Present Value Birr 104 million ($6 million) Birr 36.9 million ($2.2 million)

Total Net Present Value per Beneficiary

Birr 2,365 ($137)

Birr 835 ($51)

Internal Rate of Return 39% 33%

Given the assumptions, the economic analysis showed qualitatively similar results and reinforces the above conclusions.3 The improved seeds intervention shows a total economic net present value of $6 million, a present value per beneficiary of around $135, and an economic internal rate of return of 39%.

3 At the time of modeling the analysts had not been able to obtain a clear picture of distortions that could potentially affect the economic value of the inputs used for the cultivation of chickpeas. While the analysts judge that it is unlikely that a fuller accounting of economic distortions would change the go/no-go assessment of the project, an updating of the assumptions should be undertaken as information and its sources become available or more open.

For improved post-harvest handling, the economic NPV is about $2.1 million, the present value per beneficiary is approximately $50, and the economic IRR is 33%.

These results mean that both interventions have positive net economic benefits for the Ethiopian economy. The present value of the net incremental financial and economic flows of benefits is positive and the internal rates of return are in the acceptable range for both interventions. These outcomes are, however, contingent upon how closely the reality of project implementation matches the assumptions outlined in the model, a point which will be expanded upon in the section on sensitivity analysis.

Sensitivity Analysis Findings

Since cost-benefit analysis is by nature an ex-ante evaluation, it is important to subject the model’s main variables, which cannot be known with certainty in advance, to a sensitivity analysis, to determine which variables can change the model’s outcomes substantially if they deviate from initial expectation. Table 10 shows the results of the sensitivity analysis conducted, where we varied several model variables to determine the impacts on the project’s NPV and IRR. Such sensitivity tests help assess the robustness of the results and shed light on some of the variables that require close monitoring during project implementation.

For example, for the improved seeds intervention, if yield gains are not the 50% assumed in the basic modeling but only 40%, the present value of the incremental net benefits will be slightly over 20% lower while the internal rate of return will be, close but not quite, 20% lower. In other words, within such range of yields, the sensitivity index of the indicators of project worthiness (e.g. percentage change in project worthiness indicators divided by percentage change in yield) approaches one from below.

Analogous proportional results hold for other variations (gains or losses) in yields relative to the basic model. Even if yield gains are as small as 10% rather than the assumed 50%, the project would still make economic sense as indicated by the present value of the incremental net benefits. Nonetheless, a net present value between zero and $1.0 million associated with yield gains of only 10% suggests, that yield gains of less than that would threaten the viability of the project.

The regional coverage trials suggest that changes in coverage are likely to be associated with proportional changes in the measures of project worthiness. Noticeably, the project would still make economic sense if coverage reaches only 1% of the region rather than the 5% assumed in the basic model. Accordingly, it would seem that, in general, achieving yield gains is more important than the extent of regional coverage.

In contrast, the financial worthiness of the project is highly sensitive to the price of Kabuli chickpea.

Changes in such price elicit much more than proportional changes in the project worth indicators.

Accordingly, all the value chain actors – including farmers – that stand to gain from the project will be highly vulnerable to sustained shortfalls in the international prices of Kabuli relative to what was envisaged in the basic analysis.

Table 10: Sensitivity Analysis Improved Seeds Intervention

1. Improved Seeds Intervention Yield Change--percentage change relative to basic model yield increase (50 %)

IRR-- percentage change relative to basic model IRR

NPV-- percentage change relative to basic model NPV

100% 89% 111% 60% 55% 66%

20%* 19% 22% 0% 0% 0%

-20% -18% -22% -50% -47% -55% -80% -80% -88%

Regional Coverage—percentage change relative to basic model coverage

IRR-- percentage change relative to basic model IRR

NPV-- percentage change relative to basic model NPV

100% 112% 101% 60% 75% 62% 0% 0% 0%

-40% -21% -35% -80% -63% -74%

Kabuli Price—percentage change relative to basic model price

IRR-- percentage change relative to basic model IRR

NPV-- percentage change relative to basic model NPV

8% 19% 22% 3% 9% 10% 0% 0% 0% -5% -12% -15% -9% -22% -27%

2. Reduction in Post-Harvest Losses Intervention Loss Over 5 Years—percentage change relative to basic model loss change

IRR-- percentage change relative to basic model IRR

NPV-- percentage change relative to basic model NPV

31%* -70% -83% 15% -26% -41% 0% 0% 0%

-15% 38% 43% -31% 64% 85%

Regional Coverage—percentage change relative to basic model coverage

IRR-- percentage change relative to basic model IRR

NPV-- percentage change relative to basic model NPV

100% 77% 105% 60% 51% 64% 0% 0 0%

-40% -26% -41% -80% -70% -83%

* Interpretation examples: If the improved seeds intervention leads to yield increases that are 20% larger than assumed in the basic model, the IRR and the NPV would be 19 and 22 % larger than in the basic model; if the post-harvest losses intervention shows losses 31% larger than envisaged in the basic model, the IRR and NPV would be 70 and 83 % lower than in the basic model.

As relates to interventions to improve post-harvest losses, the indicators of project worthiness are highly sensitive to the degree of reduction in such losses. If for example, rather than a 13% reduction in losses over 5 years, the reduction is only of 11%, the internal rate of return would be 37% and the net present value of benefits 43% lower.

The indicators are much less sensitive to deviations from assumed regional coverage. The percentage change in project worthiness as suggested by the internal rate of return and the net present value is at most about the same magnitude as the percentage deviation in regional coverage. This suggests that, in terms of the post-harvest losses intervention management practices that translate into smaller losses are more important than a focus on coverage.

Implications of the Benefit Cost Results

The results discussed previously demonstrate that the project has potential to generate significant economic benefits for Ethiopian farmers and for the rest of the country. However, there are several potential roadblocks at the macro- level which are important to take into account. At the broader policy level, it is important to address such constraints as the limitations on using land as collateral for credit.

Also, the generally poor environment for conducting business in Ethiopia (as noted, for instance in the World Bank’s Doing Business reports), and the requirement for the banking sector of having to buy government bonds worth 27% of all loans extended, present challenges to the chickpeas market. This last factor is particularly worrisome, given that it constrains a bank’s loan capacity while many chickpea value chain actors point to a lack of credit as a key obstacle to the industry. Also of special importance is the protection of property rights and contract enforcement, as well as a competitive market as opposed to the public sector or publicly supported subsectors or institutions.

These constraints could have the effect of discouraging private sector investment in this value chain and constrain Ethiopia’s ability to expand its world markets sales. While not every project should necessarily focus on institutional and policy reforms, USAID should contemplate strengthening its capacity to carry out significant policy dialogue activities with its Ethiopian counterparts and link evolving project intervention to progress in the larger policy/institutional reform areas.

At the micro level, AMDe’s interventions contribute to Ethiopia’s Agricultural Growth Program but AMDe is also very dependent on the success of other AGP interventions. As noted in the Ethiopian Agricultural Transformation Agency’s Roadmap for Chickpeas (2012), the realization of this project’s potential hinges on:

Farmers gaining access to improved chickpeas varieties and adopting them

The sustainability of the improved seed production and distribution system, and its responsiveness to market demand

The availability of pesticides

Adoption by farmers of proven agronomic best practices with the assistance of extension services

The adequacy of storage facilities and their management

The efficiency of trade and marketing channels

Reductions in market price distortions as a means to promote efficiency and financial self-sustainability

Meeting quality standards for increased exports The ability of cooperatives and unions to improve their trade networks in importing countries and to be seen as reliable providers (more below)

Such assumptions are more nuanced and more difficult to gauge. Controls and project management will also need to be carefully monitored.

Given AMDe’s orientation in support of cooperatives and their unions, one fundamental assumption of the project’s potential impact hinges on the ability of cooperatives to perform as efficient, commercial endeavors, able to compete in open markets without having to rely on preferential treatment. As the Ethiopian economy develops, current niches of protection for specific sectors and entities are likely to weaken. Efficiency in implementation will be a precondition for success.

In this respect, the present condition of co-ops is not encouraging, and in light of the empirical record the prospects are mixed at best. Despite past financial support, co-op management and administrative systems are rudimentary, and the institutions are deemed not creditworthy. Additionally, most cooperatives lack the essential business orientation at the management level to compete in private markets. This concern is further reinforced by the fact that AMDe and its parent endeavor AGP are very complex, multifaceted projects. Such projects tend to overload the management of their partners and recipients, as well as that of the public and private sector institutions involved in their implementation. AMDe project implementers must be aware of the potential for theses risks to lessen or negate the positive impacts that the project would otherwise achieve. Fostering a competitive business orientation among recipient management teams should be a major goal for AMDe implementers.

In light this, co-ops and unions should identify their comparative advantages in the value chain. For instance, instead of attempts to engage in seed production activities, co-ops and unios should consider increasing their ability to purchase high quality seeds from specialized seed producers and concentrate on distributing the seed among their members for grain production and marketing. Alternatively,, unions should consider increasing their capacity to help co-ops produce high quality grain and link members to the international Kabuli grain market. Accordingly, unions could link efforts with extension agents to train co-op’ farmers in grain production, which includes the use of high quality seed demonstration and integrated pest management practices

To raise the chances of project success, AMDe management should be selective of the co-ops/unions assisted and only support for a limited time institutions able to develop strong business strategies. Central to the support of efforts to raise farmers’ incomes will be the co-ops’ capacity to facilitate demand-driven, extension-type, and short-term training services to their members.

In relation to the international market, the seed system for Kabuli chickpea production is still young, and not all Kabuli varieties are competitive in the international market. Therefore the system involving seed-producing farmers should be strengthened and export marketing efforts should concentrate on Kabuli varieties with good prospects in the international market.As the interest in Kabuli chickpeas relates primarily to its potential for exports4, free market mechanisms to link input suppliers and producers to the international market must not be dictated. Only such mechanisms offer good prospects for long term financial viability.

To achieve the export potential of chickpeas, it is necessary to promote and enhance trade relations with importing and potentially importing countries. While there are project efforts in this direction, the impression developed during the course of this analysis was that there still is a way to go in building up the necessary institutional foundations of entities such as the Ethiopian export associations.

The latter area of work requires further strengthening intelligence on existing or potential markets, solidifying links with domestic providers and other countries’ importers, and influencing the quality of chickpeas produced. In terms of work with domestic providers, it will be essential that they become aware of characteristics that end users (exporters) demand. It is yet to be seen whether co-ops can play a significant role in satisfying the necessary preconditions for conducting business.

To help create the right incentives for export association strengthening, USAID should provide the seed capital and technical support on the quality of links between the associations and the grain providers.

4 While this is currently the case (most Kabuli chickpeas are grown for export), it is certainly possible that Kabuli type chickpeas will eventually supplant Desi in domestic markets, given its desirability in general. Either way, the point remains that effective competition on international markets will hinge in the long run on the efficiency of input and value chain linkages, which in turn requires competitive processes.

Bibliography

Alemayehu Seyoum Taffesse, Bernard Tanguy, David Spielman and Eleni Gadbre-Madhin, Cooperatives for Staple Crop Marketing, IFPRI (2010).

Alemayehu Seyoum Taffesse, Paul Dorosh and Sinafikeh Asrat, Crop Production in Ethiopia: Regional Patterns and Trends, IFPRI (March 2011).

Bekele Shiferaw and Hailemariam Teklewold, Structure and Functioning of Chickpea Markets in Ethiopia: Evidence Based on Analyses of Value Chains Linking Small Holders and Markets, ICRISAT/EIAR (2007).

David J. Spielman, Dawit Kelemwork, and Dawit Alemu, Seed, Fertilizer, and Agricultural Extension in Ethiopia, IFPRI (March 2011)

Policies to Promote Smallholders Intensification in Ethiopia (draft, 2009).

Ethiopian Agricultural Transformation Agency, Pulses and Oilseed Program Final Chickpea Roadmap, ATA (May 2012)

Ethiopian Economic Association and Ethiopian Economic Policy Research Institute, Evaluation of the Ethiopian Agricultural Extension with Particular Emphasis on the Participatory, Demonstration and Training Extension System (PADETES), EEA/EEPRI (2006).

Government of Ethiopia, National Economic Parameters and Conversion Factors for Ethiopia, Ministry of Finance and Economic Development (2008).

International Food Policy Research Institute, Pulses Value Chain in Ethiopia, IFPRI (July 2010).

Richard Jones, Patrick Audi, Bekele Shiferaw and Eastonce Gwata, Production and Marketing of Kabuli Chickpea Seeds in Ethiopia: Experiences from Ada District, ICRISAT (May 2006)

USAID, Agricultural Growth Program – Agribusiness and Market Development (AGP-AMDe Project Value Chain Strategy and First Annual Work Plan (March 2012)

World Bank, Doing Business 2012 (2011).