Improvement of Postharvest Grain Systems

Grain Storage, Processhig and Marketing Report No. 103 Mlay 1985

Feasibility of a Grain Price Stabilization Program in Haiti

I FOOD & FEED GRAIN INSTITUTE MANHATTAN, KANSAS 66506

KANSAS REPORT SUMMARY

Title of Report/Publication: Feasibility of a Grain Price Stabilization Program in Haiti Authors: Roe Borsdorf, Kathy Foster, and Ekramul Haque

Period of Report/Publication: May 1985 Project Title: Improvement of Postharvest Grain Systems

Contract Number. AID/DSAN-CA-0256

Co tractor: Food and Feed Grain Institute, Kansas State University, Manhattan, Kansas 66506

SUMMARY STATEMENT

Public-sector participation in the Haitian marketing system can occur areas: grain price stabilization, in three utilization of DARNDR bulk storage facili­ ties, and data collection and analysis. A large number of possibilities exist as to the type of actions which the Government of Haiti could undertake in the general area of grain price stabilization. As set forth in the report, certain grain price stabilization actions involving direct intervention do not appear to be warranted at this time due to limiting factors. FEASIBILITY OF A GRAIN PRICE STABILIZATION PROGRAM IN HAITI

Prepared by

Roe Borsdorf, Kathy Foster, and Ekramul Haque

for the AGENCY FOR INTERNATIONAL DEVELOPMENT UNITED STATES DEPARTMENT OF STATE

AID/DSAN-CA-02 6 Improvement 5 of Postharvest Grain Systems

at the

FOOD AND FEED GRAIN INSTITUTE Kansas State University Manhattan, Kansas 66506

Charles W. Deyoe, Director

I PREFACE

This technical report has been prepared for USAID/Port-au-Prince scope of work attached based on the in Appendix A. This report aspects of grain marketing covers three general in Haiti: possible price (including analysis stabilization activities of past price data), proposed and requirements data collection processes which will facilitate implementation tion activities, of any price stabiliza­ and the utilization of DARNDR's bulk storage facilities. TABLE OF CONTENTS

Section Page

List of Tables......

List of Figures...... v Executive Summary ...... vii . . . . . Grain Price Stabilizetion ...... Aspects ...... Implications 2 Alternative Approaches...... 4 II Data Collection and Analysis ...... Data Requirements ...... Review of Sampling ...... 7 Designs...... Market Data ...... 8 ...... 8 Postharvest Loss Data ...... Data Analysis 8 ...... 8 III Analysis of Commodity Prices ...... Data Quality ...... Analytic Approacn .11 11 Long-Term Price Trends ...... 11 Market Relationships ...... 12 Seasonal Patterns ...... 18 Price/Production Relationships ...... 21 Marketing Margins and Costs ...... 34 IV Present Grain Storage Facility Status ...... DARNDR Facilities ...... 45 On-Farm Facilities ...... Market Facilities .45 ...... 46 45 V Consideration of Possible GOH Actions...... 49 Grain Price Stabilization Activities ...... 49 Utilization of Storage Facilities ...... Data Collection and Analysis ...... 54 Appendix A: Statement of Work ...... 5 7 Appendix B: Data Tables ...... Appendix C: Description 57 of DARNDR Bulk Storage Facility Appendix at Bon Repos 93 Y D: Rehabilitation of DARNDR Bulk Storage Facility at Bon Repos 101 Appendix E: References ...... 101

105 LIST OF TABLES

Table Page 1 Average Annual Rate of Increase of Long-Term Trends, 1972-1984 . . . 18 2 Calculated Temporal Marketing Margins ...... 40 3 Comparison of Spatial Marketing Margins and Costs ...... 42 4 Comparison of Temporal Marketing Margins and Costs ...... 42

V LIST OF FIGURES

Figure Page 1 Corn, Average Monthly Price with Trend, Port-au-Prince ...... 13 2 Superior Rice, Average Monthly Price with Trend, Port-au-Prince . . 14 3 All Rice, Average Monthly Price with Trend, Port-au-Prince .. .. 15 4 Millet/Sorghum, Average Monthly Price with Trend, Port-au-Prince . 16 5 Red Beans, Average Monthly Price with Trend, Port-au-Prince .... 17 6 Market Locations ...... 20 7 Harvest Periods by Crop ...... 22 8 Corn, Average Monthly Price and Harvest Period, Port-au-Prince 23 9 Superior Rice, Average Monthly Price and Harvest Period, Port-au-Prince ...... 24 10 All Rice, Average Monthly Price and Harvest Period, Port-au-Prince . . . ."...... 25 11 Red Beans, Average Monthly Price and Harvest Period, Port-au-Prince ...... 26 12 Millet/Sorghum, Average Monthly Price and Harvest Period, Port-au-Prince ...... 27 13 Corn, Seasonal Price Patterns, Port-au-Prince ...... 29 14 All Rice, Seasonal Price Patterns, Port-au-Prince ...... *. 30 15 Superior Rice, Seasonal Price Patterns, Port-au-Prince ...... 31 16 Red Beans, Seasonal Price Patterns, Port-au-Prince ...... * 32 17 Millet/Sorghum, Seasonal Price Patterns, Port-au-Prince ...... 33 18 Corn, Seasonal Price Patterns, Semi-Rural Markets ...... 35 19 Superior Rice, Seasonal Price Patterns, Regional Markets ...... 36 20 Red Beans, Seasonal Price Patterns, Semi-Rural Markets ...... 37 21 Millet/Sorghum, Seasonal Price Patterns, Semi-Rural Markets . . .. 38 22 Example of Seasonal Price Pattern Shift ...... 51

vii EXECUTIVE SUMMARY

Public-sector papticipation in the Haitian marketing areas: grain system can occur in price stabilization, utilization three ities, of DARNDR bulk storage and data collection and analysis. facil­ od in the sense These three areas are all interrelat­ that data collection and analysis grain price stabilization is a necessary precedent actions, and options to storage facilities for the utilization of are related to any possible bulk actions in the grain price stabilization area.

Grain Price Stabilization Activities Price stabilization can be accomplished purchase, through the regulation of imports; storage, and resale of domestically-produced the commodities; or the development of tne market system. Governmental import restrictions currently exist These import restrictions for cereal grains and flour. could be reformulated to predetermined quantities allow for the importation of of basic cereals at established basic objective of points in time. The this approach is to dampen seasonal during years of production price rises, especially shortfalls. This approach regulation such that would call for import imported commodities would arrive the appropriate times. in limited quantities at However, it should be realized is consumer-oriented that such an objective and does nothing for the producer. The second approach to grain price stabilization purchase, activities is public-sector storage, and resale of domestically-produced prices for commodities. commodities in Haitian markets The this type are highly correlated. Therefore, of price stabilization activity countrywide would have to be implemented basis, with the basic purchasing on a seasonal points being rural markets. price patterns and the price/production The the GOH relationships suggest intervened in the market, the that if tion seasonal price curves for normal years would shift upward and produc­ resemble the shape of the curves would those for years of production change to prices does shortfalls. Analysis of commodity not reveal any evidence that except in marketers receive windfall the case of production shortfalls profits, individual and supply/demand imbalances markets. There is the possibility in create the that GOH intervention opportunity for windfall profits would marketplace for marketers since prices would begin to act like prices in the in production shortfall years. The question may arise as to why the GOH could not simply ities at a price above the purchase the commod­ low harvest price which would price, then release the commodities act as a farm support price later in the season to moderate increases. If this is the objective seasonal basic factors of price stabilization actions, concerning moderation of four sidered: seasonal price increases must (1) global supply/demand balance, be con­ quantity of (2) timing of stocks released, stocks released, and (4) location (3) vention of stocks released. Such in the marketing system implies inter­ price stabilization that the GOH agency responsible must have the manpower, for such a function. talent, and facilities to perform This is net the case at the that grain price present time. It does not appear stabilization activities involving in procurement, the intervention of the GOH storage, and resale of cereals Seasonal price is warranted at this movements are not of the time. extreme to require such intervention,

ix nor are marketing margins excessive, in either Further, spatial or temporal terms. the lack of knowledge concerning the operation of the current system precludes the development market of any activities which might rationally into the system. In addition, fit the lack of understanding of production factors prevents the GOH from being cost able to determine appropriate prices commodities. for

The third basic approach to grain price stabilization encourage the is for the GOH to development of the market system. accomplish The activities designed to this are the creation of an appropriate storage system applicable either individual farms or to to village-level storage; the measurement harvest losses both at of post­ the farm level and within the marketing collection and dissemination system; and the of market information to as many possible within the participants as system. The management of these assigned activities should be to DARNDR. The present capabilities extensive of this agency are such that organizational improvement will resources not be necessary, but additional will be required in order to allow it to successfully carry out above activities. For futher the information on grain price stabilization ities, activ­ see Sections I, III, and V.

Utilization of DARNDR Storage Facilities Due to the inability of the GOH to effectively by intervene in the market system means of price stabilization actions involving the purchase, storage, resale of commodities, any consideration and that the existing bulk storage facil­ ities can be used to support price stabilization activities is inadvisable. The recommended action is for DARNDR to sell the storage facilities current unfinished condition in their to interested parties in public sector. the private sector or the More information on facility utilization can be found in Sections IV and V.

Data Collection and Analysis

The direction towards establishment of a sound data collection mechanism correct position for the is the GOH. This effort should be collection of expanded to include marketing and postharvest loss data. should At the same time, attention be given to procedures by which DARNDR can improve its ability analyze data and present the to results of this analysis to Priority should be given decision-makers. to the collection, analysis, and current information dissemination of on supply, demand, and patterns, price; the analysis of cropping production costs, supply responses, analysis and demand projections; and the of marketing patterns and systems. The nation as a whole benefit from the would development of the capacity within diagnose total the public sector to food and fiber marketing system appropriate problems and to prescribe system-wide solutions. Sections II and V contain further details on data collection and analysis.

x SECTION I

Grain Price Stabilization

The central policy issue in agriculture has long been that of tion, or the policy of maintaining price stabiliza­ the price of a commodity within the system. Government at a certain level policy has often been maintaining prices aimed at establishing or for agricultural products that market forces were would not result if free allowed to determine the objectives of prices of these products. such policies have been producer The increases in income support and stability, production levels due to producer consumer price protection, response to higher prices, and moderation of chance fluctuations in prices. The Government of Haiti (GOH) currently has no program of price for domestically-produ ed grain. stabilization sector There is no government intervention since all act.Lvities are handled in this of by private enterprise. The the study team's work, as presented purpose USAID/Port-au-Prince, in the statement of is to present work prepared by stabilization alternative strategies activities to be for grain price considered by the GOH. Before actions in this area, it undertaking any is essential that the GOH consider and implications of price certain aspects stabilization actions. The to Introduce a grain price GOH should not attempt stabilization program without basic understanding the processes and potential outcomes. Aspects A policy of grain price stabilization implies three which may be taken: possible type of actions (1) a farm support price, and (3) a combination (2) a maximum consumer price, of a farm support price These actions require and maximum consumer price. the participation of the government storage, and sale of commodities in the purchase, whose prices are to be either limited. In order to do this, supported or the government must have the ities to purchase, store, and necessary facil­ resell. commodities. More importantly, have the organization, ability, it must efficiently and knowledge to perform and effectively. The such operations latter implies that the government a sound and comprehensive data must have collection system and an understanding marketing system into of the which it intends to insert itself. Farm support prices are based on the concept for a commodity of providing a sufficient so as to encourage production. price there will If farmers are assured be a market for their commodities that provide at a profitable price, a powerful incentive for increasing this can this agricultural production. assurance, producers are not Without labor, willing to invest in farm and accept the risks involved inputs, expend relationship in increasing their production. between the price received Thus the by the farmer for his commodity the cost of producing and that commodity is crucial. The eationale behind the fixing of a maximum consumer prices in line with price is to hold food consumer income and the ability The expected result to purchase food products. of such an action is the out the system. stabilization of prices through­ Therefore, by fixing prices expected at the retail level, that no other price regulation it is measures will be necessary.

1 A combination of a farm support price and a maximum government establishment consumer price involves of a floor price and a ceiling commodity. The price for a given resulting margin between these salers, two prices is fixed. Whole­ processors, and retailers thus have an administratively specified margin within which to operate. The above price stabilization measures are generally or recurrent price designed to reduce random fluctuations due to chance influences Regardless of or short-term cycles. which intervention method a government tant factors selects, the most impor­ are reliable data and information, knowledge about marketing procedures, and knowledge about production, marketing, and consumption terns. Additionally, analysis pat­ must then be conducted to understand underlying factors and linkages the that operate in the market system. such, any Without expectation of success is futile. Over the past two decades an increasing number countries have of governments in developing considered it necessary to intervene order to deal in the marketing system in with food problems. These interventions of price setting, have been in the form or the operation of price stabilization developing countries schemes, For most this has involved the establishment import and distribute of state agencies to grains, to purchase and grains, and resell domesticaly-produced in some cases to carry out rationing and subsidization schemes. Implications It should be recognized that price stabilization tation procedures policy choices and implemen­ are extremely critical matters. often Confusion arc vagueness exist concerning the objectives to be achieved by policy Conflicts arise as to whether choices. the primary objective of price policy should be income stabilization, stabilization range, stabilization of prices within or price fixing. The problems a given with trying to select more these goals for price stabilization than one of policy is that high farm prices consumer prices are mutually and low exclusive. Therefore, a the outset of the decision must be made at program a3 to the ultimate objective mechanisms used of the policy. The to implement and achieve price vary stabilization objectives will greatly, according to the national institution responsible for program and the nature the of trade in commodities concerned. Price stabilization policies can be applied ac any level market.ng chain: of the production and producer, wholesale and retail. any However, intervention at one level will inevitably have an impact on other levels above For example, price stabilization and below. policies leading to uniform prices, irrespective the establishment of of distance from the market neglect fundamental or supply source, transport and utilization considerations; on quality are adverse effects noticeable in many areas where institutional competition is limited by price fixing; and pressure to protect farmers from the effects of adverse price trends induces the government to provide subsidies. The effects of a farm support price can vary with the type of farm Commercial farmers will population. generally react quickly to subsistence farmers price changes, while react more slowly over a number to the relationship of seasons. In addition between the cost of production a given commodity, and the price received for producers will also consider sibilities. alternative production pos- If non-food products present a more favorable price situation,

2 this may result in a deterioration of the local food supply. can also result in Adverse affects food marketing channels since e. product reduces business reduction in the volume of activity due to the fact generally that non-food commodities have their own marketing charnels. The setting of a maximum consumer price tends to depress ceive. Since marketers prices farmers re­ wish to maintain or increase their attempt to pay a lower margins, they will price to the producer if they above a given fixed can not sell the product price. A reduction in farm production due prices tends to discourage to the fact that it is not profitable labor. to invest in inputs and The result is therefore a scarcity of the product in the marketplace. This leads to a creation of gray or black markets. Someone who pay a higher price to the is willing to producer can sell the product to price above the fixed price. the consumer at a Consumers are willing to pay this black market because the product price on the may not be available at the regular market. A maximum fixed price in the consumer price can also cause keting margins due a widening of mar­ to scarcity of the product, the force down farm prices, ability of marketers to and the lack of price competition price is fixed by when a product's government. Without price competition, given to marketing less attention is costs since marketers are price paid concerned with forcing down the to the supplier. The net result pricing efficiency. is that the market loses its Another result of the lack of deterioration in price competition is a the quality of the product that Since the price is available on the market. is fixed by the government, quality the consumer is no longer a factor and is forced to purchase the product its quality. at its fixed price regardless of Ultimately, therefore, both the producer and the consumer are losers under this type of scheme. The administrative setting of both a farm support price price for a given product and a maximum consumer results in a fixed margin or processors and wholesalers. resale price for all This generally leads to increased ing costs and limited average market­ competitive adaption to producer ences. This type and consumer prefer­ of price stabilization action will of gray or black lead to the development markets, a decline in product quality, efficiency as noted and the loss of pricing above in the setting of maximum essential disadavantage consumer prices. The of this specific arrangement, income countries, particularly in low­ is that merchants are services prevented from offering low-cost to consumers, even though most services customers would like to have available. Thus the government low-cost is sheltering local monopolies hindering the more efficient merchants and from expanding their output by their margins and possibly cutting making way for even 6i-eaLer economies. While the moderation of chance fluctuations in can be a worthy prices to farmers and consumers objective under certa'n circumstances, think that it is dangerous to price stabilization will be always merchants, beneficial. Once producers, or consumers have experienced the margins, comforts of stabilized prices or especially where government subsidies tant to are involved, most are reluc­ forgo such arrangements. Competitive methods and reduce pressures to introduce new costs are diminished, and crystallize market organizations tend to in forms that protect the inefficient the more and hinder the expansion progressive and productive firms. of the inherent Price stabilization schemes carry danger that resources will not move in accordance with the changing demands of society.

3 No sound role models exist on which to Many base government intervention other countries have stabilization actions. consistently programs, but these programs characterized by inefficiency, are chronic low levela of effectiveness, distortions of the marketing and system as well as the production extremely high levels of subsidization, system, which is costly to fiscal process. As price stabilization the government as it policy becomes more sophisticated attempts to intervene more extensively or in The marketing system, becomes more crucial for the need administrative measures, controls, support to purchase commodities and financial at a given price, to transport ities to areas of consumption, these commod­ and co store the commodities However, even if the organizational, until needed. financial, and physical made available to implement means can be the price stabilization policy, seen whether the results it remains to be achieved by the policy under the costs of such consideration justify far-reaching intervention actions. of approaching Perhaps alternative ways the problem of price irregularities in the marketplace would be more effective.

Alternative Approaches

Rather than direct intervention into the marketing system, native forms of thrre are alter­ action which governments can pursue to reduce price fluctua­ tions in the marketplace.

If the objective of government policy is to increase ent areas can farm prices, two differ­ be addressed. The first is in question. to restrict imports of the commodity By doing this, the supply of the commodity will be this in turn will cause an reduced, and increase in the price of domestically-produced commodity. The second area is the development of store their products the ability of farmers to on the farm and thus be able sonal price to take advantage of sea- increases. Together with price this program, a credit system information system will enhance and a the small farmer's ability commodities. The credit to store system relieves the farmer and the price of cash flow constraints, information system helps the farmer determine the most profit­ able time to sell his product. If the objective of government policy is to prevent consumer, excessive prices to the the main course of action is one of liberalization and regulation imports so as to rel ease commodities of on the markets in an high effort to dampen prices due to scarcity of supply. If the objective is to moderate low producer prices at harvest sumer prices during the off-season, and high con­ a combination of import regulations improvements in on-farm and storage can be implemented to addition, specific achieve price goals. In activities can be introduced operations. to facilitate marketing One such activity is the creation provide information of a market news service to on prices to consumers, producers, way, all participants and marketers. In this in the marketing system tion and can have access to the same informa­ make better judgments concerning activity ,ales and purchases. Another is the development of quality standards for products which that high-quality goods should dictate command higher prices. A third to assure credit at a reasonable action would be cost to marketers in marketing costs. Finally, an attempt to reduce infrastructure could be strengthened adequate wholesale and by providing retail market space to be If operating costs leased to the private sector. can be reduced by these actions, the market becomes more price-efficient. Even though these actions do not constitute a direct intervention market, they still require into th--­ that the government have access production, consumption, to data on price, and market flows, as well as marketing system. an understanding of the The government can not act in isolation and hope to have a positive impact on prices. SECTION II

Data Collection and Analysis

As previously stated in Section I, before implementing a ization program, accurate grain price stabil­ data must be available on all consumption, prices, aspects of production, and market flows. Knowledge of losses system is also helpful. occurring in the The collected data must then be analyzed determine relationships between in order to factors existing in the system and lying causal agents. This the under­ will allow the policy actions to focus areas where on relevant there is a potential for positive impact. The current status of data collection in Haiti can be described as abominable. Agricultural production data, as exemplified millet/sorghum by the data for corn, rice, and in Annex B, appear to be merely sumption observational guesses. Con­ data do not seem to exist in to any form. No organized process exists collect data on production or consumption. been While market price data have collected, this has not been an ongoing gaps activity as evidenced by the many in the data sets. Data on postharvest losses are limited to a studies and do not represent a reliable few overview of the situation. It is evident that data collection and analysis needs have not been addressed in the past.

Data Requirements The study team considers that the collection of data form of cultivated on production, in the area and yields, should be given the proposed data collection the highest priority in process under the ongoing ADS II Project. data are to be meaningful If the for analysis and planning purposes, production must be limited estimates of to geographical units no larger ment. Estimates based than one depart­ on larger geographical units, such planning regions (two as the government to three departments each), will jeopardize estimates for analysis the use of and planning. Analysis involving units means that large geographical accurate relationships can not be determined tion and between produc­ consumption, or production and price. While production estimates are of first priority, alent need. they are not the only prev- To be able to use production estimates process, in a meaningful analytical consumption and marketing data must the percentage be readily available. Although of crop production moving into the as market system has been stated 25 percent, in reality no known figures To exist to support this assumption. determine consumption levels, market production flows, and farmgate prices in the sector, household budget surveys and must household marketing surveys become part of the data collection process. surveys should In addition, household budget be carried out in the urban sector tion during the same data collec­ period so that these results can later be used to corroborate market flows.

Conducting such a series of production/consumption collection surveys would lead to the of data on consumption levels, market However, flows, and farmgate prices. the mere fact of conductinE: the will surveys is no guarantee that answers be provided as to the effects of price obtain on supply response. In order to this information, the data collected must be carefully analyzed and interpreted.

Preou~PUg96k Review of Sampling Designs Based on the preceding production and considering consumption data requirements, the need to minimize sampling and operational constraints, errors, non-sampling errors, the study team recommends and Cuevas and Magill. the list frame proposed by However, there are some addressed specific factors not sufficiently within the list frame proposal. yield One is the process of collecting data on an inter-crop production system. The proposed collecting production information system for production on a farm basis will probably in a sufficient manner. However, enumerate production yields determined on the and estimated land area basis of are not satisfactory for planning purposes. Therefore, analytical and yield data should be procedure which is designed collected by a sampling to specifically sample yields. the reference period Another factor is for the survey. Since continuous cropping patterns seem to under normal weather conditions, be the idea of conducting surveys will not allow data limited to be collected on all crops the year. Insteadl, specific planted throughout data collection periods tified to coincide will need to be iden­ with the production periods according to location across the country.

If the methodology and operating system being developed tion activity will not continue for this data collec­ to exist after the end then efforts being of the ADS II Project, put forth for their development not fund continuation are futile. If the GOH can of the process, the process now and the should either be abandoned funds allocated to more productive endeavors, or the donor should commit itself to continuing agency project. this activity after the Data covering a end of the ADS II short time period of collection extremely accurate, work, even if have no great value for analytic and planning procedures. Market Data One data collection requirement that has the need for not yet been given consideration market information on prices, is collection volumes, and flows. process will need a This data totally different design from considered for the collection the design being of production and information will have consumption data. Market to be collected within the not by sporadic organized markets of Haiti, one-time surveys but rather as a continuous process which measure the movements in will prices and quantities of commodities.

Postharvest Loss Data The measurement of postharvest losses in cereals design from also requires a different that of production and consumption loss measurements data collection. Postharvest can not be included in the duced and consumed. measurement of quantities This is because postharvest pro­ on-farm losses must be separated losses and market system into losses. Quantitative losses through accurcte weighing must be measured procedures over the period interview survey of storage, not by techniques. Quality losses knowledgeable must be assessed by persons in grain quality, by means of visual inspection of commodities at specific time points.

Losses vary by crop, production year, type of pests, conditioning length of storage time, and storage procedures, processing rates of consumption, methods, transportation modes, and climate. Given this enormous variability, reliable

8 statistics regarding the type, location, causes, and degree of grain losses are difficult to obtain. postharvest However, these statistics are necessary if postharvest losses are to be reduced.

Data Analysis

Once the data have been collected, the next step is to determine the that will be used to analyze process the data and present it to decision-makers usable format. Although a large in a quantity of price data exists (see and Foster), little attertion Borsdorf has been given to the analysis The analytical work of such data. that has been done has been sporadic dinated effort. and without coor­ Even though existing price data low quality, sets are incomplete and of far more analysis could have been attempted than what was actually done. The only way for policy-makers to arrive at sound decisions concerning policy actions in agriculture is through the analysis of agricultural data. analysis will provide answers This to questions concerning the relationships between production, market flows, and prices. It can also shifts in one determine how price commodity will affect supply responses The for other commodities. end result of analysis is the ability to make intelligent choices are arrived which at through the kiowledge thus obtained. SECTION III

Analysis of Commodity Prices

The analysis of commodity prices is undertaken native in an effort to present alter­ strategies to the GOH concerning grain price stabilization actions. This analysis involves the examination of historical price patterns, domestic production levels, market relationships, and market margins.

Data Quality

An examination of the market price data for cereals and Borsdorf and Foster beans compiled by reveals a data collection effort Most data by diverse GOH agencies. sets are incomplete because of the discontinuation of data collec­ tion by those responsible for doing so, and the inability of the study locate data which had previously team to been collected. Conflicting average annual prices series of have not been reconciled. No comment statistical quality can be made on the of the price data sets since procedures collection and estimation are unknown. A review of available procedure data sets suggests that the used to construct price averages collected. is the simple average of samples Without the raw data, sample distribution to assure can not be tested so as that the simple average price is representative of actual market prices.

Farmgate or actual prices received by producers Market have never been measured. flows are not known. As previously stated, the production data able are incomplete and of poor avail­ quality. With such faulty production any resulting work on price/production data, relationships may be meaningless.

Analytic Approach

In view of the data constraints set forth above, a specific was devised in order analytic approach to address the question of price actions. Four stabilization policy primary price movement factors long-term were considered: (1) the trend, (2) seasonal price movements, (3) production disturbances which are extremely high seasonal price movements caused by production, and (4) supply/demand a decline in shifts which are changes in the supply flowing into markets amount of to meet demand at any given point latter factor is exposed in time. The to many market constraints, such edge and infrastructure as imperfect knowl­ deficiencies. However, since it is not market flows are unknown, possible to address this aspect of the marketing system and its effect on prices. Four major crops were analyzed: corn, rice, millet/sorghum, proxy price and red beans. A series was constructed from various data sets available in order to have a long enough price series to represent historical movements. proxy price series is based This on the Port-au-Prince market since most complete data the longest and sets were available for this market. bility of this To test the relia­ constructed time series, the original Prince have data sets for Port-au­ been measured against oth.r markets ments to determine if price move­ are correlated and thus indicate a significant degree of relationship between markets.

Pre Ou3- Page Long-Term Price Trends

Figures 1 through 5 present the average monthly prices for for five commodities: corn, 1968 through 1984 all rice, superior rice, millet/sorghum, beans. These and red figures also illustrate the commodities. long-term price trends for these For the most part, price movements 1-4) have for cereal grains (Figures followed the same basic patterns movements over this time period. Price for beans (Figure 5), however, have been far more erratic than for cereal grains. All commodities those show evidence of severe average monthly prices at peaks in the different periods in these time indicate that severe series. These peaks production declines in mid-1974 1976 through mid-1975, mid­ through mid-1977, and 1980 through mid-1981 caused drastic increases in prices due to decreased availability of supply. The long-term trends of average monthly prices for both cereal beans are somewhat similar grains and over the 1968-1984 time period. commodities, the price For all five trends between 1968 and 1972 slight upward are either flat or show a incline. This indicates a stability for this period. in average annual prices After 1972, however, price trends a either rose dramatically new flat trend or began a steeply-inclined to upward trend. This direction of price trends change in at this period in time coincides crease in world prices with (1) an :.n­ of agricultural commodities due production of cereal to a decline in total grains and legumes, and (2) increase the beginning of a large in worldwide inflation brought on by the decline in production coupled with other economic factors. The long-term trend for corn after 1972 is illustrated moves in Figure 1. The trend abruptly upward in 1973 and remains flat until 1978. Thereafter trend moves upward at a the significantly steep rate. However, 1983 and 1984 prices, considering the it appears that if no production affect prices disturbances occur to over the next few years, the price trend will level out at this plateau and remain a flat trend.

The long-term trend for rice is diagrammed in Figures 2 shifts from a slight and 3. The trend upward tendency to a steep upward in a near movement that results tripling of prices within a period of 11 years. As with corn, and 1984 prices for rice indicte 1983 that in the absence of production bances, the price distur­ trend may shift to a flat line. The long-term trend for millet/sorghum is shown in other commodities, Figure 4. As seen in the the price trend for millet/sorghum upward in shifts dramatically 1973 to a new trend. In contrast ever, to the other commodities, how­ the price movements in 1983 and 1984 do not give a strong indication that an expected plateau in the price trend will occur. The long-term trend for red beans is illustrated in same basic change Figure 5. It follows the from a nearly flat trend in 1978 to a steeply-inclined trend over the remaining time period. Although the basic trends for these commodities are similar, one distinction should be pointed out in the case of corn. While the trends commodities are upwardly for the other inclined beginning in 1973, flat until 1978. the corn trend remains This reaction suggests that millet/sorghum have had production may a dampening effect on corn prices by acting as a substitute cereal.

12 L4 P,

3

2

19O 1969 1970 197 1972l 1973 19714 '1975 -19797 17199 180 18 92 18 91 Figure 1. Corn, Average Monthly Prices with Trend, Port-au-Prince

Source: Table 25, Appendix B 23 22 21 20 19­ 18 17: 16 15 14 -cd 13 1­ 12 -P

11 ' 10 0 9 8 7 6 5 4 3

1968 '1969- '1970 '1971 '1972 ' 1973 '1974 195 196 17 97 '179 98 '98 '12 '98 '181

Figure 2. Superior Rice, Average Monthly Prices with Trend, Port-au-Prince

Source: Table 25, Appendix B 23 22 21 20 19 18 171

15 16-

14 - '

13 -

12 U2

10 0 9 8 7 6 5 4 3

19697

Figure 3. All Rice, Average Monthly Prices with Trend, Port-au-Prince

Source: Table 25, Appendix B 6 --I

4,P

U)

2­

196 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 198 1983 1984 Figure 4. Millet/Sorghum, Average Monthly Prices with Trend, Port-au-Prince

Source: Table 25, Appendix B 13­

12

11

10­

8'

i

6­

5­

4­

3­

1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984

Figure 5. Red Beans, Average Monthly Prices with Trend, Port-au-Prince

Source: Table 25, Appendix B The correlations between corn and millet/sorghum are prices (Table 24, Appendix B) not only significant but also extremely time high until 1976, which is the same period in which corn showed a flat price trend. The correlation between prices after 1981, while still significant, is substantially lower. further suggests that millet/sorghum, This while not as preferred a cereal may have acted as a strong substitute as corn, for corn in the Haitian diet from to 1978, but that due to population 1972 pressures and increasing consumer ences, this relationship prefer­ was not sustained after this time. A comparison of the inclined price trends of the five scribed in Table commodities, as de­ 1, reveals a high degree of similarity. true for corn This is especially and millet/sorghum when the flat trend for corn in 1973-1978 is taken into consideration.

When the trends are deflated into real price terms, it prices for corn, is found that producer millet/sorghum, and beans have Producers increased in real terms. of rice have not fared as well, since the price for their commodity has slightly decreased in real terms. The rate of price increase due to inflation commodities. is identical for each of the These results indicate that the accurately constructed proxy price series depicts the actual situation and therefore can be considered as a reliable source of information.

Market Relationships

An extensive network of markets of various types and all the domestically-produced sizes exists to handle food ciops being sold kets in Haiti. These mar­ are organized and operated by the private sector with no government intervention.

Table 1 Average Annual Rate of Increase of Long-Term Trends 1972-1984

(Percent)

In In Rate Increase Current Real Due to Prices 1 Prices Inflation Corn 13.7 4.5 Millet/Sorghum 9.2 10.4 1.1 9.3 All Rice 8.6 -0.6 9.2 Superior Rice 9.2 -0.04 9.2 Red Beans 9.7 0.5 9.2

IBased on CPI deflator, Table 1, Appendix B.

18 The market system in Haiti can be described marketers generally as a vertical system in specialize in a given which the encompasses route or geographical a given set of markets. area which price Rather then searching out opportunities, these marketers new market or prefer to establish lasting with buyers and sellers relationships within this given set of markets. This vertical system causes certain problems in the analysis tionships. In order for of market rela- any analysis to yield sufficiently to indicate market high correlations integration, trade must generally means that products be uni-directional. This must continually flow towards the manner larger markets. Because of in which the marketers operate uni-directional in Haiti, there is no trade. In addition, evidence of since market flows are unknown, no way to relate one market there is to another based on market determine if any two flows, and no way to given markets should be more highly correlated than others.

Previous studies (as noted in the references) are inconclusive of whether a high on the question degree of market price correlation relationships were exists in Haiti. Market tested by regressing Port-au-Prince against other market price series market prices series by specific which price data set. The markets from data were available for regreson 6. The resulting analysis are shown in Figure correlations between markets through 23 are presented in Tables in Appendix B. Only 11 correlations 13 regression of Pcrt-au-Prince out of 166 derived from the time series against insignificant, other market price series were or 7 percent. Those correlations were mostly which were not significant brought about through the availability of only short (small time series numbers of observations) for analysis. While a large percentage of the correlations were variation significant, there is a in the degree to which che wide other Port-au-Prince market is correlated markets. The wide variation to i due to data gaps (missing unequal time series within data and a specific set), the geographical markets to the Port-au-Prince relationship of market, and the specific time series being analyzed.

A higher degree of correlation between the Port-au-Prince urban markets market and other is shown in the later time time series (1977-81) than in series. When plotted, all the earlier regressions between the Port-au-Prince and other urban markets show market a nearly identical linear relationship narrow range of price and a very difference between markets. graphical location of An examination of the this narrow price range reveals other urban markets that 50 percent of the have slightly higher prices rice, and millet/sorghum, than Port-au-Prince for corn, while the remaining 50 prices. This price percent have slightly lower pattern does not hold true the indication for red beans, however, where is that the prices on the higher Port-au-Prince market are than on the other urban always markets. Although price between the Port-au-Prince differences exist market and other urban is it indicated that markets, at no point average price differences are any commodity. greater than 20 percent for These results imply that each as an individual urban market can be considered end market whose aim is to support the needs of the popula­ tion served by that urban market. The price correlations between the Port-au-Prince series for regional time series and the time and semi-rural markets present a more narrow range of

19 C HAITIEN +CAMP COQ

* Urban markets ONAIVES A Regional markets _( + Semi- rural markets PONT DE L'ESTERE +CERCA SVAJAL A PONT SONDE • --. ST MAR C SHINCHE

1 ESARMES

I - FRANC + ER"I

A ENPORT-AU r- PRINCE ACROIX DES OUQUETS

MIRAGOANE 0 ARON + CAMP PERRIN A FOND DES PETIT GOAVE

+ GAILLARD +FORW DES Cj A MEL TH/OTTEPigs

Figure 6. Market Locations correlations than was exhibited between Port-au-Prince and other markets. The plotted regressions, urban while for the most part identical linear relationships, showing nearly are more widely dispersed the urban markets. than in the case of In numerous cases, the regressions prices for corn and illustrate that average millet/sorghum will be higher rural markets in the regional and semi­ than in Port-au-Prince, while and red beans. the reverse is true for rice This result is in accord with ally to be the variation in prices lcgic­ expected, since corn and millet/sorghum areas and therefore are staples in the rural may command a higher price beans will in rural markets. Rice and command higher prices in urban and markets where consumers are able to pay the higher price willing associated with these products. price differences between Average the Port-au-Prince market semi-rural markets and the regional and illustrated by these regressions do not exceed 25 percent. The only market flow information available is These limited that presented by Duplan (1974). market flows for the Port-au-Prince products market show that most are moving into the Port-au-Prince of the market from regional rural markets which show the and semi­ highest correlation with the series. The Port-au-Prince Port-au-Prince times market is also more highly lying regional and correlated with out­ semi-rural markets that are graphical areas located within nearby geo­ having an adequate road infrastructure linking them to Port­ au-Prince.

Two predominant factors can be identified in the Haiti. First, market price relationships in all urban markets can be the considered as individual markets, existing correlations indicate and that the markets have nearly long-term price trends and identical general seasonal price movements. relationships exist between Secondly, close the Port-au-Prince market semi-rural markets. and the regional and Although wide price variations sions exist in plotted regres­ due to the geographical location markets, of outlyzng regional and the correlations indicate semi-rural that similar patterns exist price trends and seasonal in long-term price movements for all markets.

Seasonal Patterns Seasonal price movements result from the fact that more specific production occurs at one or periods in the year, while produced consumption of the commodities is constant throughout the year. Products thqt are sold of harvest or shortly thereafter at the time command relatively low prices which exceeds the demand due to a supply for that product in the progresses, prices marketplace. As the year in the marketplace for a given to the need commodity will increase due to cover marketing costs (storage, risk-bearing, carrying costs) and also due to the fact that demand begins to outweigh supply. The harvest periods for different crops are diagrammed in Figure monthly prices and corresponding 7. Average crop harvesting periods for superior rice, red beans corn, all rice, and millet/sorghum are illustrated through 12. These figures clearly in Figures 8 indicate that there are seasonal in price, and that such variations variations will change from general pattern and intensity year to year in the of movement. The seasonal a recurring phenomenon movement of prices is and one that is closely linked to the harvest period. More important to the understanding of this aspect need to determine of price m(vements is the average seasonal price movements, the rates of change during

21 R-) R') Beans

-1 Sorghum

0 F: 0 Rice

Corn

Jan Feb Mar Apr May June July Aug Sep Oct Nov Dec Month

Figure 7. Harvest Periods by Crop

Source: DARNDR 6 (

LI

w4- co

30

2

1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 Figure 8. Corn, Average Monthly Prices and Harvest Period, Port-au-Prince

Source: Table 25, Appendix B 23­ 22 21­ 20­ 19­ 18- QlJ 17 4 16­ 15 ­

13' a) 12 o 0 11 Ij. 10 9 8 7 6 5 4 3------3------" - " 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 Figure 9. Superior Rice, Average Monthly Prices and Harvest Period, Port-au-Prince

Source: Table 25, Appendix B 23­ 22­ 21­ 20 19 18 17 ' 16­

15- c

13 ~r

12 - 0

10 9­ 8­ 7' 6­ 5 t _t 4

1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 198'1 Figure 10. All Rice, Average Monthly Prices and Harvest Period, Port-au-Prince

Source: Table 25, Appendix B 14­

13­

12­

11­

10.

Cc

70

9- 7- -.

4­

1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984

Figure 11. Red Beans, Average Monthly Prices and Harvest Period, Port-au-Prince

Source: Table 25, Appendix B 6 a

U)

3 0 0

2­

1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 Figure 12. Millet/Sorghum, Average Monthly Prices and Harvest Period, Port-au-Prince

Source: Table 25, Appendix B so-called normal seasons and the comparable deficit rates of change during production. The measurement periods of determination of seasonal movements will of (1) margins between allow the sequences price lows and price highs, between these lows and (2) the time price highs, (3) the rate of increase movements, and (4) the differences in seasonal and between years of years of production shortfalls. normal production comparison These measurements will of marketing margins provide for a possible and costs over time and impacts on price movements allow appraisal of if the GOP inserts itself ing system into the market­ with a price support program. The average seasonal price patterns are years without divided between price patterns production shortfalls (so-called in for years normal years) and price following production shortfalls. patterns Figures 13 through average seasonal price patterns 17 illustrate for corn, all rice, superior and millet/sorghum based rice, red beans, on the Port-au-Prince proxy of these commodities price time series. Each has its own unique seasonal pattern which is directly related to the harvest periods irndicated in the figures. In the case of corn, Figure 13, a seasonal harvest period peak occurs at the beginning in June, and prices decline of the January, thereafter to a low some two months after in December and the end of harvest. The increase is 25 percent average seasonal price above the market price lows in years of normal produc­ tion. In years with production higher, but also increases shortfalls, the price at a more rapid rate increase is not only above market lows. with highs being 45 percent

Figures 14 and 15 depict seasonal price movements for rice. superior rice both have relatively All rice and the same basic patterns of and two low periods per year, two high periods related to the two production the patterns are generally seasons. Although the same, each has a unique in the rate of use and decline. shape of price movement The percentage of increase is substantially below in the price rise that for corn. In normal rise in the years there is a 10 two months following the percent August December-January harvest. harvest, however, prices After the for all rice increase to percent above market lows, slightly over 20 while the price for superior percent. Price rice climbs only 10 increases in years of production shortfall are 30 percent above market lows. Seasonal price patterns for red beans years, are illustrated in Figure prices show a slight increase 16. In normal of 10 percent after the harvest, with a more substantial December-February increase of 15 percent harvest period. However, after the June-July production shortfalls create price, with an increase a dramatice shift in of slightly over 45 percent harvest period. after the December-February The price reaches its peak the summer harvest in May and rapidly declines period, followed by during another increase of 20 percent market lows. above

Figure 17, containing seasonal price patterns relatively slight for millet/sorghum, increase in price after shows a production harvest for normal years. shortfall shift the Years of price increase upwards to percent of market lows slightly over 30 within 2-3 months after harvest. more or less This pattern remains at the same level until slightly before the next harvest.

28 a) (I) 60

S50 U 0 0 -- .1 Average Seasonal C1 300 Price Movements Following Production 20 10 c-i 0

, -- Average Seasonal Price Movements Without Production Shortfalls, 1968-84

Harvest Period

J F M A M J J A S 0 N D Month

Figure 13. Corn, Seasonal Price Patterns, Port-au-Prince

Source: Table 25, Appendix B 60 50 40 30 Average Seasonal Price Movement During Period Following Production

o10 o 0 SAverage 0 Seasonal Price Movements Without Production Shortfalls, 1968-84

Harvest Period

J F M A M J j A S 0 N D Month

Figure 14. All Rice, Seasonal Price Patterns, Port-au-Prince

Source: Table 25, Appendix B 50

o " 1968-84,verage Seasonal Price Movements Following Production Shortfalls,

O 20 1:0 10

0 a)Average Seasonal Price Movements Without Production Shortfalls, 1968-84

Harvest Period

J F M A M J j A S 0 N D Month

Figure 15. Superior Rice, Seasonal Price Patterns, Port-au-Prince

Source: Table 25. Annpndiy R U) 60 0

Average 20 Shortfalls,Seasonal 1968-84 Price Movements Following Production 20

S10­ (U C.) 01

0- Average Seasonal Price Movements Without ro Production Shortfalls, 1968-84

Harvest Period

F M A M J j A S 0 N D Month

Figure 16. Red Beans, Seasonal Price Patterns, Port-au-Prince

Source: Table 25, Appendix B a) Ci) 6o 50 40 0Average 30 1968-84 Seasonal Price Movements Following Production Shortfalls,

a 0) Ci 10 4-)

C) 0 s-s 0-. Average Seasonal 1968-84 Price Movements Without Productf on Shortfalls,

Harvest Period

J M A M J J A S 0 N D Month

Figure 17. Millet/Sorghum, Seasonal Price Patterns, Port-au-Prince

Source: Table 25, Appendix B An analysis was conducted on the prices markets1 of these commodities in the to determine if (1) the semi-rural same general seasonal patterns and (2) the degree of held true, seasonal price increase was Port-au-Prince nearly the same as for the market. The results are presented in Figures 18 through 21. In the case of corn, Figure 18, seasonal price patterns as movements show the same general for Port-au-Prince, but the degree tially higher. of price increase is substan- This is especially true for where years of production shortfall, prices increase to nearly 75 percent above the market lows. degree of price increase in The greater the semi-rural markets is most fact that imports of likely due to the corn into Port-au-Prince may that Port-au-Prince dampen price increases, and appears to be a more price-competitive be noted that there is market. It should a price lag difference of one markets and Port-au-Prince, month between semi-rural which is as to be expected which allow new given market flows corn to show up sooner in semi-rural markets than in Port-au- Prince.

Price movements for superior rice, illustrated in Figure double curve as for 19, show the same Port-au-Prince seasonal movements. quite closely with Price movements track Port-au-Prince, with the exception December-January of prices following the harvest period being slightly slightly lower higher in normal years and in yea.os of production shortfall. can This illustrates that rice be considered a commercial crop rather than a subsistence crop. Seasonal price movements for red beans in the semi-rural markets are Figure 20. These price patterns shown in Port-au-Prince foilow the same basic pattern market with two exceptions: as in the shortfall (1) prices in years of production are higher in the Port-au-Prince and market after all harvest seasons, (2) price movements in normal years are higher in Port-au-Prince semi-rural than in market after the June-July period. The price patterns for millet/sorghum, diagrammed in tial differences between Figure 21, show substan­ Port-au-Prince and the semi-rural the markets. Prices in semi-rural markets not only show also dramatically higher increases, follow a substantially different but they price pattern. If one compares the movements for corn in Port-au-Prince seasonal and semi-rural markets seasonal price movements for with the millet/sorghum in both markets, terns are similar. This the price pat- reflects the fact that millet/sorghum a substitute seems to act as for corn in both market groups.

Price/Production Relationships

Since available production data are of low quality, can be made are the only calculations that crude estimations of the degree to which production shifts will affect prices of the commodities.

For rice, an analysis of production data and yearly average that a decrease of 1 percent prices indicates in production would lead percent in price. to an increase of 1.5 A decrease of 10 percent in production would lead to an increase of 9 percent in price. This is nearly a 1:1 ratio.

1 Superior rice had to be calculated based on the lack of sufficient regional markets due to data in semi-rural markets. No calculations could be made fcr all rice.

34 CJ8o

ci)60 70 \ Average Seasonal Price Movements Following Production Shortfalls 1968-84 50 //

30)

430 C)10

S0 Average Seasonal Price Movements Without Pro uc ion Shortfalls, 1968-84

Harvest Period

F M A M J j A S 0 N D Month

Figure 18. Corny Seasonal Price Patterns, Semi-Rural Markets

Price Data Source: Borsdorf and Foster U1)

o 40 3 Average Seasonal Price Movements 20 - .'1968-84 Following Production Shortfalls a,10

o C) PL4 Averagefalls, 1968-84Seasonal Price Movements Without Production Short-

Harvest Period

J F M A M I j A S 0 N D Month Figure 19. Superior Rice, Seasonal Price Patterns Regional Markets Price Data Source: Borsdorf and Foster oj U] Q1) Q 50 40 Average Seasonal Price Movements Following Production Shortfalls, 1968-84

0

Q)0

SC 1968-84

Harvest Period

J F M A M j J A S 0 N D Month Figure 20. Red Beans, Seasonal Price Patterns, Semi-Rural Price Data Markets Source: Borsdorf and Foster ci)

t 59 f 0 o30 - - Average Seasonal Price Movements Following Produc­ tion Shortfalls, 1968-84

20 Ca 10 Q 0

)Average Production SeasonalShortfalls, Price 1968-84 Movements Without

Harvest Period

J F M A M J j A S 0 N D Month

Figure 21. Millet/Sorghum, Seasonal Price Patterns, Semi-Rural Markets

Price Data Source: Borsdorf and Foster For corn, analysis of production data and yearly average a decrease of 1 percent prices indicates that in production would lead in price, a decrease to an increase of 2 percent of 10 percent in production 7.5 would lead to an increase percent in price, and a decrease of an increase of 15 percent in production of 11 percent in price. would lead to This commodity also shows ratio. nearly a 1:1

For millet/sorghum, an analysis of production failed to show data and yearly average a relationship between production prices possible and price. There are explanations for this lack two of correspondence: either data sets are bad, the available or millet/sorghum is so close the calculation of a substitute for corn that a relationship between millet/sorghum production and is impaired. price For red beans, an analysis of production indicates data and yearly average that a decrease of 10 percent prices in production would lead crease of 13 percent in price. to an in­ A decrease of 20 percent to an increase of in production would 19 percent in price. Again, this is nearly a 1:1 ratio. Marketing Margins and Costs Marketing margins can be divided into two major categories. gory is spatial marketing The first cate­ margins, or the margins sales price between purchase price which result from the purchase and marketplace of a commodity from the and the resale of that producer or commodity in another marketplace limited time period (less within a than one month). The second marketing margins, category is temporal or the margins between purchase to holding a purchased price and sales price due commodity for an extended tion of selling period of time in expecta­ the commodity at a higher price as seasonal prices increase. Exact spatial marketing margins in Haiti sented are unknown. Based on in references given in evidence pre­ Appendix E, it could be stated marketing margin between that the average the price paid to the is approximately producer and the retail price 25 percent of the retail price of average of the commodity. An analysis monthly prices in all Port-au-Prince semi-rural markets" in relation proxy time series reveals to the difference that the expected spatial as a percentage of price average Port-au-Prince sales percent for corn, 17 percent prices should be 12 for superior rice, 12 13 percent for millet/sorghum. perccnt for red beans, and If the 25 percent stated above average marketing margin is accepted as a sound approximation the calculated of the total margin, then spatial margins between markets since this would should be reliable estimates leave a difference of 8 to 13 percent between the farmgate and wholesale prices. Temporal marketing margins, as shown in Table seasonal price patterns 2, are calculated from the established in the previous normal production. subsection for years of These calculated temporal gain a marketer margins express the expected could achieve by purchasing a commodity at its holding it for resale until low price and the price reaches its highest increase percentage of over the purchase price.

1 Superior rice is calculated based on regional markets. tions No calcula­ could be made for all rice.

39 Table 2

Calculated Temporal Marketing Margins (Percent of Purchase Price)

Commodity Port-au-Prince Average of All Proxy Time Series Semi-Rural Markets Corn 25 31 All Rice First Season 8 Second Season 13

Superior Rice First Season 10 Second 16 Season 8 4 Red Beans First Season 9 6 Second Season 13 8 Millet/Sorghum 9 26

Source: Figures 13-21 Table 25, Appendix B Borsdorf and Foster

To determine whether marketing margins can be considered they must be compared normal or excessive, to marketing costs. There are marketing costs: five broau categories of (1) carrying cost, or the reLurn commodity on the money invested in a for the purpose of resale, (2) holding and quality cost, or the cost of storage maintenance of the p"oduct purchased for resale, (3) risk cost, the cosL of' accepting risk of ownership or of a commodity due to the for physical (quantity) or quality potential '.oss, (4) transport cost, or the moving a product from one market cost of are? to another, and (5) overhead the average return a marketer should cost, or receive over and above all other order to compensate him for the costs in effort expended in marketing a commodity. overhead cost should be considered The as a return to labor and management. While very little research on marketing costs of available has been done in Haiti, a review references in Appendix E assisted crude in the development of a minimum marketing costs. These set of minimum marketing costs are as follows: Carrying cost - 4 percent of purchase price of the commodity per Holding cost - month 0.2 percent of purchase price of commodity the per month Risk cost - 1 percent of purchase price of the commodity per month

40 Transport cost - 6 percent of the purchase price of the commodity Overhead cost - 11 percent of the purchase price of the commodity A comparison of marketing margins and costs description of the is presented so as to marketing system in which provide a pate. The reader must the GOH may decide to remember that the results partici­ calculations, often developed presented herein are with only the most crude a base. This approach tenuous of data to be used has been taken in order as which should be to express all considerations weighed by the GOH prior to undertaking any grain stabilization activities. price A comparison of spatial marketing margins Marketing margins and costs is presented are large enough to in Table 3. cover marketing costs and margins. allo7 for net A comparison of temporal marketing Historically, margins and costs the average expected is shown in Table 4. all margins have not been marketing costs. As a high enough to cover pocket result, marketers have expenses such as storage recovered only their out-of­ to costs. The returns have fully reimburse them for not been sufficient thq money, labor, and management marketing effort, invested in the nor for the risk assumed. The net margins generated by the immediate resale Table 3, are used to of commodities, as shown support the negative margins in the product for late which result from holding sale, as shown in Table evidence to suggest 4. There is no available tijat any amount of excess marketers. Given profit is being the high level of generated by transport expenses, costs incurred with it appears that carrying cost and costs. margins are only large This is not to say theje enough to cover Such windfall are no "windfall" profits profits could be generated in the system. prices can be forced in years of production upwards in the marketplace shortfall when Windfall profits may also due to supply shortages. occur in a single marketplace imbalances on any given due to supply/demand market day. The first failure and poor management case is a result of production of food product imports. tight supply/demand situation The second is due in Haiti which way to a ities from one market cause diversion of commod­ to another based on Because of a lack of the expectations of information, these expectations marketers. may lead to losses may not be correct, in one marketplace with and concurrent windfall profits another marketplace. in There is one other element in marketing involves the margins to be considered. processing margins for This sUgh as polished products which undergo rice and milled corn. a change in form, the prices for In the case of rice, paddy and polished rice an analysis of costs may reveals that a portion be covered by the marketing of the milling a 65 percent margins shown in Tables milling yield for paddy, 3 and 4. Using the return to milling and based on the data provided varies widely. Local by ODVA, indicate that market prices in the the return to milling (processing ODVA area percent of the sales margin) ranges from -10 price of polished rice, to +10 season. These processing depending on the year costs are not included and the 3. When the processing in the costs shown in margins are 0 or negative, Table then this cost is covered

41 Table 3 Comparison of Spatial Marketing Margins and Costs (Percent of Purchase Price)

Calculated Calculated Commodity Marketing MMargin Marketing 2

Corn 22 16 Superior Rice 27 16 Red Beans 22 16 Millet/Sorghum 23 16 ISpatial marketing margin plus 10 pecent for wholesale/producer difference to match price with crude estimates in references. 2 Carrying cost, risk cost, transport cost, and overhead cost for 10-day period. a

Table 4 Comparison of Temporal Marketing Margins and Costs (Percent of Purchase Price)

Port-au-Prince Average Proxy Time Series of All Semi-Rural Markets Commodity Months to Margin Achieve Margin Cost1 Margin Achieve Margin Costl Corn 25 5 37 31 6 40 All Rice First Season 8 3 25 _ Second Season 13 3 25 _ - - Superior Rice First Season 10 3 25 Second Season 16 2 8 2 20 20 4 2 20 Red Beans First Season 9 3 25 Second Season 6 3 13 3 25 25 8 2 20 Millet/Sorghum 9 7 45 26 5 35 ICarrying cost, holding cost, risk cost, and overhead cost indicated. for months

42 by the net margin as determined from Table 3. When further related poral margins, Table to tem­ 4, no excess profits can be determined.

43 SECTION IV Present Grain Storage Facility Status

DARNDR Facilities Historically the GOH has not been involved storage in ownership or operation facilitiev. However, in of grain the 1970s the GOH began constructing grain storage bins and other metal associated equipment. The Commercialisation Agricole Service National de (SENACA) indicated that silos each a total of 10 modules of were to be located at Bon 5 and Repos, St Marc, Gonalves, Mirebalais, Petit Goave. Adequate documentation was not available to plans and programs for these determine the facilities. The study team facility near Port-au-Prince, visited the Bon Repos but it was not possible to see the facilities at any other locations.

The as-built layout of a module at Bon Repos and the detail facility status are given of the Bon Repos in Appendix C. This unfinished five modules, each facility consists of containing one combination drying-storage bins, a scalper (cleaner) bin, four storage and other associated equipment. imported from the All equipment was U.S. and installed by a local estimated contractor on site. The storage capacity of this facility is 5,000 MT of dry grain sorghum. On-Farm Facilities

Farm sizes are so small and fragmented that it is titative data on impossible to gather quan­ facilities being used for grain been conducted storage. No study has yet in this regard. However, a general description can be given the various on-farm of storage methods being used. In the production of corn, corn in the husk is semi-dried to 45 days after in the field for 30 maturity. The large producers percent harvest the grain at moisture content and sun about 16 dry the grain to about 15 percent a concrete drying surface moisture on for about a day. The semi-dry unhusked small farmers harvest the ear corn and hang the bundled drying ears from tree trunks for and storage. The storage structure final Gouanes used is locally known as "gouane." are used exclusively for corn, and are built at very tools and materials available low cost with on the farm. Gouanes do protection against insects, not provide adequate molds and rodents. Desrouilleres the grain losses (1981) evaluated in traditional storage including Madian gouanes used in Salagnac areas. Three loss evaluation and criteria were used: kernels affected by insects (1) number of and fungi, (2) affected weight 1,000 kernels, and in a sample of (3) percent weight loss. The on the basis study indicated that losses of all three criteria increased Weight with the storage duration. losses were found to vary from 0.2 to 14.9 percent for a of 32 days to 200 days, with storage time 9.8 percent maximum loss for 6 However, it is unclear a 1 3-day storage. whether weight losses due to drying are included in this figure. In the production of sorghum, the semi-dry and sorghum heads are harvested stored in the house in wooden by hand boxes or in the attic area fire. Sorghum is shelled above a kitchen in the same manner as crete drying corn and spread on the con­ floor in front of the house or on mats to further dry the before storage in wooden grain boxes, sacks, or other containers.

45 Previous Page 01 1 Rice production is concentrated being mainly in the valleys, with the largest area of production. Artibonite Sorenson and Chung rice heads were harvested (1973) reported that from the paddy stalk, transported by manual labor, and then from the field threshed on concrete floors with sticks or by repeatedly or mats by beating trampling under the feet. dried on mats or concrete Rice was then sun­ floors. An alternative method in the field to dry. was to leave the rice Dried heads were then harvested quarters on a raised and stored in the living platform without any walls and roof. The rice threshed when marketed. was

In the last few years, new rice varieties have been many farmers introduced and accepted by in the producing areas. These from Texas, varieties, such as long-grain are easier to dry than the rice harvesting traditional short-grain varieties. and threshing methods have New involve also been adopted. These harvesting the entire paddy methods before. stalk at a higher level The fields are generally of moisture than flat, and they are dry season. All of these elements during harvest make it possible to change method. Under the new method, the threshing threshing is done in the canvas sheet by holding field on a large a handful of stalks at the heads against a stone, bottom and beating the lumber, a rubber tire, ohject. The or any other convenient threshed rice is then transported hard of to the farmer's house, which is a multipurpose concrete in front drying floor. Sun drying on this drying floor of rice is done or on mats. A typical farmer rough rice in a 2-season harvests about 2 MT harvest year. After of drying, the grain is usually stored in a large flat woven basket without any lid. In bean production, the entire bean plants are harvested, the attic above the dried, and stored in kitchen area where the final Beans are threshed drying is accomplished. before consumption and marketing. Beans may be redried after threshing by spreading on a drying floor or a mat. Because of the extreme limitations of data, the on-farm can not volume of the grain stored be quantified. However, it that a lack is evident from various references of sufficient on-farm storage capacity is partly responsible lower for crop prices during the harvest.

Market Facilities An accurate description of these facilities can not be lack of data given because of the and fragmented nature of operations. Storage facilities mainly consist of warehouses or general-purpose the markets. The warehouses rooms in are generally owned the government by the traders. In Haiti, has not yet become involved in grain storage and marketing operations.

In relation to market facilities, one ongoing project in to combine the need for Haiti that attempts producer storage with the participation in need to encourage producer the market system is a project funded by the Deutsche Welt­ hungerhilfe (DWHH). The storage component of this project initially involved the 8 x 10 x 3 m pilot warehouse construction of a and an adjacent grain N~gres. The project drying floor in Fond des was then extended to include the construction of 18

46 similar warehouses - 10 in the nite), northeast, 3 in the central and 5 in Fond des Negres. region (Artibo- Construction was begun region in December, 1982, in the northeast and in the other two regions in May 1983. By March construction was begun 1985, 6 warehouses in the N~gres were completed. northeast and 2 in Fond The concrete warehouses des and construction were built by local engineers crews, including the area farmers. The estimated capacity of storage the project is 1,800 MT. The warehouses were built on the the expectation of generating grain from the area farmers profit by buying selling during harvest when prices when prices are higher. are low, and Any profit is to be the warehouse and the equally divided between farmer whose grain is sold. is run by a manager Each operational warehouse selected by the area farmers. chemicals and The project also provides technical assistance to the warehouses, and is installation of rice considering mills in the project area.

47 SECTION V Consideration of Possible GOH Actions

A large number of possibilities exist could undertake as to the type of actions which in the general area the GOH considering of grain price stabilization. these possibilities, the When decisive criteria reader should keep in mind exist which allow any that no allegations determinations to be made concerning that (1) farm prices are too prices are low, even at harvest time, the major constraint in increasing (2) low the lack of marketing agricultural production, and infrastructure is the (3) of output major constraint in the expansion of a given commodity and the improvement of' producer income. Grain Price Stabilization Activities There are three basic approaches to grain the GOH can price stabilization activities take under consideration. Price which through the stabilization can be accomplished regulation of imports; the domestically-produced purchase, storage, and resale commodities; or of The the development of the market specific objectives that the system. stabilization GOH wishes to achieve actions are unknown. through grain price As pointed out in objectives need to be established Section I, definite before any actions are again be emphasized that price taken. It should stabilization can be implemented purposes and some of these for different purposes serve conflicting interests. The GOH can consider using the regulation stabilization. of imports as a form Governmental import restrictions of pr 2e grains such currently exist for cereal as corn and rice, and price These controls are in effect governmental actions help for flour. limiting maintain cereal prices the supply. These import at high levPld by restrictions could be reformulated for the importation of predetermined to allow lished quantities of basic cereals points in time. The price F.t estab- policy involving flour could reformulated to coincide also b'! with regulated import controls. The basic objective of using dampen this approach to price stabilization seasonal price rises, especially is to during years of production This approach would call shortfalls. for import regulation such would arrive in limited that imported commodities quantities at the time upward movement. seasonal prices begin their Prior to the new harvest season, so as not to drive imports would be curtailed harvest prices lower than conditions. However, normally expected for current it should be realized that sumer-oriented and such an objective is con­ does nothing for the producer. approach may The impact of such even cause a decline in producer an harvest season, prices, especially during since marketers may not be the without the willing to purchase and store assurance of a price increase grain costs. The sufficient to cover marketing study team considers this approach ment of to be valid only if the potential impact of import a measure­ prices, regulations on market prices, and the market system is carried producer the GOH out prior to taking any wishes to give detailed consideration action. If data and analysis to such an approach, far more will be required than can be further emphasizes presented in this report. This the fundamental need for data collection and analysis as described on pages 54 and 55.

Previou3 Page 49 Another approach to grain price stabilization activities purchase, store, and s for the GOH resell domestical ly-produced to of commodity prices commodities. The analysis in Section III sets the sible impacts of stage for determination such an activity on the of pos­ prices for commodities market system and market in Haitian markets prices. The that the are highly correlated. consequence of a reduction This means in supply due to market area would be felt GOH purchases in one in many other market areas. price stabilization activity Therefore, this type would have to be of basis, with the implemented on a countrywide basic Purchasing points purchases would being rural markets. be spread evenly over In this way, the markets and a serious market could be avoided, impact in one thus protecting all markets. Average seasonal price patterns extreme as shown in Section III except in years of production are not considered problem, shortfall, however this not a marketing problem. is a production production The seasonal price patterns relationships suggest and the price/ seasonal that if the GOH intervened price curves for in the market, the normal production years shape of the curves would shift upward would change to resemble and the shortfalls. those for years of This would mean that seasonal production expected in prices would rise more normal production years. rapidly than the GOH Taking as an example has the capacity to the case of corn, store approximately 1 storage facilities were 10,000 MT (if all bulk in the condition to do out of the market in so). If this amount were a normal year, it would taken price curve would be expected that the shift as shown in Figure seasonal but it is 22. Not only is the also shifted to the left, curve higher, age increase which indicates that the will rise more rapidly. rate of percent- act in GOH purchases in the the same way as a production market system would free market shortfall because available system is being reduced. supply to the would have The private-sector marketing less cr;,wodity to handle, system consequently prices to producers and consumers would te higher. Analysis of commocity prices does not reveal receive windfall profits, any evidence that marketers except in the case supply/demand imbalances of production shortfalls in individual markets. and GOH intervention There is the possibility would create the opportunity that keters since for windfall prices in the marketplace profits for mar­ production shortfall would begin to act like years. Also, additional prices in be created in supply/demand imbalances specific markets by the could purchases. decrease in market supply These reactions could due to GOH further heighten the expected pattern seasonal price shown in Figure 22. The seasonal price patterns in Section III indicate throughout the harvest that prices decline season, reaching their the harvest season. lows shortly after the GOH purchases, if begun end of assist in maintaining during the harvest season, prices above the normal should ing the average seasonal lows, thereby annual market price. increas­ the amount This increase would be taken out of the market in proportion to by the GOH, as described duction relationships in the price/pro- in Section III. At this point, the question may arise purchase as to why the GOH the commodities at could not simply a price above the low act as a farm support harvest price which would price, then release the commodities later in the

'Approximately 5 percent of production.

50 a)

C) ­ t 50 40 Stock Rel-U1ss Beginning Latern Than March Stock Re:aL..; Begi.ing in March SStu,- ck RcA1as±tL Bc2r ir in Hovwleble

C Expcctud Seasonal Price Pattern with GOH Purchas:- P S0

Q- Normal Average Seasonal Price Pattern, Semi-Rural Markets, Figure 18

Harvest Period

J F M A M J j A S 0 N D Month

Figure 22. Example of Seasonal Price Pattern Shift season to moderate seasonal price increases. If this is the objective price stabilization actions, of seasonal "our basic factors concerning price increases must moderation of be considered: (1) global ance, (2) timing of stocks supply/demand bal­ released, (3) quantity of stocks released, and (4) location of stocks released. In the case of Haiti, the global supply/demand insufficient supply. balance is essentially one If quantities are removed of agency not currently from the market system by involved in the marketing an further imbalance process, this will create due to the fact that existing a how the actions marketers will be unsure of the intervening agency of will affect them. Therefore will react as if they there were a production shortfall. Timing of stock release is critical to the success of action. If stocks are released any price moderation at the time when seasonal upwards, such as in November, prices begin to move one would not expect moderated since marketers the season price to be will have already purchased level nearly equal commodities at a price to the GOH purchase price commodities and they will have to sell at prices which will generate these operating a sufficient margin to cover expenses. The need to cover their purchases operating expenses also since the release price applies to GOH would have to be high enough the original purchase price to recuperate plus any marketing costs. the release price The only way to reduce is if the GOH is willing costs. If to subsidize these stocks are released beginning marketing upward in March when prices turn, the expected effect take a strong of a new seasonal high and the curve would begin price may be dampened, to follow the normal pattern although at a higher as shown in Figure 22, peak than when stocks are Stocks released released beginning in November. after March, when seasonal would have prices have reached a tendency to turn the their peak, during market downward towards the new harvest period. the level prevailing However, since the have a sound understanding market system would not of when stocks would be not be able to prevent released, these actions may seasonal prices from going even higher before leveling out. Stock release quantities will have to be carefully avoid destabilizing calculated in order to individual markets. Because tities the release of large of stock would drive prices quan­ sharply downward and marketers out of the system, could force some stock releases should occur However, when selling into in measured amounts. a fragmented system composed such as exists in Haiti, of small marketers, the GOH will incur much it is more higher marketing costs since expensive to sell in small lots than in large quantities. The last factor to be considered in this type of stabilization location for the release action is the of stocks. If stocks the other stated are held at Bon Repos storage facility locations, and tions would have then release from these some adverse price effects loca­ prices in on different markets. rural markets would continue Seasonal Figure their upward movement as 22. However, seasonal price shown in areas movements in markets located would be decreased much near storage more than shown in Figure volume being released into 22 because of the the market, even though it over time. This would may be done in small lots damage producers and consumers prices high, and would in rural areas by leaving destabilize the market system into which stocks serving the market areas were released. Any expectation destabilization of stock release could cause in these markets even before stocks are physically released.

52 Such intervention in the marketing system implies sible for price stabilization that the GOH agency respon­ must have the manpower, to perform such a function. talent, and facilities This is not the addition, given the case at the present time. current calculated marketing In sector, it may be necessary margins in the private for the GOH to subsidize such intervention because the operating costs of the marketing costs incurred assumed to be much by the GOH can higher than those incurred be by the private sector. It does not appear that grain price .Lrtervention stabilization activities of the GOH in Drocurement, involving the warranted at storage, and resale of this time. Seasonal price cereals is require such movements are not of intervention, nor are the extreme to marketing margins excessive, spatial or temporal terms. in either Further, the lack operation of of knowledge concerning the current market system the activities which precludes the development might rationally fit of any of understanding into the system. In addition, of production cost factors the lack to determine appropriate prevents the GOH from prices for commodities. being able the need These deficiencies emphasize for data analysis as detailed later in this section. The third basic approach to grain price stabilizacion courage the development is for the GOH of the market system. to en­ -rices is to increase If the objective of stabilizing the producer price at price increases, harvest and to damp-enL then several different seasonal actions can be considered. In order to help farmers obtain higher prices priate storage system for their products, an applicable to e'ther individual appro­ storage should farms or to village-level be developed. Given current IV, conditions as set this provides the opportunity forth in Section With to cure existing deficiencies the creation of a storage in the system. system which will allow the quality of their harvested producers to preserve crops, higher prices commodities as seasonal can be acheived for the prices increase. In addition quantity of grain available to higher quality, the is also greater, thus should Ie dampened as seasonal price increases the supply of grain over opment of such a time is increased. storage system will require The devel­ market credit the accompanying for farmers to allow introduction of them to carry their own activity gives the producer inventories. This the opportunity to obtain products and a higher price for his to become more firmly entrenched in the marketing system. Another area which should be addressed is postharvest level and within the losses both at the farm marketing system. The occurr*.ng in these degree of postharvest systems should be assessed losses and amount of by measuring the type, loss, and based on these location be developed findings loss control procedures and introduced into should system both the on-farm or and the marketing system. village-level storage This specific set of greater quantity of grain actions will lead to for farmers and marketers a dampen seasonal to sell, thus tending price increases. It will to command also allow producers and a fair price for a quality marketers to product since the quality not of the product has deteriorated during: storage. Finally, development of the marketing system can collection and dissemination be encouraged through of market information the possible within to as many participants the system. The specific as concerns the needs of the marketing different types of information system as procedures need will have to be defined will have to be developed and o serve these needs.

53 The management of these activities should be assigned capabilities of this to DARNDR. The present agency are such that extensive will organizational improvement not be necessary, but additional resources will be required allow it to successfully in order to carry out the above activities. Utilization of DARNDR Storage Facilities Due to the inability of the GOH to effectively intervene by means of price stabilization in the market system actions involving the purchase, resale of commodities, any storage, and consideration that the existing ities can be used to support bulk storage facil­ price stabilization activities If utilization of the unfinished is inadvisable. DARNDR storage facilities following options are the is desired, the only alternatives which can be considered viable at the present time. 1. Sell the storage facilities in their current unfinished condition terested partie in the private to in­ sector or the public sector. would not require any This option investment on the part of DARNDR other than the cost of administering the sale. 2. Complete the construction of the facilities and sell them parties in the private to interested sector. This option requires resources in the storage the investment of facilities, which constitutes risk for DARNDR. The a considerable foreign exchange cost will will probably be very high, and DARNDR realize only a minimal rate of return on its investment. 3. Complete the construction of the facilities and lease parties in the private them to interested sector. Again, this option erable investment will require consid­ on the part of DARNDR, which may see only a slight return. It is not recommended that the bulk storage bins be used for bagged products. Technical the storage of considerations prevent this option from being considered as a viable alternative. The study team recommends that option DARNDR. I be the course of In this case, no further action taken by or detail is necessary. However, option 3 is selected, further if option 2 specific actions will need to Appendix D be undertaken. should be consulted for details.

Data Collection and Analysis

The direction towards establishment of a sound data collection correct position for mechanism is the the GOH. This effort should collection of marketing be expanded to include and postharvest loss data as described in Section III.

Further, the analysis of collected data should not be developed and be ignored. Plans should a program implemented, coincidental of data collection with the implementation procedures, by which analyze DARNDR can improve its data and present the results ability to of this analysis to decision-makers. suggestions herein are intended The as guidelines for the general which should be encompassed analytic areas by such development programs.

54 The first area is the collection, analysis, information and dissemination of on supply, demand, and price. current Government planners and implementing agencies alike must have this kind of information on a regular if they are to perform effectively. and timely basis The same information is needed the functions involved in marketing to support policy analysis. Further, sector needs this type of the private information in order to make day business as well decisions on its day-to­ as its investments. The collection, semination of the information analysis, and dis­ is clearly the responsibility of a public agency. Private industry can not effectively perform such a function. It is suggested that first priority designed be given to this area, and tr' ensure accurate and that a plan be timely reports. Among the needed, the following priority types of reports is suggested: (1) current acreage, yield, and crop estimates of production of grains by department, wholesale market (2) average monthly prices for grains by market, market (3) current weekly wholesale prices for grains by market, (4) Average monthly retail for grains by market, (5) market prices average monthly market prices grains by department, received by farmers for (6) monthly stocks of grain handlers, wholesalers, by position (farm, first processors) by department, reports of dewand (7) current situation conditions and price outlook current situation for grains by market, and (8) and outlook on marketable surplus and flows of basic grains by department.

The second area is the analysis of cropping patterns, responses, production costs, supply and demand projections. This projection involves the analysis, evaluation, by department of (1) farm production and cropping costs for grains, (2) shifts patterns, (3) shifts in demand in patterns, and (4) production to alternative levels of response prices. Such work is essential planning by public implementing to the support of agencies as well as by private information is a prerequisite industry. This to such activities is (1) determining number, type, size, the proper and location of storage an, establishing marketing facilities, (2) appropriate price support levels, ment and (3) properly designing procure­ sales programs, and (4) designing programs for public and private marketing industry development.

The third area is the analysis of marketing patterns and systems. marketing analysis is now No organized conducted in DARNDR or elsewhere This information is in the government. essential to support the development policies and efficient of 'ound marketing marketing programs, both within other implementing at the ministry level and agencies. This should organization, include analysis of (1) market structure, conduct, and channels, performance, (2) marketing costs and margins, (3) functions, infrastructure the impacts of physical and affecting the development institutional and performance of systems, and (4) the Haitian marketing impacts of existing and alternative marketing and pro- cessing technologies.

The nation as a whole would benefit from the development the public sector to diagnose of the capacity within total food and fiber marketing and to prescribe system problems appropriate system-wide solutions.

55 -i~ APPENDIX A

Statement of Work 1. Background

USAID is pleased to report that GOH's new Minister of some laudable initiative Agriculture has shown in refocusing national attention problems. He has asked USAID on food storage to participate in this review, provide some technical recommendations specifically to as to what can be done funded (1971-1976) grain with the USAID­ silos, at Bon Repos near Port-au-Prince, smaller, equally unused, as well as silos in Cayes and Mirebalais, nationwide in the context of a price stabilization plan. 2. Proposed USAID response to GOH inquiry USAID wculd like the Kansas State University Food send some short-term and Feed Grain Institute to technical assistance, funded provement of Postharvest through the ST/AGR/APS Im­ Grain Systems project No. alternative approaches 0786, to formulate some for the Ministry. Specifically, supply agricultural we would like a food economist working with a Ministry counterpart to: A. Review food price trends, beginning with 1983 "Haiti: Rod Kite and Shirley Pryor's Food Policy Options" study, losses and assess the impact of on the uncertainties that crop they impose on producer behavior. B. Recommend desirable program and policy changes, if any, that larize prices such that investments might regu­ in food production would increase.

C. Review and advise on 1. A phased series of household production/consuraption how price shifts surveys to determine in one staple commodity affect supply responses for another. 2. Collection of quantitative data on used to postharvest crop losses that identify points at which conservation can be this connection, can feasibly be improved. reference should be made In to the literature on USAID's brate Pest Management Verte­ Project with Denver Wildlife. 3. Measurement of the degree to which small farmers could improve situation if staple their food prices were more stable. At the same time, USAID would like a food storage technologist working with a Ministry counterpart to: D. Assess the farm-level and national level food available in Haiti storage infrastructure now and make recommendations as to how it might be more effec­ tively utilized. E. Advise USAID and Ministry of Agriculture that could on additional technical options be sought to reduce postharvest food losses at all levels. F. Advise USAID and Ministry on potential use of unused silo facilities in the appropriate sort of price stabilization initiative. After the arrival of the study team, verbal discussions resentatives of USAID/Port-au-Prince were held with rep­ and GOH for the purpose of clarifying the

59 issue to be addressed under the Scope sions, of Work. As a result of these the Scope of Work was redefined dicus­ to encompass three main areas: sible price stabilization actions pos­ (including analysis of past review and evaluation of proposed price data), a data collection processes, and and physical utilization the economic of bulk storage facilities.

60 APPENDIX B Data Tables Table 1

Consumer Price Index1

Year 1970 = 100 1976 = 100 1960 76.0 1961 78.9 1962 78.4 1963 81.8 1964 89.4 1965 91.4 1966 98.9 1967 96.1 1968 97.3 1969 98.7 1970 100.0 1971 109.5 1972 113.1 1973 138.8 1974 159.5 1975 186.3 1976 199.4 1977 100.0 214.2 107.4 1978 208.0 1979 104.3 228.1 114.4 1980 269.2 1981 135.0 291.3 146.1 1982 315.7 1983 158.1 336.3 168.4

1CPI for Port-au-Prince weights: food 68.5%, clothing 19.53%, housing 11.91%.

Source: Zuyekas Departement du Plan World Bank

63 Table 2

Rice Imports

Tonnage Year Imported(MT)

1962 65.9 1963 31.0 1964 54.6 1965 30.0 1966 13.8 1967 16.8 1968 0.2 1969 9.0 1970 3.8 1971 47.9 1972 197.2 1973 2,515.4 1974 18.6 1975 1,106.0 1976 17,765.2 1977 741.6 1978 198.7 1979 115.8 1980 160.4 1981 898.3 1982 577.2 1983 7,500.0 Source: Delatour

64 Table 3

Corn Imports (MT)

All Corn 1 All Corn Year Corn Corn2 Products 3 Products 4 1962 4,300 1963 2,700 1964 2,200 1965 2,400 1966 4,400 1967 6,000 1968 7,000 1969 2,000 1970 11

1971 826 12 1972 8 5,465 2,000 8 1973 58 673 3,000 53 1974 19 6,645 1,000 19 1975 3,971 3,581 5,000 3,971 1976 94 20,000 1,000 94 1977 433 16,000 19,650 1978 123 5,200 1979 196 15,000 1980 1,191 5,000 1981 464 10,000 1982 298 5,000 1983

1 Source: Kite and Pryor 2 Source: Roe 3Source: USDA, Foreign Agricultural Circulars 4Source: FAO, Trade Yearbook 65 Table 4

Rice Production

Area Harvested Yield Year (1000 ha) Production (Kg/ha) (1000 MT) 1962 54 1,111 60 1963 55 1,164 64 1964 60 1,133 68 1965 65 1,108 72 1966 70 1,086 76 1967 70 1,100 77 1968 72 1,069 77 1969 75 1,107 83 1970 75 1,067 80 1971 75 1,080 81 1972 76 1,079 82 1973 76 1,079 82 1974 45 2,222 100 1975 48 2,292 110 1976 48 2,308 112 1977 40 2,260 90 1978 52 2,200 114 1979 54 2,261 122 1980 50 2,200 110 1981 50 1,900 95 1982 50 2,100 105 1983 50 1,900 95

Source: FAO, Production Yearbook

66 Table 5 Rice Production

Area Year Harvested Yield Production (1000 ha) (Kg/ha) (1000 MT) 1962 54 1,110 60 1963 55 1,160 64 1964 60 1,130 68 1965 65 1,110 7 1966 70 1,090 76 1967 70 1,100 77 1968 72 1,070 77 1969 75 1,110 83 1970 75 1,070 80 1971 75 1,070 80 1972 76 930 71 1973 76 1,050 80 1974 76 1,080 82 1975 76 1,080 82 1976 76 850 65 1977 76 1,110 85 1978 51 1,780 91 1979 51 1,870 95 1980 50 1,820 91 1981 50 1,900 95 1982 50 2,100 105 1983 50 1,900 95

Source: USDA

67 Table 6

Corn Production

Area Harvestea Year (1000 ha) Yield Production (Kg/ha) (1000 MT) 1962 300 763 229 1963 300 767 230 1964 300 773 232 1965 300 780 234 1966 305 766 234 1967 320 769 246 1968 300 733 220 1969 305 793 242 1970 310 774 240 1971 320 788 252 1972 330 779 257 1973 330 788 260 1974 238 1,050 250 1975 238 1,050 250 1976 239 1,048 250 1977 210 800 168 1978 260 962 250 1979 234 784 183 1980 245 714 175 1981 200 925 185 1982 185 919 170 1983 200 900 180 Source: FAO, Production Yearbook

68 Table 7

Corn Production

Area Harvested Year (1000 ha) Yield (Kg/ha) Production 1962 300 760 229

1963 300 770 230

1964 300 770 232

1965 300 780 234

1966 305 770 234 1967 320 770 246 1968 300 730 220 1969 305 790 242 1970 310 770 240 1971 320 790 252 1972 330 780 257 1973 320 780 250 1974 320 780 250 1975 320 780 250 1976 320 560 180 1977 320 780 250 1978 248 1,050 260 1979 250 980 245 1980 245 960 235 1981 250 1,180 295 1982 250 1,000 250 1983 250 1,000 250 Source: USDA

69 Table 8

Sorghum/Millet Production

HarvestedArea Year (1000 ha) Yield Production (Kg/ha) (1000 MT) 1962 260 706 184

1963 260 708 184 1964 260 712 185 1965 270 693 187 1966 270 694 187 1967 270 696 188 1968 270 700 189 1969" 280 746 209 1970 280 750 210 1971 280 754 211 1972 280 768 215 1973 280 771 216 1974 220 995 219 1975 220 1,000 220 1976 223 1,009 225 1977 152 730 111 1978 200 775 100 1979 157 787 123 1980 160 688 110 1981 163 675 110 1982 155 710 110 1983 155 742 115 Source: FAO, Production Yearbook

70 Table 9

Sorghum/Millet Prodiction

Area Harvested Yield Year (1000 ha) Production (Kg/ha) (1000 MT) 1962

1963

1964

1965

1966

1967

1968

1969

1970

1971 1972 216 670 144 1973 216 670 144 1974 216 670 144 1975 216 670 144 1976 216 670 144 1977 200 930 185 1978 210 900 189 1979 210 ^60 181 1980

1981

1982

1983

Source: USDA

71 Table 10

Agricultural Production (100 MT)

Millet/ Year Corn Sorghum Rice Beans 1977 168.3 110.8 1978 90.3 46.5 161.4 99.5 1979 113.5 45.6 183.3 123.3 1980 122.1 51.9 186.2 125.2 1981 124.1 52.6 179.2 120.8 1982 119.7 50,8 175.7 117.8 115.8 1983 170.9 50.3 106.6 113.4 46.7

Source: DARNDR estimates

Table 11

Planted Area by Department (Hectares)

Corn - Beans Department 1978 1979 Millet 97 -- 9 1978 1979 1978 1979 West 62,791 32,312 2,224 2,296 48,231 38,825 57,780 North 18,058 19,387 31,978 3,853 4,275 15,485 19,448 Northeast 17,461 16,947 1,995 1,341 3,013 2,784 11,688 12,806 Northwest 13,848 14,617 8,374 8,717 713 535 16,993 17,027 6,931 South 29,974 36,094 4,521 6,753 4,963 23,678 25,775 21,193 Southeast 19,568 20,706 1,497 31,948 Grande-Anse 862 14,836 17,218 9,356 9,924 30,231 33,865 1,736 2,483 Central 18,805 22,748 6,595 6,103 24,879 27,515 2,621 3,355 Artibonite 14,854 17,553 25,047 28,812 31,563 32,306 31,436 32,450 24,843 25,832 30,585 31,122 Haiti 248,373 233,749 51,614 51,003 189,413 197,232 165,856 156,698

Source: DARNDR estimates

72 Table 12

Agricultural Production, 1978

Corn Paddy Rice Beans Sorghum Transversal Area (ha) 74,438 29,680 21,554 Yield (MT/ha) 0.82 66,687 Production 4.00 0.60 (MT) 61,039 118,720 0.84 12,932 56,017 South Area (ha) 69,953 5,920 Yield (MT/ha) 23,536 38,051 0.66 1.50 Production (MT) 0.47 45,272 8,880 0.62 11,190 23,524 West Area (ha) 53,018 4,680 Yield (MT/ha) 28,460 41,902 0.99 1.40 Production (MT) 0.59 0.96 52,488 6,552 16,791 40,226 North Area (ha) 36,334 42,800 Yield (MT/ha) 16,128 10,055 0.67 1.80 Production (MT) 0.72 0.33 24,344 77,040 11,612 3,318 Country Area (ha) 233,743 83,080 89,678 Yield (MT/ha) 0.78 156,195 Production 2.54 0.59 (MT) 183,143 211,192 0.79 52,524 123,085

Source: Ministry of Plan

73 Table 13 Market Price Correlations for Corn1 Port-au-Prince and Other Urban Markets Market Time Series Correlation (R2 ) La Gra Data Cap-Haftien Cayes 1968-74 Gonalves 0.67 1968-74 0.69 Hinche 1970-74 Jacmel 0.72 1968-74 0.69 Jacmel 1969-74 Jeremie 0.57 1968.74 0.322 Port-de-Paix 1969-74 St Marc 0.48 1968-74 0.57 0.79

IHSI Data Cap-Haltien Cayes 1968-761968-76 Gonalves 0.77 1968-76 0.79 Jacmel 1968-76 J4remie 0.70 1968-76 0.58 Port-de-Paix 1968-76 St Marc 0.71 1968-76 0.64 0.86

DARNDR Data Cap-Haltien Cayes 1978-81 1978-81 0.81 Gonalves 1978-81 Hinche 0.93 1978-81 0.77 Jacmel 1978-81 Jeremie 0.97 1978-81 0.85 Port-de-Paix 1978-81 St Marc 0.69 1978-81 0.49 0.89

IPrice data source: Borsdorf 2 and Foster. Not significant at 5% level.

74 Table 14 Market Price Correlations for Superior Rice1 Port-au-Prince and Other Urban Markets Market Time Series Correlation (R2 ) La Gra Data Cap Haltien 1968-74 Cayes 1968-74 0.75 Gonalves 1968-74 0.79 Jacmel 1968-74 0.77 Jacmel 1969-74 0.56 Jeremie 1968-74 0.68 Port-de-Paix 1969-74 0.56 St Marc 1968-74 0.45 0.63

IHSI Data Cap Haltien 1968-76 Cayes 1968-76 0.62 Gonalves 1968-76 0.68 Jacmel 0.84 1968-76 J6remie 1968-76 0.77 Port-de-Paix 1968-76 0.79 St Marc 1968-76 0.83 0.80

DARNDR Data Cap Haltien 1978-81 Cayes 1978-81 0.85 Gonalves 1978-81 0.85 Jacmel 0.79 1978-81 0.79 Port-de-Paix 1978-81 St Marc 0.78 1978-81 0.74 0.93

1Price data source: Borsdorf and Foster.

75 Table 15 Market Price Correlations for Red Beans1 Port-au-Prince and Other Urban Markets Market Time Series Correlation (R2 ) La Gra Data Cap Haltien 1968-74 Cayes 1968-74 0.80 Gonalves 0.80 Hinche 1968-74 1970-74 0.82 Jacmel 0.65 Jacmel 1968-74 1969-74 0.59 Jere'mie 1968-74 0.57 Port-de-Paix 1969-74 0.77 St Marc 1968-74 0.63 0.63 0.83

IHSI Data Cap Haitien 1968-76 Cayes 0.86 Gonalves 1968-76 1968-76 0.82 Jacmel 1968-76 0.72 Jeremie 0.66 Port-de-Paix 1968-76 1968-76 0.81 St Marc 1968-76 0.68 0.80

DARNDR Data Cap Haitien 1977-81 Cayes 0.80 Gonalves 1978-81 1977-81 0.60 Hinche 1978-81 0.77 0.77 Jacmel 1977-81 Jer~mie 1977-81 0.65 0.56 Port-de-Paix 1977-81 St Marc 0.49 1977-81 0.042 0.86

1 Price data source: 2 Borsdorf and Foster. Not significant at 5% level.

76 Table 16 Market Price Correlations for Millet/Sorghuml Port-au-Prince and Other Urban Markets

Market Time Series Correlation (R2 ) La Gra Data

Cap Hattien 1968-74 0.65 Cayes 1968-74 Gonalves 0.46 1968-74 0.45 Hinche 1970-74 Jacmel 0.62 1968-74 Jacmel 0.70 1969-74 0.68 Jr~mie 1968-74 St 0.202 Marc 1968-74 0.41

IHSI Data

Cap Haltien 1968-76 0.64 Cayes 1968-76 0.79 Gonalves 1968-76 0.80 Jacmel 1968-76 0.52 J~remie 1968-76 0.52 Port-de-Paix 1968-76 0.73 St Marc 1968-76 0.54

1Price data source: 2 Borsdorf and Foster. Not significant at 5% level.

77 Table 17 Market Price Correlations for Corn I Port-au-Prince and Regional Markets Market Time Series Correlation (R2 ) La Gra Data Chantal Ducis 1968-74 1968-74 0.53 Croix des Bouquets 1968-74 0.58 Fond des Negres 1969-74 0.81 Miragogne 1969-73 0.76 Pont de l'Estere 1968-70 Pont Sonde 1972-74 0.51 0.69 0.41

IHSI Data Croix des Bouquets 1968-76 0.86

DARNDR Data Chansolme Chantal 1978-81 1978-79 0.63 Croix des Bouquets 1978-81 0.172 Fond des N~gres 1978-81 Lib 0.79 Leon 1978-811978-81 Limb6 0.782 1978-81 0.74 Pont de 1'Est re 1978-81 0.49 Pont Sonde 1978-81 Thiotte 0.87 1978-81 0.85 0.85

IPrice data source: 2 Borsdorf and Foster. Not significant at 5% level.

78 Table 18 Market Price Correlations for Superior Rice I Port-au-Prince and Regional Markets Market Time Series Correlation (R2 ) La Gra Data Croix des Bouquets 1968-74 Miragoane 1969-73 0.89 Pont de l'Est~re 1968-70 0.222 Pont Sond6 1972-74 0.81 0.172

IHSI Data Croix des Bouquets 1968-76 0.89

DARNDR Data Pont 1 de l 'Est~re 1978-81 Pont Sonde 1978-81 0.80 0.41

1Price data source: 2 Borsdorf and Foster. Not significant at 5% level.

79 Table 19 Market Price Correlations for Red Beansl Port-au-Prince and Regional Markets

Market Time Series Crrelation (R2 ) La Gra Data Chantal Ducis 1968-74 1968-74 0.66 Croix des Bouquets 1968-74 0.70 Fond des N~gres 1969-74 0.88 Miragoane 1969-73 0.75 Pont de l'Estere 1968-70 Pont Sonde 0.62 1972-74 0.292

IHSI Data Croix des Bouquets 1968-76 0.89

DARNDR Data Chansolme 1978-81 Chantal 1978-79 Croix 0.53 des Bouquets 1978-81 Fond 0.69 des N~gres 1978-81 Leon 1978-81 0.322 0.32 Limb6 1978-81 0.59 Pont de l'Est~re 1978-81 0.78 Pont Sonde 1978-81 Thiotte 0.71 1978-81 0.60 0.72

1 Price data source: 2 Borsdorf and Foster. Not significant at 5% level.

80 Table 20 Market Price Correlations for Millet/Sorghum Port-au-Prince and Regional Markets

Market Time Series Correlation (R2 ) La Gra Data

Chantal 1968-74 Ducis 0.68 1968-74 0.60 Croix des Bouquets 1968-,74 0.62 Fond des N~gres 1969-74 Miragogne 0.65 1969-73 Pont 0.78 de l'Estere 1968-70 Pont 0.75 Sonde 1972-74 0.64

IHSI Data

Croix des Bouquets 1968-76 0.64

1Price data source: Borsdorf and Foster.

81 Table 21

Market Price Correlations for Corn I Port-au-Prince and Semi-rural Markets

Market Time Series Correlation (R2 ) La Gra Data Camp Perrin 1968-74 Cerca Carvajal 1973-74 0.61 Croix Fer 0.63 1972-74 0.332 D~sarmes 1972-74 0.59 0.80

IHSI Data Petit Goave 1968-76 0.80

DARNDR Data Camp Coq 1978-81 Cerca Carvajal 1978-81 0.63 Croix Fer 1978-81 0.79 Darbonne 1978-81 0.84 D~sarmes 1978-81 0.73 Fort des 0.73 Pins 1978-79 Gaillard 1978-81 0.86 Marfranc 1978-81 0.162 0.61

IPrice data source: Borsdorf 2 Not and Foster. significant at 5% level.

82 Table 22 Market Price Correlations for Red Beansl Port-au-Prince and Semi-rural Markets

Market Time Series Correlation (R2 ) La Gra Data Camp Perrin 1968-74 D~sarmes 0.80 1972-74 0.62

IHSI Data Petit Goave 1968-76 0.89

DARNDR Data Camp Coq 1978-81 Cerca 0.64 Carvajal 1978-81 Croix 0.58 Fer 1978-81 Darbonne 0.45 1978-81 Desarmes 0.64 1978-81 Foret 0.52 des Pins 1978-79 Marfranc 0.55 1978-81 0.59

IPrice data source: Borsdorf and Foster.

83 Table 23 Market Price Correlations for Millet/Sorghuml Port-au-Prince and Semi-rural Markets

Market Time Series Correlation (R2 ) La Gra Data Camp Perrin Cerca Carvajal 1968-74 1973-74 0.53 Croix Fer 1972-74 Desarmes 0.45 1972-74 0.54 0.85

IHSI Data

Petit Go~ve 1968-76 0.74

1Price data source: Borsdorf and Foster.

84 Table 24 Market Price Correlations between Corn and Millet/Sorghuml Port-au-Prince

Data Series C;orrelation (R2) La Gra 1968-74 0.88 IHSI 1968-76 0.88 DARNDR 1978-81 0.58

1Price data source: Borsdorf and Foster.

85 Table 25

Proxy Price Series Average Monthly Prices Port-au-Prince

Courdes per Marmite

Year/ ole Month Red All Corn Millet Superior Beans Rice Rice

1968Jan 1.00 1.40 Feb 3.00 4.20 3.50 0.92 1.25 3.18 Mar 3.50 3.50 0.90 1.25 3.30 Apr 3.65 3.50 0.90 1.26 3.33 May 4.55 3.50 1.52 0.60 3.79 5.25 Jun 1.85 4.28 2.00 5.50 4.40 4.50 Jul 1.72 Aug 1.82 4.31 4.20 1.50 1.78 4.31 Sep 3.66 4.20 1.30 1.72 4.00 3.43 4.65 4.18 ov 1.05 1.60 Dee 0.89 3.25 4.90 4.25 1.50 2.94 5.50 3.50 0.88 1.50 3.00 5.00 3.56 1969Jan 0.96 1.60 Feb 3.25 4.30 3.68 1.06 1.60 3.33 4.09 Mar 3.50 1.10 1.60 3.25 Apr 4.09 3.50 0.90 1.40 3.75 May 4.09 3.50 0.90 1.40 3.75 Jun 4.09 3.50 0.90 1.40 3.38 Jul 4.09 3.50 0.90 1.40 3.00 Aug 4.50 3.50 0.87 1.40 Sep 2.94 4.25 3.41 0.79 1.22 3.50 Oct 0.80 4.10 3.42 1.35 3.75 3.55 3.75 Nov 0.90 1.25 Dec 3.12 3.90 3.68 0.81 1.60 3.37 4.25 3.25 1970 Jan 0.73 Feb 1.55 3.62 4.00 0.87 1.50 3.31 3.25 Mar 4.31 3.50 0.90 1.45 3.25 Apr 4.31 3.50 0.95 1.30 May 3.37 5.00 3.75 1.10 1.40 3.68 Jun 1.17 5.00 4.06 Jul 1.50 4.81 5.00 1.05 1.50 4.37 Aug 3.43 5.07 4.12 1.00 1.50 3.25 Sep 1.10 4.93 4.00 Oct 1.50 3t25 5.00 1.25 1.50 4.37 Nov 3.20 6.31 0.95 1.40 5.12 Dec 3.00 5.69 0.90 1.40 4.62 3.00 4.93 4.00

86 Table 25 (cont.)

Gourdes per Marmite

Year/ Whole Month Red All Superior Corn Millet Beans Rice Rice 1971 Jan 0.90 1.140 3.15 4.29 Feb 0.90 1.140 5.50 Mar 3.00 5.00 5.50 0.95 1.145 Apr 3.20 4.48 5.70 1.15 1.60 May 4.50 5.47 6.15 1.15 1.55 Jun 4.05 5.25 5.90 1.25 1.50 Jul 3.65 4.89 5.50 1.15 1.50 3.60 4.87 Aug 5.80 1.20 1.65 3.65 4.97 5.90 Sep 1.15 1.60 4.05 5.98 5.80 Oct 1.15 1.60 Nov 4.50 5.00 5.62 1.12 1.55 Dec 4.37 4.05 4.87 1.06 1.56 3.90 4.75 4.10 1972 Jan 1.00 1.40 Feb 4.00 5.77 5.25 0.90 1.25 4.00 5.63 Mar 5.12 1.02 1.30 4.25 5.77 Apr 5.25 1.25 1.50 4.06 5.25 May 5.50 1.25 1.50 4.00 5.51 5.50 Jun 1.42 1.55 Jul 4.25 5.50 5.37 1.25 1.50 Aug 3.56 5.35 1.30 1.50 4.30 Sep 4.05 5.50 4.25 1.52 1.70 Oct 5.25 5.50 4.75 1.60 1.70 Nov 5.50 6.04 5.50 1.50 1.70 Dec 4.41 6.04 5.50 1.70 1.80 4.12 6.73 6.12 1973 Jan 2.00 1.95 Feb 5.12 5.60 7.00 2.25 2.06 5.31 5.70 Mar 6.62 2.53 2.35 5.56 5.90 7.31 Apr 2.P- 2.62 May 5.50 6.30 9.56 2.73 2.56 Jun 4.68 6.59 9.25 3.12 2.72 6.50 7.51 Jul 9.25 3.25 2.75 6.75 7.29 Aug 9.00 3.25 2.75 5.81 7.29 Sep 9.00 3.00 2.58 6.25 7.03 Oct 8.68 2.40 2.25 Nov 5.87 6.42 8.37 2.16 2.32 Dec 4.94 6.50 6.81 2.17 2.32 4.81 6.50 6.25

87 Table 25 (cont.)

Gourdes per Marmite

Year/ Whole Month Red All Superior Corn Millet Beans Rice Rice 1974 Jan 2.37 2.48 Feb 5.75 6.09 6.50 2.45 2.53 Mar 6.31 6.45 6.88 2.64 2.60 Apr 6.69 6.91 7.38 2.75 2.70 May 6.81 7.00 7.50 2.64 2.50 Jun 7.00 7.12 7.50 2.58 2.50 Jul 6.63 7.33 7.63 2.50 2.45 Aug 5.70 7.50 7.63 2.58 2.50 Sep 5.40 7.31 8.00 2.18 2.43 Oct 5.40 7.55 7.88 2.40 2.50 Nov 6.45 7.70 8.40 2.29 2.43 Dec 6.50 7.41 8.50 2.48 2.50 6.13 7.87 8.25 1975 Jan 2.25 2.96 Feb 5.36 7.88 7.00 2.60 2.73 Mar 5.50 7.88 8.00 3.00 3.15 7.50 9.15 8.00 Apr 3.76 3.50 May 8.33 10.27 11.75 4.56 4.56 Jun 10.59 12.28 12.87 5.20 3.39 Jul 11.00 10.34 12.62 4.41 4.90 7.65 9.56 11.67 Aug 4.41 4.53 Sep 6.416 11.12 13.57 3.80 4.84 Oct 6.32 11.96 14.60 2.88 4.24 Nov 6.17 11.12 13.57 2.38 4.51 6.50 10.50 Dec 12.81 2.46 4.20 7.40 11.54 11.81 1976 Jan 2.92 3.85 Feb 8.45 11.60 9.75 2.58 4.36 Mar 9.66 12.15 11.50 2.81 4.15 Apr 9.75 11.70 11.62 2.80 4.10 May 9.50 11.20 10.00 280 3.00 Jun 9.03 10.50 10.50 2.87 2.95 Jul 6.85 8.80 10.25 2.20 2.75 Aug 6.20 7.60 8.07 2.31 2.56 Sep 5.65 10.90 7.62 2.33 2.75 Oct 5.80 10.65 8.10 2.50 3.15 Nov 6.30 8.70 9.00 2.65 3.85 Dec 7.20 9.75 9.75 3.30 4.15 6.10 9.60 9.75

88 Table 25 (cont.)

Gourdes per Marmite

Year/ Whole Month Red All Superior Corn Millet Beans Rice Rice 1977 Jan 3.23 3.70 Feb 8.70 11.45 11.88 3.05 3.60 Mar 8.55 11.35 11.78 3.20 3.95 Apr 8.95 11.05 11.47 3.94 4.80 May 11.36 11.25 11.69 4.01 4.95 Jun 14.40 11.50 11.94 3.94 4.75 Jul 9.90 8.00 8.30 3.94 4.75 Aug 6.05 10.05 10.43 3.94 4.75 Sep 5.95 10.00 10.38 3.23 4.15 Oct 6.50 10.00 10.38 3.16 3.56 Nov 6.55 10.00 10.38 2.75 3.30 Dec 6.30 9.05 9.39 2.75 3.15 5.95 8.10 8.41 1978 Jan 2.49 3.70 Feb 7.15 8.50 9.'8 2.49 2.90 Mar 8.35 8.60 9.16 2.57 3.05 Apr 8.05 8.75 9.38 2.49 3.15 May 8.05 9.20 9.54 2.53 2.90 Jun 8.40 9.10 8.98 2.53 3.50 Jul 7.60 8.25 8.78 2.31 3.45 Aug 7.45 9.85 10.33 2.01 3.05 Sep 7.55 10.05 10.42 2.02 3.20 Oct 7.95 11.45 11.33 2.15 3.25 Nov 7.60 10.00 11.30 1.90 3.50 Dec 8.55 9.65 9.55 1.95 3.60 7.75 10.00 9.75 1979 Jan 2.05 3.35 Feb 7.40 9.95 9.72 2.45 3.40 Mar 8.55 10.00 10.20 2.65 3.60 Apr 9.80 11.10 11.60 3.15 3.95 May 11.55 12.65 13.25 3.30 4.55 Jun 10.15 12.25 12.80 4.40 4.85 Jul 9.30 12.00 11.60 3.05 5.20 Aug 8.00 12.45 12.70 2.75 5.20 Sep 8.20 12.85 12.80 2.80 5.40 Oct 8.75 13.40 13.00 2.50 4.46 Nov 10.84 11.46 11.35 2.60 4.68 Dec 11.56 12.52 12.40 2.40 3.77 10.75 13.33 13.20

89 Table 25 (cont.)

Gourdes per Marmite

Year/ Whole Month Red All Superior Corn Millet Beans Rice Rice

1980 Jan 3.45 5.50 Feb 12.60 13.80 3.75 5.50 12.45 13.90 14.50 Mar 14.25 3.70 5.25 Apr 12.25 15.56 15.10 4.00 5.50 May 12.50 14.25 14.40 4.10 5.40 12.50 Jun 4.60 12.36 12.30 5.55 13.50 12.85 Jul 4.41 12.05 Aug 5.50 12.15 13.25 4.55 5.85 13.25 Sep 12.15 13.00 12.98 4.35 5.85 Oct 12.15 14.10 12.84 3.55 6.1o Nov 12.65 14.40 12.70 4.15 6.15 Dec 12.05 14.20 13.25 4.50 6.25 12.50 14.60 14.15 1981 Jan 4.80 6.19 12.05 17.21 Feb 5.50 6.23 12.27 17.97 16.80 Mar 18.25 6.00 6.68 Apr 12.57 21.64 22.15 6.10 5.15 May 9.05 17.40 5.70 5.15 20.95 Jun 8.50 17.10 16.75 5.00 5.20 Jul 8.00 14.20 16.50 4.60 5.60 Aug 8.70 14.50 3.35 5.40 17.50 Sep 8.45 14.55 17.25 3.00 5.35 Oct 8.95 14.35 4.03 5.00 16.25 Nov 9.25 12.45 16.00 3.23 4.75 Dec 10.60 13.05 16.70 3.30 4.80 10.15 12.75 16.55 1982 Jan 3.82 3.96 Feb 9.26 11.81 13.82 3.99 3.78 9.09 11.75 Mar 13.75 3.90 3.48 8.91 12.44 Apr 14.55 3.86 3.12 9.14 12.81 May 14.99 3.90 3.30 9.66 13.00 15.21 Jun 4.11 Jul 3.54 10.18 12.75 4.11 3.66 14.92 Aug 9.03 12.25 4.11 3.66 8.91 14.33 Sep 12.38 14.48 4.24 3.30 Oct 8.80 12.31 14.96 4.07 4.08 Nov 8.86 12.25 3.94 3.48 15.44 Dec 8.74 12.38 3.74 3.48 15.92 8.91 12.63 16.40

90 Table 25 (cont.)

Gourdes per Marmite

Year/ Whole Month Red All Superior Corn Millet Beans Rice Rice 1983 Jan 4.86 3.48 Feb 9.09 12.81 16.63 5.08 3.54 9.14 12.94 16.27 Mar 5.08 3.66 10.18 13.50 17.42 Apr 5.41 3.90 May 11.50 14.31 19.00 5.83 3.90 Jun 11.96 14.88 18.63 5.75 3.84 Jul 10.70 15.06 16.50 5.87 4.98 Aug 9.55 14.25 16.53 5.11 3.72 Sep 10.18 14.88 17.27 5.18 4.68 Oct 10.64 15.31 17.59 5.79 5.40 Nov 11.50 15.63 15.82 5.67 4.20 Dec 11.85 15.31 15.92 5.33 4.14 11.21 15.19 15.19 19814 Jan 5.33 4.14 Feb 12.08 14.94 16.04 5.20 4.26 12.19 14.94 16.99 Mar 5.37 4.32 12.13 15.00 17.12 Apr 5.54 4.50 11.79 14.44 15.614 May 5.58 4.80 11.85 14.69 16.29 Jun 5.45 5.10 11.85 14.75 16.36 Jul 5.03 5.10 11.33 14.75 16.36 Aug 5.03 5.314 11.33 14.75 16.36 Sep 5.45 5.34 Oct 12.19 14.88 16.50 5.50 5.34 Nov 12.13 15.31 16.98 5.25 5.40 Dec 12.09 15.13 16.78 5.25 5.34 12.13 15.19 16.85

Source: Borsdorf and Foster

91 APPENDIX C

Description DARNDR Bulk Storage Facility at Bon Repos

PeI o S Page Located 11 km from Port-au-Prince, Bon Repos is the storage facility constructed site of a bulk grain by the GOH in an attempt grain marketing to become involved in activities. Equipment for in 1977. the plant was ordered from Haytian Tractor and Equipment the US Company, S.A. of Port-au-Prince hired as a procurement agent was for the silo equipment, and contracted to a local firm. construction was However, construction was completion of the operation halted in 1978 before and the grain drying and remained unfinished to storage facility has this day. The study team documentation was unable to locate the which described how this marketing facility would fit into system in Haiti, nor the grain could any explanaticn be for the discontinuation found as to the reason of construction on the facility. The facility calls for five semi-circular modules, each storage bins and one combination consisting of four drying-storage bin. The the silo modules is shown schematic layout of in Figure 1. All bins have the following dimensions: flat bottoms and are of 28 ft diameter, 16 and perforated ft eave height, 300 roof pitch, floor 14 in above bin bottom. the Even though GOH documents capacity of each bin to be show 240 MT, a more realistic storage on U.S.Grade I corn is 200 capacity based MT (see nominal grain storage page 99). The estimated capacity calculation, total storage capacity of 5,000 the Bon Repos facility is MT for shelled corn or sorghum, or 4,100 MT for rough rice. Each bin is equipped with an unloading auger to the main and a sweep auger. In addition unloading sump at the center other of the bin floor, there unloading sumps which allow are three grain to be taken out uniformly locations. Each bin is from radial also equipped with a suction-type consisting of aeration system a 12-in diameter axial flow fan driven by a 3/4 HP motor. The drying equipment consists of a Sukup Heatway 2 fuel oil fuel oil burner (No. 1 or or kerosene). The heat output No. an estimated of the burner could be varied 410,000 Btu/hr to 1,040,000 from two nozzles Btu/hr by varying the size on each burner. The drying of the air is supplied by a centrifugal fan. The drying Sukup Speedway unit is provided with a heat flue gases from passing exchanger to prevent through the grain. The burner has a sail switch which turns it off if the fan stops. To dry grain uniformly, each drying bin Stirway is equipped with a 4 -auger II stirring system. To spread Sukup in the fine materials and grain the drying bin so as to facilitate uniformly uniform drying, each drying equipped with bin is also a Sukup Spreadway grain spreader. Each module is equipped with a rotating and cylindrical grain scalper a portable transport auger (cleaner) mounted on wheels so as from the ground level into to transfer the grain the top of a bin. The dry to any of the storage grain can be transferred bins by using either the auger transport auger or the vertical installed beside the drying bin wall and a roof auger. An evaluation of the existing status of the plant is detailed as follows: Site

The land has not been leveled and graded, and it lacks proper drainage. The grain receiving and cleaning area has not been paved.

95 'K o-i "- r 5 S 4 SN U 2

/ U

1- COMBINATION DRYING-STORAGE BIN 2,3,4,5- STORAGE BINS C- CLEANER R- ROOF AUGER S-SWEEP AUGER T- TRANSPORT AUGER U- UNLOADING AUGER FIG. f LAYOUT OF BON REPOS SILO MODULE AND MAJOR EQUIPMENT

96 Bin Foundation The foundation plinth level is low (only about In general, the foundation 6 in above the ground appears to be adequate. level). of the foundation will However, the real strength not be tested until the bin has been loaded with grain. Bins The anchoring of the bins by using a hold-down flat use of a flat plate plate seems erratic. The to press the bin down onto method of anchoring. the foundation is not a Some anchor bolts protruded good others. This means above the nuts more that either bolts of different than anchor bolts were not uniformly lengths were used, or all embedded in the concrete. the bottom tier and the foundation The sealing between appear seems adequate. The bin to be reasonably well-erected. walls and the roof sheets seemed However, the bolts on the to be over-tightened, with side wall washers in resulting damage to the many places. Mastic seals neoprene were no were properly used in obvious water leakage problems joints. There inspection in the few bins inspected, was done on a dry day. however the material The galvanization of in general has not weathered. wall and roofing The material is a few bin walls bright and shiny. On there are some rusty areas. The perforated floor seemed somewhat lacking to determine whether in strength. It was not proper and sufficient number possible used or not. In general, of steel floor supports were the floor seems adequate. unloading sumps and augers The sweep augers and the have been installed, however and close the mechanism to open the sump gates does not work. The walk-in door and the door steps were installed some doors were inconveniently at the proper height, but well. and arbitrarily located. The oblong manhole covers The doors close on the bin roofs are mechanism to open and close badly corroded. The the manhole covers from factorily. The roof safety the ground works satis­ ring appears to be properly stops at the junction installed. The bird of the roof sheet rib and places. the bin wall are in the The ladders are not proper located at convenient places. Augers The unloading auger and tube assembly in a few bins walk-in door which is was located underneath an inconvenient and unaesthetic the drive on the vertical location. The motor auger for the drying bin level. This undesirable is located below the ground location could have been floor level for the foundation. avoided by using a higher Proper transport augers screw have been used. The flightings are rusty, however.

Scalper Cleaners The cleaners are in reasonably good condition.

Motors and Drives All motors and drives, especially the outdoor ones, are badly corroded. A few motor mounts do not have motors on them.

97 Bearings and Shafts All bearings and shafts are heavily corroded, especially the outdoor belonging to ones augers, cleaners, fans, and burners.

Power Cables

The existing location of the power cable in be inconvenient front of the working area may and hazardous, especially for installation the trucking operation. The of the main power cables and power lines to each motor and control unit is incomplete.

Power Transformers Three pole-mounted transformers have been installed for single (115 V, 60 Hz) and three-phase (208/230 V, 60 Hz) power, but their condition is unknown. Motor Controllers

Many motor controllers have not been installed.

Fuel Oil Tank and Piping

The fuel oil tank and the necessary pipings have not been installed.

Other Observations

It is not known what provisions were originally grain since planned for weighing the there is no truck scale. Therefore, incoming it may be assumed that both and outgoing grain was supposed to be in bags. However, does not seem to include the plan provisions for a bagging operation, scales were not and platform found. A bagging oper tion in the plant will be slow and difficult. Security of the grain being stored in the bin is of utmost importance. Consideration against grain pilferage seems somewhat lacking. There is a lack of adequate truck access to each module. There were many unassembled parts of bins and roof sheets, equipment such as side wall perforated floors, sumps, and auger parts, motors and controls, pulleys, belts, bolts and nuts at the Bon Repos plant. These erection spares because manufacturing can not be companies provide erection only on small items such spares as bolts, nuts, washers, seals, The type and and caulking agents. quantity of the unassembled parts that remaining at this plant suggest additional bins were to be erected either in this location or elsewhere.

98 Nominal Grain Storage Capacity Calculationl

Bin diameter, D = 28 ft Eave height from the bin bottom, H, = 16 ft Perforated floor level from the bin bottom, X, = 14 in Roof pitch, G = 300 Angle of repose,o, varies from grain to grain and with moisture. Opening Let = 350 size on the bin top, k = 2 ft L, = (D/2 - k/2) . tanG = (14 - 1) (tan 30) = 7.51 ft H3 = HI + LI = 16 + 7.51 = 23.51 ft Since the roof pitch is somewhat lower' than th3 angle not be filled of repose, the bin can to the eave height (point E). The maximum grain level is point A.

L = D/2 2 . tanO(= 14 . tan 350 = 9.80 ft H = 2 H3 - L2 = 23.51 9.80 = 13.71 ft H = H2 - X = 13.71 - 1.17 = 12.511 ft Considering the conical portion of the stored product, the effective cylinder height, h, of the storage = H + 1/3 . L2 h = 12.54 + (1/3)(9.80) h = 15.81 ft

The volume of the stored product * (?r/4)(D 2 ).(h) * (ir/4)(28)2 (15.81) * 9735 ft 3 = 7788 bu On the basis of U.S. Grade I yellow dent corn (56 lb/bu), the storage capacity of the bin = 7,788 x 56 2, - 198MT

Considering some grain compaction, the nominal storage be taken as capacity of a bin may 200 MT of corn. The test weight as corn, for sorghum is about the same thus the storage capacity estimate rough for sorghum is 200 MT. For rice, the capacity will be reduced by a factor equivalent to the of the test weight of rough rice ratio to the test weight of corn. This storage capacity will give a for rough rice of about 164 MT. The total storage capacity of the facility is 5,000 MT of yellow dent corn or shelled sorghum, or 4,100 MT of rough rice.

1See Figure 2 for dimensions and notations.

99 FIG.2- BIN DIMENSIONS AND NOTATIONS

100 APPENDIX D Rehabilitation of DARNDR Bulk Storage Facility at Bon Repos Becalse of the discontinuation of construction facility on the grain storage and at Bon Repos before the drying project was completed, the facility been used. However, due to has never the fact that it has stood unused completed from 1978 to and partially the present, deterioration of ment has occurred. the structures and equip­ If the GOH desires to bring this status, considerable facility up to operational rehabilitation of the facility following will be necessary. The specific actions should be undertaken in order to accomplish this task.

A complete inventory should be taken of all equipment what is available and parts to find out on-site and in good condition, imported and what will have to be from the U.S. It may be possible equipment, to repair some of the corroded while other items will need to be replaced. Mechanical and trical parts and equipment should elec­ be cleaned and relubricated after replacement. Any usable parts repair or and equipment that have been painted salvaged should be if they show signs of weathering. The power lines should be re-routed to Undergrou,'d the back side of the silo modules. lines are not advisable, therefore Both overhead lines should be used. single and three-phase power should be obtained by the use of appropriate step-down transformers. All power lines must be properly supported. All motors, magnetic motor controllers, circuit breakers, be supplied with an appropriate and switches must power source. Security installed at the site. lights should be Electricity to operate construction ment during the tools and equip­ rehabilitation work must also beginning be made available prior to the of such work. The wire size and specifications, voltage drop siderations, all supports, all con­ electrical equipment, and safety conform to the U.S. National devices must Electric Code (NEC) 1 8 , tional Fire Protection 9 4 published by the Na- Association. All outdoor NEC equipment as defined by 19 84 must be approved for outdoor the comply use. All specifications must with any local electrical codes. also The contractor must prepare necessary wiring drawings. The installation of a truck scale is not advised, age, and handling since transportation, stor­ are not geared to a bulk mostly system. Incoming grain will in bags, and outgoing grain be will need to be rebagged, therefore tion and maintenance of a truck opera­ scale will be difficult. The scales and the use of platform sewing of bags by hand are the procedures recommended for this facility. Once the facilities are operational, 200 MT of wet at a rate of grain should be available 50 MT per day, in order to test the capacities of the dryer, storage bins, and augers.

Qualifications of the Contractor

The contractor should be a company or person(s) from the of experience in electrical U.S. with ten years and mechanical erection and operation of grain drying and storage of the type equipment used at this facility. should have a good The contractor financial standing. The qualifications should be of the contractor evaluated by a qualified consulting engineer. Getting a U.S. company interested in a capable small foreign project like this substantial amount of money. will involve a However, the success of the project obtaining depends on a capable U.S. company as contractor.

103 Consulting Engineer The consulting engineer will provide periodic short-term to ensure the appropriateness, technical assistance compliance of contractual progress of the contractor's obligations, and work. The consulting experience in the design, engineer must have broad construction, and supervision age projects. It is estimated of similar grain stor­ that two TDY trips to Haiti The first trip, of a one-week will be required. duration, should coincide commencement of work with the contractor's on the construction site. weeks, should The second trip, to last start immediately prior three to the end of the construction continue through and the end of the test run.

Cost of Rehabilitation

The exact cost of rehabilitation can not be itemized survey estimated without a complete of all storage units and equipment. However, a estimate for rehabilitation preliminary is $500,00C to $700,000.

104 APPENDIX E

References Albert, Roland M., Assessment of Requirements for Designing and Implementing a Pilot Area Frame Survey in Haiti, SRS, USDA, USAID/Port-au-Prince, February, 1985. Borsdorf, Roe and Kathy Foster, A Compilation of Market Price Data for Cereals and Beans, Haiti, Food and Feed Grain Institute, May, 1985. Coffey, W.J. et al., Social Institutional Profile of the Les Cayes :lain Basin: Towards a Coordinated Rural Regional Development Strategy, USAID/Port-au-Prince, January, 1984. Cuevas, Miguel and David J. McGill, Assessments of Requirements for Designing and Implementing a National Agricultural Survey in Haiti, U.S. Bureau of the Census, January, 1985. Delatour, Leslie and Paul Duret7 Rice in Haiti, USAID/Port-au-Prince, February, 1984. Denver Wildlife Research Center, Vertebrate Damage Agriculture Control Research in Annual Report 1983, U.S. Fish and Wildlife Service, 1984. Departement de l'Agriculture, des Ressources Naturelles et du Developpement Rural (DARNDR), unpublished data. Desrouillres, J.B., La Conservation Traditionnelle Pertes du Mas en Milieu Rural: et Agents Responsables, Universite d'Etat d'Ha ti, July, 1981. Duplan, V. and J. La Gra, Commercial Activities in Rural Haiti: A Community- Centered Approach, IICA, October, 1974. Duplan, V. and J. La Gra, Les Marches Publics d'Halti, IICA, 1975. Duplan, V. and J. La Gra, Transport des Produits Agricoles vers Port-au-Prince, !ICA, June, 1975. Dupont, Ernest and Richard Swanson, Implementing a Program: Farming System Research A Case Study in Haiti, USAID/Fnr-tau-Prince, September, 1984. Elliot Berg Associates, Haiti: Agricultural Policy Reforms, PL480 Title Project Paper, Background III Document, April, 1984. FAO, Plan d'Opration - Projet CCP/HAI/010/SW1, Centres C4 realiers pour' la Prevention desPertes Locaux Apres Recolt, Programme de Cooperation FAO/Gouvernement.

FAO, Production Yearbook, 1962-83, Rome.

FAO, Trade Yearbook, various issues, Rome.

FAO, unpublished data. Food Sector Strategy Team, Food and Agriculture Sector Strategy for Haiti, USAID/Port-au-Prince, February, 1982.

p.107 Institut Haltien de Statistique et d'Informatique, Bulletin Statistique

Annuel, various issues. Institut Halcien de Statistique et d'Informatique, unpublished data. Jaffe, Joann M., Structural Monopoly and Independent Stagnation Household Production in the Agrarian Economy of Haiti, unpublished MS thesis, Cornell University, 1983.

Johnson, J.L. and J. La Gra, Le Syst~me Interne de Commercialisation Agricole d'Halti: Une Analyse de Frix, IICA, November, 1975. Kite, R.C. and S.A. Pryor, Haiti: Food Policy Options, USAID/Port-au-Prince, November, 1983. La Gra, J., W. Charleston and G. Fanfan, Agricultural Product Prices in Haitian Marketplaces, IICA, March, 1975. Locher, Uli, The Internal Marketing Sstem of Agricultural Port-au-Prince Products in Haiti, IICA, 1974. Lundhal, M., "Price Series Correlation and Market Evidence Integration, Some from Haiti", in The Haitian Economy Man, Land and Labor, Croom Helm, London. Murray, G. and M. Alvarez, The Marketing of Beans in Haiti: an Exploratory Study, IICA, June, 1973.

Muskin, J.A., An Analysis of Grain Prices and Policy Options USAID/Port-au-Prince, in Haiti, August, 1983. Organisme de Developpement de la Vall~e de l'Artibonite, Annuel Rapport 1983/84, DARNDR, July, 1984. Pinchinat, A. and J. Alix, Situation de la Riziculture enHati, ODVA/IDR Project, 1981. Roe, Terry L., An Economic Evaluation of the Haitian Agricultural Marketing System, USAID/Port-au-Prince, February, 1978. Service de Statistiques Agricoles, Evolution de Certains Produits au Cours des Annes 1977-1978-1979, DARNDR, July, 1980. Service National de Commercialisation Agricole (SENACA), des Bulletin des Prix Produits Agricoles sur les March6s de la R6publque, various issues. Service National de Commercialisat4on Agricole (SENACA), Coat Estimatif Ple Mise en Marche des Silos de Bon Repos, DARNDR, 1982. Service National de Commercialisation Agricole Commercialisation 6 (SENACA), Projet de C ralire (Stock de Rgulation et Stock de Rserve) DARNDR, December, 1978.

108 Service National de Commercialisation Agricole (SENACA), unpublished data. Smucker, Glenn R., Supplies of Credit Among Haitian Peasants, Development Alternatives, Inc., February, 1983. Sorenson, Orlo, and Do Sup Chung, Grain Marketing and Market System Development, Haiti, Food and Feed Grain Institute, December, 1973. USAID, Project Paper Supplement No,. 1, Agricultural Development Support II (521-0092), August, 1982. USDA, Agricultural Statistics, various issues, 1969-83. USDA, Foreign Agricultural Circular, various issues, 1961-80.

World Bank, Economic Note for the Haitian Subgroup, November, 1983. Zuvekas, Clarence Jr., Agricultural Development in Haiti, an Assessment of Sector Problems, Policies, and Prospects under Conditions of Severe Soil Erosion, USAID/Port-au-Prince, May, 1978.

109