PP-AA Z-2t1? 90

The

P.L.-480 Title III Program in

An Evaluation of the First 16 Months

August 1978 - November 1979

USAID/Bangladesh December 13, 1979 TABLE OF CONTENTS

I. Introduction

II. Summary Findings and Recommendations

III. Production, Prices and Public Food Distribution

A. Production Trends

B. Crop Yields

(1) Current Situation and Trends

(2) Yield Factors

(a) Price Incentives

(b) Fertilizer

(c) Irrigation

(d) Research and Extension

(3) Constraints

(a) Agricultural Credit

(b) Land Tenure

(c) Fertilizer Distribution

C. Foodgrain Supply and Per Capita Consumption

D. Prices and Public Food Distribution

(1) Elasticity of Demand

(2) Trend Analysis

(3) Seasonal Analysis IV. The P.L. 480 Title III Agreement

A. Background

B. Terms of the Agreement

C. Price Incentives

(1) Objectives

(2) Performance

D. Price Constraints - OMS

(1) Objectives

(2) Performance

(a) Program Operations

(b) Impact on Coarse Rice Prices

(3) OMS Objectives Reconsidered

(a) Rice Price Constraint

(b) Promotion of Private Sector Activity; The OMS Price

(c) Food Security; OMS When Needed

E. Local Currency Proceeds; Loan Forgiveness

F. Self-Help Measures

G. Recommendations

(1) For the Amendment to the Current Agreement

(2) For the Next P.L. 480 Agreement

Annex A - The Impact of Procurement Annex B - Estimated Reserve and Storage Requirements Annex C - Food Distribution by Income Status; Urban vs. Rural Distribution Annex D - The Crop Reporting System Annex E - OMS and Flour Mill Allocations

Tables (14) Charts (9) I. Introduction

This is the third and most comprehensive joint U.S.-BDG evaluation of the P.L. 480 Title III Program. The basic Title III Agreement was signed on August 2, 1978, and amended on June 29, 1979. The first joint evaluation occurred in January 1979, and the second was completed the following June. Unlike the first two evaluations which were narrowly focussed, this one is broader in scope. It is intended to provide the basis both for significant changes to the current Agreement and for future Title III Agreements.

Part II of this Report contains a summary of overall findings and recommendations; Part III examines the agricultural sector and the food system context in which the Program must operate; Part IV constitutes both an evaluation of the Bangladesh Government's performance with respect to Title III objectives and a reassessment of those objectives, with recommendations for change. The recommendations are confined mainly to the Amendment to cover the remaining term of the current Agreement (remainder of U.S. fiscal year 1980). A separate paper will provide detailed recommendations with respect to the next P.L. 480 Agreement.

The research that provided the basis for this Report included visits to eight of the country's 62 subdivisions, undertaken during the period, Oct. 15 - 26, selected mainly on the basis of geographical diversity. The subdivisions visited were:

Khulna Division

Satkhira ( District) Magura ( District)

Dacca Division

Tangail () Goalondo (Faridpur District)

Rajshahi Division

Kurigram () Thakurgaon fDinajpur District)

Chittagong Division

Brahmanbaria () Sunamganj ()

Those participating in the evaluation, including the field trips, were Richmond Allen, W. T. Oliver and Jon O'Rourke from USAID; Dr. David Kunkel from USDA; and Dr. Richard Newberg, on temporary du. - with USDA. Accompanying the USAID/USDA teams on the field trips were Nizam U. Ahmed and Harun-al Rashid of the USAID Program Office; and Messrs. A. Siddiqui and A. Aziz, both Deputy Directors in the Directorate of Supply, Distribution and Rationing, Ministry of Food. II. Summary Findings and Recommendations

A. Findings; ProductionFood Distribution and Prices

(1) Stability and adequacy of production - Foodgrain Produc­ tion in Bangladesh is relatively stable on a year-to-year basis. From the Mid-1960's, and excluding the war years of 1971-72, production has not dropped more than a million tons (8 percent) from any one crop year to the next. The more important problem is the perennially inadequate level of production and its failure to increase significantly over time. Per capita foodgrain availability, i.e., production and imports com­ bined, is 6 to 7 percent lower than in the late 1960's. Thus, given the country's marginal subsistence level, even moderate year-to-year pro­ duction declines can assume a significance of crisis proportions.

(2) Seasonality of production - Unlike the situation in most countries where the total crop is harvested in two or three months of the year, Bangladesh farmers now harvest foodgrains virtually the year-around. Instrumental in this situation has been the rapid growth of boro rice and wheat which are harvested during February-May and now account for about 20 percent of toklal foodgrain production. (Aman rice, which is harvested during November-January, accounts for about 55 percent of the total and aus rice, barvested during July-October, for the remaining 25 percent.)

(3) Need for crop diversification - Given the year around pro­ duction potential, small farmers in most (but not all) areas can protect themselves from seasonal shortages and crop risks th-ough crop diversi­ fication. The need is for improved cropping systems, including new crops (potatoes, vegetables, oilseeds, etc.), and increased access to iputs, especially irrigation.

(4) Price patterns - Despite the relative stability in the long term production trend, there have been extremely wide year-to-year price nwings in which crop shortfalls, low stock levels, and anticipa­ tion of reduced stock levels have all played a role. Thus, prices rose by 170 percent between 1974 and 1975, and by 70 percent between 1978 and 1979. On the other hand, as would be expected from the pattern of near year-around foodgrain production, seasonal price variations in Bangladesh have been relatively modest. The natural tendency is for prices to be highest, to about an equal extent, during the "lean seasons" of March-May and August-October. In fact, owing to the consistently higher level of offtakes under the Public Food Distribution System (PFDS) during August-October, prices have been highest during March-July.

(5) Farmers' liquidity - Many farmers are in their best cash position of the year during August-October when they have just marketed their jute and mesta crops. Others have either just sold or are holding for consumption their aus rice. -2­

(6) Sto.k levels and offtakes - Decisions with regard to off­ take levels have been based mainly on stock availabilities. Thus, dis­ tribution has sometimes been high even when crops are good and prices low. As a result, strategic reserves do not accumulate; and since donors tend to base their allocations on stock levels, there is little incentive for the government to seek their accumulation.

(7) Rice demand vs. wheat demand - Cross-elasticity of demand between rice and wheat has been generally strong, with wholesale wheat prices averaging about 65 percent of coarse rice prices. However, the relationship is relatively weak during March-May when both the domes­ tic wheat harvest and substantial supplies from the Food-for-Wo'k (FFW) Program arrive on the markets, and tends to be esrocially close during periods of generally reduced purchasing power.

B. Title III Program

(1) Incentive prices to farmers - Effective procurement programs, based on generallyadeguate incentive prices to farmers, were carried out during both the 1977/78 and 1978/79 rice harvests, and in the wheat harvest of 1979. The Government has shown itself to be more interested in procure­ ment for meeting the perceived needs of theration system than in the more appropriate objective of providing adequate incentive prices.

(2) OMS Program - Evaluation of the OMS Program may be considered in terms both of program performance and of the appropriateness of program objectives.

(a) Performance

1. Although coarse rice prices were well above the target price of rice in most areas during most of 1978-79, OMS sales have not approached the intended levels (total sales of 126,000 tons during Sept. 1978 - Oct. 1979 vs. the 400-600,000 tons allocated for the full 24-month period). Low stock levels were partly responsible (given the Title III provision that OMS need not take place when foodgrain stocks fell below 500,000 tons); yet stocks were adequate to support NR sales of 503,000 tons during the same period.

2. A high proportion of OMS sales was to flour iills. Such sales, while not prohibited by the terms of the Agreement, did not contribute to the main program goal of constraining coarse rice prices.

3. The objective of coarse rice price constraint was not achieved. Low stock levels and highly unusual market conditions had much to do with the program's failure in this respect, but so too did erratic program management and the evident preference of BDG officials for MR as a means of moving foodgrains to the rural areas. -3­

4. Even under favorable conditions, it is unlikely that the rice price constraint objective could have been achieved to any significant degree. Seasonal differences in cross elasticity of demand between rice and wheat required a degree of fine tuning of the rice/wheat pricing structure that was not allowed for in the Agreement and would have been extremely difficult to effect in practice.

5. As market prices of wheat and atta rose far above the OMS price, the BDG established rigid operational controls. Dealers who purchased wheat at Tk.87/md. could resell at Tk.100/md., or in some areas, Tk.95/md. Individual allotments were restricted to from 2 to 5 seers per person, depending on the area. Licensing and selection of dealers became increasingly controlled and even politicized. In some areas, only MR card holders could purchase OMS wheat. Justified as a means of preventing "windfall profits", the effect of these measures was to turn OMS into another ration channel.

(b) Objectives Reconsidered

1. The program was designed to moderate seasonal price swings. As noted above, the more important problem is large year-to-year fluctuations caused by major crop shortfalls or stock shortages.

2. As far as seasonal price swings are concerned, these can and in the past have been influenced by appropriate variations in the pattern of offtakes through the various channels of the PFDS.

3. A long term objective of the program was reduc­ tion of the highly subsidized ration system. Constraint of rice prices through OMS would presumably obviate the need for controlled sales at artificially low prices. In fact, given the low, fixed, price of OMS wheat, the immediate result was to add yet another subsidy channel to the existing system.

4. While the Title III Agreement called for the promotion of private enterprise in the wheat trdde, the establishment of an OMS price only barely above (later, belcw) the procurement price operated to preclude attainment of that objective. Private traders were not provided a margin within which to procure, store and later resell wheat.

5. If the basic objective is the increased avail­ ability of a reasonably priced foodgrain, wheat (or atta) should suffice for the purpose. From that point of view, the constraint of coarse rice prices should not be a matter of great concern.

(3) Use of Local Currency Proceeds; Loan Forgiveness - Following -4­

are the principal findings with regard to the local currency use and loan forgiveness aspects of the Agreement.

(a) The BDG has reported depositing the local currency equivalent of ^16,500,000 of Title III wheat sales (OMS and MR) to the Title III Special Alccount. Further, they have reported "allocating" some $15.1 million of this amount to USAID's Fertilizer Distribution and Improvement Project. However, there has been no indication that proceeds have been transferred to the Project Account, let alone disbursed from that account for project use.

(b) In allocating funds to the Fertilizer Project, the specific intended use appears to be financing of the fertilizer subsidy, as opposed to construction of fertilizer godowns or other more tangible aspects of the Project.

C. Goals and Objectives

(1) The major objectives of the Title III Program should be to accelerate agricultural development and to help provide food security. The latter may be defined as having a nutritious foodgrain available throughout the country at a price within a reasonable range above the procurement (support) price to farmers.

(2) Major price surges reflect supply shortfalls arising from periodic events (such as the 1979 drought) to a greater extent than seasonal supply variations within the course of a single year. Hence, ioajor emphasis should be placed oa responding to year-to-year supply rather than seasonal variaLions. In meeting the food security objective, fullest emphasis should be placed on wheat (and atta).

(3) Foodgrain production should be encouraged by the establish­ ment of procurement prices announced well in advance of planting seasons. The primary objective in the establishment of procurement prices and their implementation should be the provision of incentives to farmers. As a by-product, the grain being procured by the Government will contri­ bute to the augmentation of its food reserves, but the Government should not seek procurement for other purposes.

(4) In implementing the price support program, full and adequate use should be made of private intermediaries, especially in order to reach small farmers and remote areas that would otherwise not be served.

(5) Farmers, dealers and consumers should be encouraged to store for off-season use to reduce the load on Government facilities. They will undertake to do so to -he extent they can foresee a margin between the procurement and market price of wheat sufficient to cover costs of ­ chase, assembly, storage and resale. The OMS price should be high enough to cover the procurement price plus dealers' -storage and handling costs. -5­

(6) Foodgrain subsidies should be reduced and ration operations increasingly directed to the benefit uf the poor. To this end:

(a) Distribution to Statutory Rationing (SR) and less essential categories should be reduced, and increasingly shifted from rice to wheat;

(b) Modified Rationing (MR) should be directed to the poorest by making assets and income the sole criterion of eligibility;

(c) Food-for-Work (FYW) should be used when and where it can economically and properly supplement productive employment for the very poor;

(d) Relief should be directed to those unable to earn a living;

(e) Ration prices should be progressively raised. The near­ term objective should be ration prices equal to the procurement prices of rice and wheat. The longer term objective should be prices sufficient to cover costs of procurement and distribution;

(f) Food distribution should emphasize low-cost alternatives acceptable to the lowest income groups, products not strongly competed for by people who can afford to pay more, e.g., wheat rather than rice; wheat as atta rather than refined wheat products; gur rather than refined sugar; lower cost oils, etc.

(7) Taka proceeds from the sale of PL 480, Title III commodities should be used for development projects which directly impact on increasing foodgrain production in the next five years.

D. Recommended Actions

Following are the principal recommendations for inclusion in the Amendment to the current Agreement. A complete listing appears in Section VI.F.

(1) OMS

(a) Eliminate the rice price constraint objective;

(b) Establish the OMS price of wheat at a price sufficiently above the ptocurement price to provide private traders a margin for the storage, handling and resale of wheat;

(c) Base OMS sales on an atta trigger price and sell wheat in any subdivision when the trigger price is exceeded; -6­

(d) If the market price rises above the trigger, raise the OS price accordingly. (e) Eliminate dealer restrictions and make a positive effort to sell to atta dealers;

(f) Eliminate resale price and volume restrictions on OMS.

(2) Local Currency Use

(a) Permit loan foregiveness upon proof of disbursements from project accounts;

(b) Eliminate the Fertilizer Project from the list of projects qualifying for loan foregiveness.

(3) Supply and Attribution of Wheat

(a) Add 200,000 tons of wheat already budgeted for Bangla­ desh in FY 1980 to the existing Title III allotment of 742,000 tons, making a new total for FY 1978-80 of 942,000 tons;

(b) Of the 942,000 tons, 400,000 tons may be attributed to NR; the remainder to OMS or, to the extent not needed, to BDG sbocks. Provided that terms of the Agreement are complied with, Title III wheat not used (i.e., attributed to MR or sold through OMS) may be treated as having been sold, and that amount of the loan eventually written off.

(c) It is to be understood that the terms of the Amendment to the current Title III Agreement are to apply to all of the wheat supplied - the original 742,000 tons as well as the 200,000 tons being added from U.S. FY 1980 availabilities. Since attributions from wheat already received have not approached entitlements under the terms of the original Agreement, this will not pose a problem in accounting terms. -7-

III. Production, Prices and Public Food Distribution

A. Production Trends

During the strife and chaos which accompanied the struggle between East and West Pakistan and which led to the separation into independent nations, production in what is now Bangladesh fell off drastically. Since then, agricultural production has recovered and increased to new peaks. However, overall output in 1979 was only marginally above the pre-independence (1969-70) base. The long-term trend in foodgrain production has been upward at a rate of about 1.6 per cent per annum, with year-to-year variations about the trend line of up to 8 per cent. Rice production has increased by about 1.2 per cent per annum from the late 1960' to 1978/79. Wheat production has increased sixfold since 1969/70 and now accounts for nearly five per cent of total foodgrain output. Production of commercial crops such as jute and sugar cane has declined.

The major increases in producti3n have come in boro rice and wheat, which together have increased from about 700,000 MT in the mid-1960's to about 2.6 milliun MT in 1978/79 and will likely exceed 3 million MT in the near future. Meanwhile, aman and aus have increased only about 12 per cent. The rapid growth of boro rice and wheat has had two important consequences, in addition to the impact on total supply. It has tended to reduce weather­ caused year-to-year fluctuations, especially from flooding, and to even CaUt the domestic supply available during the course of a single year. The result is a relatively low year-to-year variability in production and a relatively low level of seasonal variations in supply and price. Table 1 shows annual output from different crops.

Acreage of rice forall crops combined has increased very little over the past 10-15 years. Rice acreage remained almost unchanged at about 22.5 million acres until the late 1960's, then increased by about 2 million acres, but since then has remained unchanged at 24-25 million acres. The -us aid aman crops have changed very little in area cr production but the boro grew rapidly in the late 1960's both in area and yield and has continued to increase, albeit more slowly since. Wheat acreage continues to expand rapidly. Total crop acreage has held at 30-32 million acres throughout. Table 2 shows acreage by crop since 1969/70.

B. Crop Yields

(1) Current situation and trends

With Bangladesh near the capacity of its cropped acreage, future production gains must come primarily from yield increases. Yields of rice have gone up somewhat, principally in the boro crop, and overall because of the increased boro acreage. Also, wheat yields have more than doubled since the 1960's. Nevertheless, foodgrain crop yields in Bangladesh are extremely low; there is tremendous scope for increases. -8-

High yielding variety (HYV) strains account for only 15 per cent of overall rice acreage. The proportion of acreage planted to HYV varies according to crop, as follows: aus (8 million acres), 12.5 per cent; T.amay (10 million acres), 9 per cent; and boro (2.7 million acres), 54 per cent.- The HYV varieties yield collectively only about 900 kilograns per acre. Local variety yields are 320 Kg. for aus, and 500 Kg. for both T.aman and boro. In addition, some 4 million acres are sown to broadcast (or deepwater) aman, yielding some 400 Kg. per acre. Research results in stations indicate a potential several times these levels for all rice varieties. Table 2 summarizes yield levels and trends for the principal rice crops since the 1960's.

(2) Yield Factors

(a) Price incentives

Since the 1977/78 aman season, the Government, through its rice and wheat procurement programs, has provided generally adequate incentive prices to farmers. The relative failure of the 1978/79 rice procurement program (total procurement of 353,000 tons) has been cited as evidence of an inadequate procurement price in that year, a judgement facilitated by the fact that the procurement price was not increased from 1977/78 to 1978/79. In fact, the "failure" of the procurement program in 1978/79 was due primarily tc the poor aman crop in that year. Procurement was more than adequate in relation to the size of the crop (see further Annex A). Incentive price policy is discussed further in Section IV.

(b) Fertilizer

Growth in fertilizer use has been one of the brighter spots in the agricultural sector. From the base years 1967/70 to 1978/79 fertilizer offtakes have grown from an average 236,000 tons per year to 741,000 tons. In nutrient terms this growth has been from 110,000 tons to 345,000 tons. About 80 per cent of the total fertilizer has been applied to foodgrains. Thus, nutrients going to foodgrains increased from about 90,000 to about 270,000 tons. Considering that very few Bangladesni farmers use recommended doses of fertilizer, and that use is at less than optimal efficiency even allowing for that fact, the additional 180,000 tons should have been sufficient to increase production by about 1.5 million tons. The actual increase during this period was 1.35 million tons, but then 1978/79 production was reduced sonewhat by drought. There is a fairly close correlation between increases in fertilizer use and increases in food grain production over the period 1965/6 to 1978/9. Most of the discrepancies can be explained by weather and other factors. As Table 4 shows, the fertilizer/rice price ratio (number of seers of fertilizer required to buy a seer of rice) has varied considerably over time. Yet there has been little correlation between these ratios and the trend of fertilizer

./Data refer to the 1977/78 crop year, except 1978/79 in the case of T.aman. -9-

ratio, wniie I1UCUd.LLr, lo off takes. Doubtless this is because the the corresponding ratios largely favorable in comparison with remained or declines in Asian countries. Rapid increases prevailing in other as related to other factors, such fertilizer offtakes are probably weather, or - as in the drought available supply, distribution effici.ency, in farmers' cash availabilities. aftermath of 1979 - abnormal swings

(c) Irrigation Bangladesh has been Development of irrigation systems in to the need. woefully inadequate relative irrigated under Between 1960/61 and 1970/71 the area acres under low increased from 64,000 to 820,000 Government (BADC) programs then to 32,000 under tubewells. Since pumps, and from essentially zero lift about 1.4 million acres and irrigated area has increased to low lift pump since 1970/71 million (Tables 5, 6 and 7). Thus, tubewells to about 0.4 1,800,000 increased from about 850,000 to the total BADC irrigated area has 6 per cent of the total area or from about 3 per cent to about acres, in 1972/3 when low lift The largest increase was reported cultivated. to 1,230,000 acres. The reportedly increased from 864,000 irrigation alone between the the substantial increase in irrigation production impact of appears to measure. The most notable impact 1960's and 1978/9 is hard to appear rice and wheat acreages. It would been in the expansion of boro have the area under irrigation but needs to be done not only to expand that much methods and facilities. the efficiency of existing irrigation also to improve land use offer the major of irrigation and more intensive More efficient use The focus should in food production in Bangladesh. prospects for increase shallow well pumps tubewells, more efficient animal-powered be on small-scale Small pumps are and even solar-powered small pumps. (e.g., flow pumps), Bangladesh. small fragmented holdings of especially important for the

(d) Research and Extension national to be general agreement that the There appears not large numbers of employees, is and extension network, despite research many different organizations Fragmentation of effort among up to the task. have been widely cited both in research and extension and lack of coordination organizations exist Separate research and extension as major problem areas. research Livestock not only has a separate for different commodities. is a tendency ministry from crops, and there organization but also a separate medicine to diverge. for animal science and veterinary by farmers, but on cropping systems is much needed Research in research and relatively little attention until recently had been receiving some good needs are far from being met, extension organizations. While rice varieties be produced. More location-specific research is now beginning to -10­

are being identified or developed, including higher yielding deep water rice reportedly capable of producing 1.5 tons of rice per acre. Micro nutrient deficiencies have recently been identified and are now receiving increased attention. These deficiencies, such as zinc and sulfur, may be factors in low fertilizer response.

In short, the need is for increased research both on rice varieties and on nutrient deficiencies, for increased extension activity, and for improved coordination between the research and extension branches.

(3) Constraints

In reality, there is no clear-cut division between "factors in" and "constraints to" higher crop yields. Thus, irrigation is a "factor", but the slowness with which it has spread is a "constraint". Fertilizer is regarded here from both points of view; its use has expanded rapidly, but there appear to be constraints to its more wide-spread adoption. The separation made here, therefore, is mainly for purposes of convenience. This section covers factors that operate mainly to constrain agricultural growth.

(a) Agricultural credit

The 1978 Lar.Q Occupancy Survey (LOS II)I / revealed that more farmers than had previously been thought have access to institutional credit at reasonable rates. 2 However, the number remains extremely low relative to the need. The need has become especially acute this year, with farmers cash short following three successive below average rice crops ­ 1978/79 aman and boro and 1979/80 aus - and abnormally low pices for the major cash crop, jute.

(b) Land tenure

About 23 per cent of the cultivated land is farmed on a sharecropping basis. Under the normal land tenure arrangement, the sharecropper bears the full cost of inputs, but must surrender half of his crop to the landlord.

The effect of this situation on sharecropper use of agricultural inputs and crop yields is mixed and not altogether clear. LOS II revealed that the percentage of tenants and owner-cultivators using fertilizer is about the same. However, USAID surveys have shown that intensity of use - the amount applied per acre - is considerably greater amorng owner-cultivators. Tenant

J/ Land Occupancy Survey, 1978, Janzzi & Peach, April 1979. 2./ LOS showed that 47 per cent of the loans obtained for use on owned land, and 22 per cent of the loans obtained for use on tenanted land, were from institutional sources. i.e., banks, cooperatives or government sources. -11­ farmers are also less likely to use irrigation, and we may surmise that in general they farm the poorer quality land. All in all, it is perhaps surprising that, as field surveys have also tended to show, yield results achieved by sharecroppers average very close to those obtained by owner­ cultivators for a given variety of cultivation. In general, tenants use more of their own labor than owner-cultivators. Perhaps this factor compensates, partially at least, for their known disadvantages.

As discussed, tenant farmers are able to obtain crop yields nearly as high as owner-cultivators for a given variety of cultivation. The significant production constraint involved in the land tenure situation arises from the fact that very few tenant farmers are willing to undertake HYV cultivation. The latter requires a significantly higher cash investment than local variety cultivation, as well as a greater risk of crop failure. Nor is the sharecropper, as a renter, motivated by considerations of improving his return to capital. In short, his relative poverty, reluctance to expose himself to risk, and lack of ownership motivation combine to inhibit sharecropper adoption of HYV technology.

(c) Fertilizer distribution

As noted earlier, price ratios between fertilizer and foodgrains appear to be favorable to the farmer, but the supply and distribution system may not be adequate notwithstanding the reported 40,000 fertilizer dealers. USAID, through its Fertilizer Distribution and Improvement Project, has been trying to improve dealers' incentives by gradually shifting the distribution system to the private sector. However, it is not clear to what extent the program has actually resulted in improved dealers' margins (even in the limited area - - to which the project has thus far been restricted), or otherwise motivated them to become active salesmen as opposed to passive distributors.

C. Foodgrain supply and per capita consumption

Over the period from 1967-70 to 1979 total production of foodgrains (rice and wheat) has increased at a rate of about 1.6 per cent per year while population has grown at a rate of about 2.6 per cent per year. Given the slow rate of growth in production, per capita foodgrain consumption could be sustained only by increasing imports substantially more rapidly than the rate of population growth. In fact, however, imports have remained virtually unchanged, save for upward spurts during crisis years (1972/73, 1974/75, and again in 1979/80). Combined production and imports increased between 1967-70 and 1975-79 by 15 per cent while population increased by about 22 per cent. The result was a decline in per capita consumption between those two periods of 6 to 7 per cent, or from about 16 to 15 ounces of foodgrains per capita per day. (See Table 8) Other sources tend to verify thi3 gap. The Royal Nutrition Survey of 1975/76 found that caloric intake per capita had declined from 2,224 in 1962-64 to 2,094 in 1975/76, a decline of 6 per cent; and a study -12­ by the Institute of Nutrition and Food Science of Dacca University fouud a 7 per cent dline in per capita caloric intake between the years 1962-64 and 1975-76.-

The decline in food supply available per capita has been reflected in the trends of rice prices and wages over this period. From 1962--64 to 1974/75 real industrial and agricultural wages declined by about 50 and 40 per cent, respectively. The declines between 1969/70 and 1974/75 were of about the same order (Tables 9 and 10). In contrast, the price of rice in real terms was sharply up in 1975. For 1977-79, the index of real wages averaged about 75 (1969/70 = 100) and the deflated price of rice averaged about 90. (Figure 5) Assuming an elasticity of demand of -0.3 (see further 2/ Section D below), this would suggest a reduction in supply of about 7 per cent.- These price data appear to verify the conclusion that an increase in supply of about one million tons per year will be needed, initially from imports, to avoid major inflationary price increases as real incomes rise. The recovery in real wages apparently taking place during _975-79 may have been a factor in the rapid rise in the price of rice in 1979.31

In short, the principal problem with respect to food supply is that combined imports and production are too low and failing to keep up with population growth. Additional problems, discussed in Annex B, are insufficient. food reserves to meet year to year variations in output and poor scheduling of imports.

D. Prices and public food distribution

(1) Elasticity of demand

Foodgrains provide 85 per cent of the total caloric intake of the population of Bangladesh. This percentage is relatively stable over a fairly wide range of per capita income distribution, though of course total caloric intake iucreases rapidly with income from the lowest to middle and upper income groups. Estimates of the income elasticity of demand for foodgrain vary widely. The IBRD cites two levels of estimates for rice:

1/ Nutrition Survey of Rural Bangladesh, 1975-76, Institute of Nutrition and Food Science, Dacca University, Dec. 1977. 2/ Rice prices in real terms are 20 per cent higher than real wages relative to the 1969/70 base period (90 t 75 = 1.2). Further, .20 i -.03 = -.073. The deflator used in making these calculations is the countrywide industrial workers cost of living index, there being no agricultural (or rural) index available. Moreover, the reliability of the basic wage data is open to serious question. -13-

(from Agarwala) .11 No estimates were -0.3 (from Alamgir and Berlage) and -0.6 combined. Based on a review of other found for wheat or for wheat and rice we think that a price elasticity studies and experience in other countries, close to reality; i.e., a 1 per cent for foodgrains of -0.3 is likely to be being equal, induce a 3.3 per cent change in supply will, other things (.01 i -.3) opposite change in price.

(2) Trend analysis System (PFDS) The Government through its Public Food Distribution and seasonal variations in supply and has attempted to attenuate year to year prices of foodgrains. are viewed In figure 1, crop production and food distribution being the BDG reporting of the fiscal year, July to June, this in the context view tends to procurement, offtake and imports. This cycle for production, offtake production variation while minimizing accentuate year-to-year to production in fact that offtake is not responsive adjustment. It suggests pattern viewpoint of aman to aman harvest, the changes. Looked at from the to in recent years adjusted rather well appears quite different, with offtakes (figure 2). compensate for production changes there have been very Despite the rather stable supply pattern, (figure 3), with three periods of large year-to-year fluctuations in prices sharp price increases in the 1970's. first 8 months of 1972 The rather sharp rise in prices in the political uncertainty. This was explained by disruption of supply and can be starting in January in prices of unprecedented sharpness followed by a rise 1975. The 1974, and finally peaking in March 1974, continuing until October the spring and summer of 1979 following third very sharp increase came in the unprecedented drought of that year. boro during 1977/78, PFDS Despite good crops in aus, aman and levels. Prices were stable that year increased above 1976/77 offtakes in 1977 to December 1978 High stocks throughout the August through December 1978. offtake probably were a factor in the high period (910,000 tons in December) levels. November/December 1978 A poor aman crop in 1978/79 and reduced the sharp rise in prices plus high levels of procurement, triggered offtakes, 1979. Offtakes increased 1979 and continuing through July beginning in January in 6 months). arrivals (only about 150,000 tons somewhat, but were limited by low

Issues, IBRD, /Bangladesh:Current Trends and Development March 1979 - p. 61. -14-

By June stocks were down to one month's off takes, but heavy arrivals were scheduled for July onward (over 400,000 tons/month). The prospect of heavy arrivals apparently helped to moderate the price rise. By August, the aus harvest, heavy July unloadings, a major increase in stocks and some increase in offtakes had begun to reverse the price upsurge. By late November, rice prices were down to their early June levels, though still nearly 40 per cent above the levels prevailing in November 1978.

The conclusions are not simple and clear cut. Price appears to be affected by a number of factors including:

1) The actual supply in the market which is a function of:

a) production less procurement

b) imports and the resulting public food distribution.

2) Expectation with respect to the supply which in turn is

largely a function of:

a) crop recently harvested

b) outlook for the next crop

c) Government stocks and expected arrivals.

Supply expectations appear to be as important as actual supplies. The 1974 price rise, for example, is difficult to explain in light of the good 1973/74 crops and fairly heavy, if declining, offtakes. Perhaps people replenished their depleted home stocks and may have been concerned at declining levels of offtake and Government stocks. (Doubtless also, dealer hoarding and speculation, reportedly at exceptionally high levels, were factors in the sharpaess of the price increases at that time.) Again, in 1979, the very small level of arrivals planned and rapidly declining stocks may have constituted a negative psychological factor going beyond the actual supply situation.

(3) Seasonal analysis

In the 1960's, 94 per cent of total foodgrain production came from the aus and aman harvests. The aus harvested in August-October accounted for about 28 per cent and aman harvested from late November to early January about 66 per cent. In other areas where aus is grown, this crop would fairly well take care of home consumption needs from August until the aman harvest and most of the commercial market needs in that period. However, there are areas where because of flooding aus is a small or non-existent crop, and in those areas, by August-October, supplies would be very low and prices rising. -15-

In more recent years, with rapid increases in boro rice and wheat output (now 20 per cent of total foodgrain production), seasonality of supply has been reduced. It is still a problem, but a minor problem compared with year-to-year supply shortages and price rises. In contrast with year-to-year price rises as high as 170 per cent in 1974 and 70 per cent in 1979, seasonal prices tend to remain within a range of plus and minus 10 per cent of the annual average. Adjusted for inflation the seasonal index covering the period 1972-77 was lowest in December at 84 and highest in April at slightly over 110. The period, August-October, generally considered as constituting the principal "lean season", was slightly below the average for the 6 year period. In recent years the seasonal price peak has been in March, April, May despite the growing harvest of boro rice and wheat during this period. In 1978 the price pattern was similar, with the peak in April and low in November-December (Figure 4). For 1979, it now appears that the peak will have come in July.

In 10 of the last 13 years 1967-79, the annual high in prices has come prior to the aus harvest (March-July). Disregarding the two years 1970/71 and 1971/72, which were not normal in the sense that the causes of shortage and price rises were mainly war and the aftermath of war, peaks in 10 of the remaining 11 years came in March-July.

It would appear that the tendency for rice prices to be higher during March-July than August-October can be explained by the pattern of PFDS offtakes. During the period, 1972-78, offtakes averaged 30,000 tons higher during August-October than dur.ng March-May.- This difference would in turn be equivalent to about 3 per cenz of average monthly consumption during those years. Assuming an elasticity of demand of -0.3, this in itself would lead to a price difference of about 10 per cent.- / The actual difference for those years was 11 per cent. Thus, August-October and March-July would be seen as about equal in terms of the rice price tendency were BDG offtakes evenly distributed as between those two periods. As the boro and wheat crops continue to gain in relative importance, August-0ctober should finally emerge as the "true" lean season. Meanwhile, the two periods deserve roughly equal attention as far as seasonal trends are concerned. Of more significance, as discussed in Annex B, the country's vulnerability to aman crop failures provides an important additional reason for directing more attention in stock level and distribution planning to the March-July period.

I/ The advent of Food For Work in 1976 and the concentration of FFW during the Jan-May dry season has served to moderate but not eliminate this tendency. Total offtakes were 40,000 tons higher during Aug-Oct than Mar-May during 1972-75, and 14,000 tons higher during 1976-78. 2/ .03 -. 10 -- 0.3 -16-

IV. The PL 480 Title III Agreement

A. Background

The legal basis for the Title III program was provided ir the International Development and Food Assistance Act of 1977 which amended the Agricultural Trade Development and Assistance Act of 1954, to include under Title III of PL 480 a new Food-for-Developmant (FFD) Program. The new legislation provided for multi-yp-r commitments with loan "foregive­ ness" in exchange for agreements on agreed development projects. Speci­ fically, as proceeds from sales of PL 480 commodities are disbursed on agreed projects the dollar equivalent of those accounts are credited against loan repayments that would otherwise be due.

Former PL 480 Title I agreements had been instrumental in en­ couraging the BDG to adopt a foodgrain procurement program to support incentive prices to farmers. In 1978, USAID advanced the notion of applying the Title III legislation to the concept not only of supporting incentive foodgrain p':ices at harvest time, but also constraining rice prices through open market sales (OMS) of wheat during the "lean" or pre-harvest season to protect the incomes of small farmers and the landless.

B. Terms of the Agreement

The Title III FFD Agreement was signed on Aug. 2, 1978. It provided for all of the basic elements discussed above, i.e.: procure­ ment prices to serve as production incentives during the harvest seasons; an OMS program to restrain prices during lean seasons; and attributions of OMS proceeds to agreed development projects. The Agreement called for the U.S. to supply 800,000 tons of wheat during a period of 24 months spanning three U.S. fiscal years, as follows: 200,000 tons in FY 1978, 400,000 tons in FY 1979, and 200,000 tons in FY 1980. Of the 800,000 tons, a minimum of 400,000 tons (200,000 in each 12-month period) was to be used for open market sales and 200,000 tons were earmarked for the MR system. The remaining 200,000 tons could be used for OMS if demand should require it; otherwise, the last 200,000 tons would be allocated to MR, Category A.

C. Price Incentives

(1) Objectives

The pricing policy contained in all recent PL 480 Agreements has accorded top priority to the provision of adequate price incentives for farmers to adopt yield increasing inputs and practices. An incentive price program requires:

-- that adequate prices be provided, with protection from m; jor seasonal (harvest-time) or abnormal year-to-year declines. In addition to a government procurement operation involving a fairly large -17­ number of widely dispersed buying points, an effective price support program should involve a large number of private dealers throughout the country acting as intermediaries, buying from farmers at harvest time and selling to the government. Dealers would perform this function only if they had reason to expect a reasonable margin between prices at harvest time and later resale.

-- that farmers have sufficient time to respond to favorable prices. This requires that price support levels be announced well in advance of the planting season.

The difficulty in arriving at appropriate support (procure­ ment) prices should not be underestimated. A dilema in this regard arises from the structural problem inherent in the tenancy system. Given the normal crop-sharing arrangement whereby a tenant is required to surrender half of his crop to a landlord, an incentive price that will produce a benefitL cost (B/C) ratio that is adequate for tenant farmers, say 2:1, becomes more than adequate for owner-cultivators. Moreover, since under present conditions B/C ratios for tenants are in fact considerably less than 2:1 (averaging about 1.8:1 and ranging down to 1.01/ or even less, depending on region and crop variety), the sudden adjustment required to provide all tenants an adequate incentive price would involve the govern­ ment in far more p.rocurement than it is presently prepared to handle, while forcing up the entire scale of foodgrain (and eventually all other) prices to the detriment of the rural poor.

B/C ratios to tenant farmers could presumably be improved by reducing the cost of inputs; i.e., subsidizing them, or subsidizing them to a greater extent than they already are. Indeed, a case can be (and frequently is) made to the effecL that it is less costly to subsidize such inputs as fertilizer than to increase output prices as a means of effecting improved price relationships for both farmers and consumers. In Bangladesh, however, where fertilizer is already heavily subsidized,2/ such a course does not seem advisable. Moreover, as was indicated in Section III.A, factor fertilizer/rice price ratios do not seem to have been a significant in farmers' decisions with respect to fertilizer use. Use itself can best be pursued by measures to encourage private sector participation in fertil­ izer distribution, along lines currently being pursued.

of 1/ B/C ratios are calculated on a cash basis and do not take account returns to capital.

2/ The Bangladeshi farmer currently pays about 40 percent of the world price, or domestic cost of production, of fertilizer. The fertilizer subsidy itself accounts for some 75 percent of the entire BDG develop­ ment budget for agriculture. -18-

In short, there is no easy solution to the problem of setting incentive prices that will be "Just right" both for owner-cultivators and for tenant farmers. Policy should continue to be geared to improving the sharecropper's lot by improving his access to institutional credit and other inputs. The incentive price should be adequate, in combination with the improved access to inputs, t- i.nduce HYV investments by at least a major portion of tenant farmers.

(2) Performance

In general, procurement prices have been adequate, and the system has functioned well enough, to provide adequate returns to production. Performance has, however, varied during recent years, and the Government's motives have clearly been based more on procurement to meet the needs of the ration system than-on farmers' incentives.

1977/78 - The procurement price is generally considered to have represented an adequate incentive price for f trmers in 1977/78. Announcement of the prices was early enough to affect farmers' decisions and, aided importantly by a record aman crop, the 1977/78 procurement program was highly succes-ful.

1978/79 - The Government did not announce a hew procurement price prior to the 1978/79 Pman season, thereby leaving itself vulnerable on two counts:

1. The "low" procurement price was blamed for the subsequent "failure" of the 1978/79 procurement program (353,000 tons procured in 1978/79, as compared with 550,000 tons the year before).

2. More importantly, since farmers' costs of production had risen since 1977/78 by 10 to 15 percent, it was at least arguable that the procurement price was no longer adequate as an incentive price.

As to the first point, the reduced procurement in 1978/79 was attributable mainly to the decline in aman production in that year, not to the "low" procurement price. Under the circumstances, efforts to force greater procurement would have been a mistake (see also Annex A).

Matters are not so clear with regard to the second point. However, it is our judgment that farmers had a sufficient "cushion" in the form of a perhaps overly generous procurement price in 1977/78 to enable them to absorb the increases in production costs over the following year. The procurement price appears to have remained generally adequate as an incentive price in 1978/79.

Increases in procurement prices for rice and wheat were eventually announced in March 1979, far too late to influence farmers' decisions with respect to boro and wheat procurement. Clearly, this was a case of pursuing procurement to satisfy the perceived needs of the ration system. -19­

1978/79 - Adequate incentive/procurement prices were announced for the 1979/80 aman crop, but the ammouncement - on June 30 - was about one month later than desirable. Procurement prices were raised again (rice from Tk.158/md. to Tk.170/md.; wheat from Tk.90/md. to Tk.ll0/md.) on Nov. 12, 1979. The latter move was taken without consultation with the Mission, and thereby represented a violation of at least the spirit of the Title III Agreement. Moreover, since the existing procurement price had been jointly agreed as representing an adequate incentive price to farmers, it was, again, an example of the BDG's pursuing procurement in order to fill BDG storage b.ns.

D. Price Constraints - OMS

(1) Objectives

As noted in Section III, one of the Government's objectives in its operation of the Public Food Distribution System has been to moderate the seasonal swings in rice prices. Thus, offtakes have been heaviest during the perceived (if not actual) major lean season of August-October, lowest during the main harvest seasons. In this respect, the PFDS has enjoyed some success; seasonal price movements have been relatively moderate, at least by comparison with the far wider price swings caused by major crop failures and/or stock shortages.

It may be added, parenthetically, that the Government had for some time operated an Open Market Operaticns (OMO) program designed speci­ fically to curtail rice prices in selected areas. Oftakes under OMO were 49,000 tons in 1976 and 27,000 tons in 1977 (3.7 percent and 1.5 percent of total PFDS offtakes in those two years, respectively), then declined to virtually nil beforz the iaunching of OMS in 1978. Given the small scale of the OMO program, its contribution to the moderation of seasonal price swings was minimal.

Whatever the degree of moderation in price swings achieved by the PFDS, it was deemed inadequate. Even with a procurement program functioning with reasonable effectiveness, small and tenant farmers were, it was felt, being whipsawed by the combination of relatively low prices at harvest and relatively high prices in the post-harvest periods. The smallest and most disadvantaged (i.e., tenants) were commonly forced to sell - to pay off debts and meet essential needs - when prices were lowest, and to buy for consumption when prices were highest. These considerations underlay the basic elements of the OMS aspect of the Title III program. As originally designed, OMS involved the following:

-- a program of OMS with the goal of limiting the increase in coarse rice prices to no more than 20 percent of the procurement price in a given year. OMS was preferred to MR for this purpose both because it would be more directly effective in constraining prices, and because it would presumably reach all of the rural poor, not merely those holding ration cards; -20-

than rice. It was recog­ -- OMS to consist of wheat rather wheat was more be more directly effective, but nized that rice sales would OMS price. Central to the theory underlying readily available at a lower was of demand between rice and wheat was the belief that cross-elasticity effect on sales would have a major dampening sufficiently high that wheat rice prices; subdivision when coarse rice -- OMS would commence in any Tk.145/md. They trigger price, initially set at prices exceeded an agreed price; prices declined below the trigger would be halted when coarse rice foodgrain dealers wheat would be sold to all licensed -- OYS initially 10 and 200 maunds, respectively, in minimum and maximum lots of at a price of Tk.85/md.; wheat at a price higher than -- Dealers were not to resell Tk.95/md.;l/ interference with the operation of -- In order to preclude would take place only during the the procurement program, OMS operaticons and the minor lean season of main lean season months of August-October March-April. to pursue OMS should its -- The BDG would not be obligated wheat below 500,000 tons, or if its foodgrain stocks at any time fall tons. stocks should fall below 250,000 endorsing the concept agreement also contained language The in the wheat of private sector involvement of encouraging the devalopment be brought guidance as to how this would trade, though without specific about.

(2) Performance to implement the BDG made a considerable effort Although the were, as the terms of the Agreement, there OMS Program in accordance with / of first evaluation report,­ a number of problems. Some noted in the market conditions prevailing be ascribed to the highly unusual these could with the OMS (notwith­ some to a combination of inexperience during 1979, preference of most 010 Program) and to the evident standing the earlier to the rural MR as a means of moving foodgrains Food Ministry officials for with section considers, first, performance areas. The remainder of this

Annex B to the Agreement, was contradictory on tnis point. 1/ The Agreement would be free to re- provisions, stated that dealers which listed detailed which were the other hand, the Agreed Minutes, sell at any price. On of the Agreement and the BDG as an integral part accepted by both the U.S. was in of Tk.95/md., and this ceiling stipulated a retail price ceiling Controllers of Food by the Food fact communicated to the Subdivisional Ministry. A First Evaluation, Sales Proram in Bangladesh; 2/ The Title III Open Market USAID/Bangladesh, June 13, 1979. -21­ respect to program operations, and second, the extent to which the basic Program objective - the constraint of coarse rice prices - was achieved.

(a) Program Operations

The most notable program failure was the consistently low level of OMS, both relative to the need and relative to the level of MR. Sales for the first two months of program operation, September-October, 1978, totalled 24,000 tons. Sales were suspended in November, owing to the original stipulation in the Agreement that OMS were to take place only during the designated lean season months. This stipulation was dropped, by mutual Agreement, in December, in which month OMS totalled only 169 tons. Sales were 28,000 tons during January-June, 1979, then were vir­ tually suspended as stocks fell below the 500,000 tons level at which point the BDG was not obliged to pursue OMS. As stocks again rose above 500,000 tons in August, OMS resumed on a relatively large scale, and for the period July-October totalled 73,000 tons. Thus, OMS for the first 14 months of the Agreement amounted to 126,000 tons. The low level of sales could be attributed in part, and at certain times, to low stock availabilities. Nevertheless, as was pointed out in the First Evaluation report, sales were at low levels in many areas even when stocks were adequate and rice prices well above the trigger point. Moreover, stocks were adequate to support MR sales of 503,000 tons (four times the level of OMS) during the same 14-month period. As noted, the Agreement con­ templated OMS sales at least equal to MR.

Othei problems in program operations that may be noted:

-- Throughout the period a disproportionately high pro­ portion of sales were made to flour mills, a development that was not contemplated in the Agreement and which could not help to achieve the basic objective of constraining coarse rice prices.L/ Sales to flour mills constituted 59 percent of all OMS sales during the period, Sept. 1978 - June, 1979. This development led to a stipulation in the Amend­ ment of June 29 that sales to flour mills could not exceed 20 percent of total OMS in any one subdivision. From that point, i.e., July - October, 1979, flour mill sales were only 9 percent of total OMS. Thus, the BDG did comply with this portion of the Agreement, as amended.

-- Sales were to be mainly to existing dealers, but in many areas new dealers were "created" to handle OMS wheat.

-- No efforts were made to sell to private non-dealers.

l/ See Annex E for discussion of OMS sales to flour millers in relation to flour mill sales under the flour mills category of the PFDS. -22- -- In 1978 and early 1979, OMS operations were initiated in some subdivisions where trigger prices had not been reached. In other areas they were not initiated though trigger prices were exceeded. In July through October 1979 trigger prices were exceeded in all areas but in some of these sales were very low, reluctantly approved and slowly executed.

The latter development was part of a process whereby, in July-August, beginning 1979, OMS became subject to increasing controls. were ju3tified The controls by the desire to prevent "windfall" profits in a situation involving a fixed OMS price (Tk.87/md.) and retail wheat prices rising rapidly to a range of Tk.160 - 200/md., according to the time and area. The controls included:

-- Increasing control and selectivity over the selection of dealers eligible to lift OMS. During the first week of each month the Sub­ divisional Controller for Food (SDCF) received his allocation for the month. In most areas, a combined committee of food officials and other local thereupon approved leaders individual dealers and passed on individual allocations. The process of reviews and decisions delayed approvals and allocations with the result that sales seldom started until mid-month;

-- Quantity limits ­ from 2 to 5 seers per person - on the amount of wheat that could be lifted by any individual (as noted, resale price limits were already in effect);

-- In many, though not all, areas OMS sales were restricted to MR card holders, thereby completing the transition of OS to what was, effect, an extension in of MR. In some areas visited by the evaluation teams, the practice was suspended for the day of our visit, presumably to resume after we had left (verified in conversations with recipients); -- Whether or not an MR card was required of recipients, in all areas recipients were registered on muster rolls along with their fingerprints.

In general, then, the process was one of a gradual in the control of the tightening OMS operation throughout the second half of 1979. The Mission, it may be noted, acquiesced in certain of these measures, i.e., che resale price and quantity controls, as necessary temporary expedients. is also to be noted, It however, that the full array of controls was continued and even increased after the stock position and market situation had returned to normal conditions.

Other aspects of OMS operations, as observed teams in October, by the evaluation may be of interest. In all operations observed, family groups were encountered, all of whom (except nursing babies) were sold the maximum allotment for that subdivision (which may well have been the case the days of on our visits only. On other days, it is likely that distribution was limited to one allotment per family, either through the MR card require­ zent or a less formal method of control). In some areas separate lines -23­ existed, one for men and accompanying children, one for women and minor childitn. Some lines were disorderly; more often they were well controlled and or.eri.y. In some areas the very poorest of the population Leemed to be in i.ne, many lacking the cash to lift their full allotment. In others, lines contained large numbers of young boys -nd girls who appeare' much better dressed and fed. In most of the subiivisions visited, OMS operations were concentrated disproportion.Rtely on the headquarters town. In some, e.g., Tangail, Kurigram and Thakuriaon, real effort was madp to spread distribution evenly throughout the subdivision.

(b) Impact on Coarse Rice Prices

The OMS program was never successful in constraining coarse rice prices within a set of prescribed limits, either in its early stages or at the new scale of prices that prevailed subsequent to the June 29 Amendment.!/ The extent to which OMS affected rice prices at all is diffi­ cult to determine. As noted in the first evaluation paper, the effect was minimal, both in the early months of September-October, 1978 and again in the March-May, 1979 lean season. The problem in the first instance was mainly operational; inadequate sales relative to the need. During March- May, Food Ministry management was not really tested. Owing to the weak elasticity of demand between rice and wheat at that time, OMS wheat was overpriced relative to the market and as a result little OMS wheat moved ever. as rice prices rose far above target prices.

By June, wheat prices had begun to rise rapidly along with rice prices, but OMS again remained imtest(' this time because of falling stock levels and the virtual suspension of >'5 operations.

Stock levels were back up to 500,000 tons and rising by August. Wheat prices had reached Tk.140-160/md. in biost areas and Tk.200/md. in some. While in most areas prices had started downward by early August, OMS operations lagged slightly the start of the price decline, not getting underway until mid-August. OMS offtakes probably were a factor along with other PFDS distribution in the price decline that continued from that point, more or less uninterruptedly, through September.

1/ The original Agreement specified a coarse rice trigger price at Tk.145/md. with a target price of Tk.150/md. The trigger price was lowered, on rtquest from the BDG to Tk.140/md. in October, 1978. By the terms of the June 29 Amendment to the Agreement, trigger and target prices were changed to Tk.172 and Tk.184/md., respectively. See Table 14 for a complete listing of rice and wheat prices before and after the June 29 Amendment. -24­

(3) OMS Objectives Reconsidered

Experience with OMS to date has led to a basic reconsidera­ tion of program objectives. This seccion examines, in turn; (a) the feasibility and importance of rice price constraint; (b) implications of the system, as it has operated, for private sector involvement; and (c) the recommended alternative role for OMS.

(a) Rice Price Constraint

As noted, the objective of rice price constraint within present target limits has not been achieved. With due regard for the extremely difficult market conditions under which the program has had to operate, it seems doubtful that specific price constraint objectives evez could be achieved. The cross-elasticity of demand between rice and wheat on which the program was predicated has been found to be strong and, indeed, fairly consistent over the long term. At the same time, however, there exists a great deal of variation in the relationship according to season (relatively strong during August-October, relatively weak during April- May); region (highest in the Western and Northern Districts); and the purchasing power of consumers. The difference between rice and wheat (or atta) prices narrowed more than seasonally as overall conditions worsened after mid-1979. Thus, the program cannot be expected to work on the basis of nationwide fixed OMS and rice target prices. A rice price constraint program based on a flexible OMS price, varying according to season and region, may be conceivable in theory, but no observer of program opera­ tions to date could believe it might work in practice.

Our se.ond conclusion with regard to the rice price constraint objective is that it is not the most appropriate target. As long as wheat (or atta) is widely available at a reasonable price, we should not be greatly concerned at the level of rice prices, The purpose of an 0MS program should be to meet the nutritional needs of the rural poor. Wheat (and atta) are at least as suitable for that purpose as rice.

(b) Promotion of Private Sector Activity; the OMS Price

The Agreed Minutes to the original Title III Agreement contained a statement of U.S. Government concern for "the importance of involving the private sector in the purchase, storage and distribution of wheat and rice", and added that the extent to which this objective was met would be a matter for review in the annual program evaluations.

The stated objective was never possible of attainment given the situation in which the OMS price was set at, a level above, or only slightly below, the procurement price for wheat.!/ To be sure,

1/ See Table 14. -25­ wheat dealers - more specifically, the relatively few dealers authorized to lift OMS wheat - could make a modest profit in the mundane business of lifting from BDG godowns and reselling at the controlled OMS price. But given the procurement/OMS price relationship, dealers were greatly in­ hibited from participating in the far more meaningful business of procuring from farmers and storing for eventual resale on the free market. That is, to the extent the OMS price served as a ceiling on the wholesale price of wheat, private dealers had no margin within which to profit by buying from farmers and reselling. Given the small scale and relative ineffectiveness of OMS, the program did not, for the most part, impose such a ceiling, and many dealers were in fact able to continue operating. The point to be noted is that dealers remained in the wheat trade only to the extent that OMS did not function as intended.

Apart from inhibiting private sector activity in the wheat trade, the existence of a procurement price higher than the OMS price is counter to the proclaimed objective of moving away from sub­ sidization of foodgrains. Budgetary costs to the government have been incurred not merely by the fact of an artificially low OMS price, but further by its failure - being a fixed price - to move with the market.

The latter point may be illustrated by reference to the way in which the system operated during the sharp runup in wheat and atta prices beginning in May-June. Wheat sold to dealers at Tk.87/md. was re­ sold to "consumers" in lots of 5 - sometimes 2 - seers at Tk.100/md. In some areas, the retail ceiling was set at Tk.95/md. The retail buyers frequently resold to atta dealers at Tk.130 ­ 140 or more, thereby pocketing one taka per seer (there being 40 seers in a maund) for their time spent in the queue and reselling. Atta dealers would then add an average Tk.20/md. for the cost of crushing, and sell the finished product at anywhere from Tk.150 to 200/md. Thus, the ex-godown to retail atta margin varied from 63 to over 100 Taka per maund (150 ­ 87 - 63, or 200 - 87 = 113). Had the atta dealers been permitted to b-iy directly from the godown, they could have realized a Tk.20/md. profit on the basis of an OMS price of Tk.130 to 180/md., depending on the market at a given time. BDG coffers would have benefitted directly from the additional 43 to 93 Tk./md.

Thus, thelow fixed price of OMS has had three pernicious effects: (1) it has operated to exclude private dealers from the procure­ ment, storage and handling of wheat; (2) it has led, almost inevitably, to the evolution of OMS into yet another ration system; and (3) it has deprived the BDG of revenues that instead have gone to unnecessary inter­ mediaries. We may add here that the artificially low OMS price has led inevitably to inequities and favoritism in the distribution system.

(c) Food Security; OMS When Needed

It may be answered in response to the above (and indeed was pointed out in the June 1979 Evaluation Report) that at certain times (notably April-May) a low OMS price is/will be necessary if OMS wheat is -26­ to move at all. The answer to this point is implicit in what has been said in Section (3)(a) above. The point is to have a widely available nutritious foodgrain. If OMS wheat is reasonably priced (i.e., at close to the procure­ ment price), but does not move because market prices are still lower, tile end is served. Under such conditions, and provided the poorest groups are served by Relief, FFW and MR A card holders, it should not matter that OMS sales are not made.

It is here that we make the link between OMS and the need for greater foodgrain security, as spelled out elsewhere in this papp;c. That portion of Title III wheat not earmarked for MR would be available for OMS or, to the extent not needed for OMS, put into a stabilization reserve for use in future contingencies. The detailed recommendations appear in Sec­ tion IV.G below.

E. Use of Local Currency Proceeds - Loau Foregiveness

Terms - The Title III Agreement specified in accordance with Section 305 (a) and (b) of PL 480 that the BDG would set up a separate, special account within 6 months of the first disbursement by CCC. Local currency proceeds generated by the sale of the eligible MR and OMS Title III wheat were to be deposited in the account, and, in t rn, funds from this special account would be disbursed to the agreed Food )r Development programs listed below:

(1) Rural Development Sector: Taka Millions Actual Agreed Attribution (a) Construction of Thana Training and Development Center Complexes 20 -0­ (b) Construction of Thana Workshops and Godowns 9 -0­

(2) Agriculture Sector:

(a) Establishment of Agriculture Training Institutes 17 -0­ (b) Shallow Tubewell Irrigation Project 40 -0­ (c) Construction of Fertilizer Godowns 25 -0­ (d) Procurement and Distribution of Chemical Fertilizer 700 227 (e) Rural Development Extension Project 20 -0­

(3) Health and Population Sectors:

(a) Construction, Equipping and Operation of Family Welfare Centers 92 -0­ (b) Zero Population Growth Project 3.5 -0­ Xc) Thana Health Complexes 270 -0-

TOTAL 1,196.5 227 -27-

These programs were chosen for their emphasis on "directly improving the lives of the poorest" and increasing their "capacity to participate in the development of their country". Upon certification that the proceeds had been used for these purposes, the USG would apply the dollar equivalent of such disbursements against the BDG's Title I loan repayment obligation, and, hence, "forgive" that portion of the PL 480 loan.

The agreement required that the BDG furnish a quarterly report of the deposits to and disbursements from the special account along with a description of the activities/projects for which the disbursements were made. The agreement also called for the BDG to complete an annual report on the Food for Development program, including a review of all operations through September 30 of each year. Copies of the annual report were to be submitted to USAID by November 1.

Findings

(1) As noted above the special account was to be opened within 6 months of the first CCC disbursement. This first CCC disbursement occurred shortly after the agreement was signed, and the first wheat arrived in November, 1978. In March the BDG notified USAID by letter that the account had been opened, and added that accumulated proceeds amounted to TK.167,378,119 The letter did not explicitly state that the deposits had been made. On October 15, 1979, the BDG wrote another letter which stated that (a) during BDG fiscal year 78/79 52,745 tons of wheat were sold through OMS and the taka proceeds, TK 122,036,000, had been deposited in the special account; (b) during the same period 78/79 30,000 tons of Title III wheat were distributed through MR and TK 62,878,000 was deposited in the special account; (c) during the first two months (July-August) of BDG fiscal year 79/80 TK 62,198,810 were generated from OMS sales (1,770 tons sold in July and 24,329 tons in August) and these funds were deposited in the special account. Thus according to this letter a total of TK 247,112,810, or approximately $16,500,000 equivalent, was reported by the BDG to have been deposited to the Title III special account as of October 15, 1979.

The letter noted also that the Ministry of Finance had "allocated" TK 226,783,000 to the project for "Procurement and Distribution of Chemical Fertilizers" in BDG FY 78/79. Again the letter did not state explicitly that the TK 226,783,000 came from the speciai account's TK 247,112,810 and was disbursed to the project. Subsequent meetings with the Ministry of Finance revealed that the BDG intended to make special account disbursements against "budget allocations" which represent expenditures already made against the FY 78/79 fertilizer project. Further the BDG expects the FY 79/80 fertilizer allocation/expenditure to exceed TK 1,000,000,000, and the BDG would app-y Title III proceeds against this amount.

In a follow-up letter USAID informed the BDG that in order for the USG to certify loan forgiveness the BDG would have to provide specified documentary evidence that the proceeds had been placed in the special account. -28-

The USG would also require specific evidence that the funds have been trans­ ferred from the special account to the project account and used accordingly. The letter also reminded the BDG that special account quarterly reports were overdue, as well as the 1979 PL 480 annual report.1 Finally, the letter informed the BDG that the terms of the agreement permit the BDG to deposit to the special account proceeds of more than 30,000 tons of FY 78/79 MR wheat as the BDG had done.

The current agreement should be amended to clarify loan forgiveness provisions. Specifically, the amendment should state that forgiveness will occur after the Finance Ministry has provided USAID with documentation that (a) funds have been deposited in the Special Account, (b) funds have been transferred from the Special Account to the project account, and (c) funds have been disbursed (i.e., vouchers paid) from the project account. The amendment should also specify that this documentation can be submitted to USAID at any time but not less than quarterly.

(2) Attribution of all Title III proceeds for the fertilizer project certainly falls within the TK 700 million noted in the agreement. However, the fertilizer project receives AID support in the form of dollar grants (roughly $25 million in US FY 79 and $30 million FY 80) to finance the import of fertilizer. The counterpart of these dollar grants is retained by the BDG for unspecified budgetary purposes. The fertilizer agreement itself contains no provision concerning specific allocation of the funds to other aspects of the fertilizer project. However, if one assumes that the AID­ financed fertilizer proceeds should be considered as supporting the fertilizer project, then one might conclude that Title III proceeds represent double USG funding of the project. One might counter by stating that Title III currencies could instead be applied to the subsidy on non-AID funded fertilizer imports, but a large portion of these imports are financed by other donors on grant or concessional basis, and, therefore, Title III would still be double funding. Thus, unless donor financed fertilizer proceeds plus Title III proceeds are less than the total BDG fertili7er project budget allocation (i.e. TK. 1,180,000,000 in FY 78/79 and an estimated TK 1,094,662,000 in FY 79/80) then a case could still be made that Title III proceeds may be double financing the project and, therefore, are not strictly "additional".

However, AID/W has determined that for purposes of this agreement the Title III additionality legal requirement is satisfied by the OMS feature of the FFD proposal. The additionality issue is therefore moot.

1/ Although an annual report was technically dui on November 1, 1978, given the very limited time between the signing of the agreement (August) and the period to be covered (through September 1978) the fact that the report was not submitted is not a substantive issue. -29-

is within the spirit Nevertheless, it is questionable whether it all its Special Account the original agreement for the BDG to disburse of none for the other agreed cover the fertilizer subsidy and reserve funds to allocation in the development projects. To avoid such a disproportionate to exclude further Special future, the current agreement should be amended project, and, instead require a Account disbursements to the fertilizer against the other agreed development more even distribution of disbursements projects. by the Title III (3) The total funds which may be generated the TK 1,196 million for which imports are likely to substantially exceed when the fertilizer project is projects have been identified. However, 496.5 million worth of projects dropped from the list of projects, only TK therefore imperative that will remain on the eligible list. It is to select new FFD projects discussions be initiated as soon as possible amended agreement. The Mission and add them to the eligibility list in the projects which are directly has determined that the BDG and USG will select or improving program admini­ related to increasing agricultural production food policy unit, the crop fore­ stration (e.g. economic research of the casting system, agricultural research, etc.). to projects which are (4) Attribution of the Title III proceeds Plan, while not strictly addi­ already included in the BDG's Development release of the budgeted taka to tional, do serve to obtain timely and full their chances for success. Delays priority projects, and thereby enhances a common and very real constraint in the release of budget allocation are and Title III would serve as an with BDG funded development projects, incentive for the BDG to reduce those delays. the current list of Nevertheless, while reviewing and expanding should strive to identify for the next amendment, the BDG and USG projects tangible beyond what otherwise projects and activities which add something would have been achieved. to date and the (5) Given the experience with this project the BDG need greater focus in legislative intent, both the USG and apparent To date these considerations discussions on local currency use. future but in the future local have been subordinate to food policy issues, perhaps even greater attention. currency use will have to receive equal,

F. Self-Help Measures and local currency use In addition to the price incentive, OMS, retained and expanded of the Title III Program, the Agreement aspects I Agreements dealing with provisions of previous PL 480 Title upon many production, food distri- Government policies in the areas of agricultural of these provisions appeared in the bution and employment creation. Most In some cases, the Self-Help "Self-Help" portion of the Agreement. restatement of policies stated provisions were themselves basically a to the furtherance of the incentive elsewhere in the Agreement as essential -30­ price and OMS objectives. The following is a listing of these measures:

(1) The BDG would establish a food policy unit which will accord high priority to the task of developing policy options with respect to the rationalization of agricultural, price and food distribution policies;

(2) The BDG would shift the PFDS toward the more vulnerable groups within the population. Specifically, there should be no further growth in offtakes in categories other than MR, FFW and OMS. MR offtakes would be limited to A cardholders (the poorest of the four MR categories) only;

(3) The BDG would reduce the subsidy and/or the quantity of the ration in the Statutory Ration (SR) outlet of the PFDS. It would, in particular, raise the ration prices of rice and wheat in the urban areas and increase the proportion of wheat to rice in the overall offtake pattern;

(4) The BDG would encourage the active participation of private dealers in PFDS food procurement and (through OMS) distribution;

(5) The BDG would ensure an adequate margin between procurement prices of rice and wheat and farmers' costs of production in order to promote adoption of HYV varieties of foodgrains, especially among small and sharecropper farmers;

(6) Procurement prices for the Spring, 1979 boro rice crop would be announced prior to the sowing season;

(7) The number of fertilizer retail outlets would be expanded;

(8) The BDG would expand cash based rural works programs which generate rural purchasing power.

The BDG was further charged with providing the U.S. Government an annual "Self-Help Report", providing a review of progress toward achieve­ ment of these objectives. The first such report, due Nov. 1, 1979, was received on Dec. 5, 1979. It was unsatisfactory in content, providing only partial information on some of the itmes. The Mission intends to request a more detailed report.

The following is offered as a brief summary of BDG progress toward these objectives:

- Item (1): The BDG Report stated simply that the food policy unit had been established. It is not clear, however, whether this unit is the full-time, professionally staffed, independent group that the Missior has had in mind. The BDG will be asked for clarification;

- Item (2): The SR, Priorities, Large Employers and Flour Mills categories have grown, in some cases substantially, in absolute terms during the period of the agreement. However, they have grown less than MR, OMS and FFW. The former categories, SR et al, have been reduced as -31­ a percentage of total offtakes from 65% for the period January - June 1978, to 52% for the period July - October, 1979. In that sense, the BDG has complied with the provision of shifting the PFDS toward more vulnerable groups; however, it has not halted the growth of the designated categories. The reduction of domestic production due to drought during this period may be considered an extenuating circumstance.

- Item (3): The BDG has reduced neither the subsidy nor the quantit, of offtakes under SR. It did raise the ration price of rice and wheat but not to the same extent as procurement prices were increased. The proportion of wheat to rice in the overall offtake pattern has increased significantly. However, this is due primarily to The extraordinary growth of wheat offtakes compared to the smaller increase in rice rfftakes.

- Item (4): As discussed in Section IV.C, given the relationship of procurement and OMS prices prevailing throughout the period under review, the participation of private dealers in procurement during the period under review has more likely decreased than increased.

- Item (5): Discussed in Section IV.C. Largely complied with.

- Item (6): As discussed in Section IVoC, this provision has not been complied with.

- Item (7): This item appears to have been largely complied with. Further information will become available through the reporting require­ ments of the Fertilizer Distrib..cion Project.

- Item (8): The extent of compliance is not known. -32-

G. Recommendations

(1) For the Amendment to the current Agreement

(a) Continue procurement operations (rice and wheat) throughout the year, both directly through Government procurement points and indirectly through dealers and other intermediaries. The procurement price for the 1980-81 aman season is to be announced, as a result of joint U.S.-BDG consultations, not later than June 1, 1980.

(b) Establish a margin between the procurement price of wheat and the OMS price sufficient to cover the storage and handling costs of private dealers engaged in procurement operations. The appropriate margin is e timated at 10 Taka per maund which on the basis of the present procurement price for wheat implies an initial OMS price of Tk. 120/md. There will be no resale price restriction on wheat, whether procured from farmers or from BDG godowns at the OMS price.

(c) In accordance with the food security objective of assuring a nutritious foodgrain at a reasonable price, decisions to sell OMS wheat should be based on wheat or atta prices. Atta is preferred because it has a more active, definable market than whole wheat. The margin between the OMS price and the atta trigger price should be sufficient to cover: (1) handling costs of dealers purchasing wheat from BDG godowns at the OMS price, plus (2) the cost of processing wheat into atta. Illustratively, the initial atta trigger price might be Tk. 150/md. and the procurement/OMS pricing structure as follows:

Tk./md.

Procurement price 110

Plus: Cost of storing and handling procured wheat 10

Equals: OMS price 120

Plus: (1) Dealers' margin for storing and handling OMS wheat about 15

(2) Cost of processing wheat to atta about 15

Equals: Atta trigger prIce about 150 -33­

(d' OMS wheat would be sold in any subdivision when the market price of atta reached the atta trigger price. The OMS price would subsequently rise with the market price of atta. The movement need not be continuous. Rather, upward adjustments in the OMS price might be made for each 10 Take move in the atta price, the point being to provide an OMS price in reasonable alignment with the market, thereby eliminating the need for resale price and volume restrictions.

(e) Summarizing the specific recommendations with respect to the proposed new OMS system:

1. Eliminate the trigger and target prices of rice (along with the rice price constraint objective);

2. Establish a trigger price for atta and sell OMS wheat in any subdivision when the trigger price is exceeded;

3. If prices continue to move above the trigger, raise the trigger price accordingly;

4. Eliminate dealer licensing requirements and make positive efforts to open sales to atta wholesalers and retailers. As one means of promoting this objective, reduce the range of limits on lot sales from the present 40 to 200 maunds to 5 and 30 maunds;

5. Eliminate resale pricr and volume restrictions on ONS;

6. Prohibit OMS to flour millers.

(f) Raise the ration prices of rice and wheat in order to reduce subsidy costs.

(g) Local Currency Use:

1. Permit loan forgiveness upon proof of disbursement from project accounts;

2. Eliminate the Fertilizer Project from the list of projects qualifying for loan forgiveness.

3. Reach agreement on a new list of projects directly related to increasing agricultural production or improving agriculture program administration. (h) Supply and Attribution of Wheat:

The original Title III agreement provided 800,000 tons of wheat during U.S. fiscal years 1978-80, of which 600,000 tons during FY 1978 and 1979. This amount was in turn reduced to 542,000 tons as a result of the -34­

P.L.-480 budget limitations and rising wheat costs during 1979. As of Nov. 30, 1979, the BDG had attributed about 175,000 tons to MR and OMS (30,000 tons for MR and 145,000 tons to OMS) leaving 367,000 tons of the first two tranches unutilized.

The amendment obligating the remaining portion of the original 800,000 tons (estimated at 258,000 tons) should specify that up to 50 per cent of the total 800,000 tons may be resold for MR, with some measures taken to assure improved distribution to the poor and to increase the proportion of distribution outside of the major towns to the more distressed rural areas. The remaining 50 per cent may be sold through OMS as the need arises, under the conditions outlined above, or to the extent not needed, added to BDG stocks.

Apart from the original 800,000 tons in the Title III agreement, the USG has programmed an additioaial 142,000 tons for FY 1980 (for a total FY 1980 allocation of 400,000 tons). The amount should be added to the original 800,000 tons, making a new total of 942,000 tons. The additional 142,000 tons will be added to stocks or used for OMS sales under the conditions outlined above.

In summary:

Availability Uses

Agreed, FY 78 and 79 600,000 Less: reduction owing to price increases -58,000

Net FY 78 and 79 542,000

From FY 80 to complete Stocks or original agreement 258,000 MR OMS Maximum Minimum Sub-total original agreement 800,000 400,000 400,000

Additional in FY 1980 142,000 142,000

Total FY 78 - 80 as Amended 942,000 400,000 542,000

(i) Provided that the terms of this Agreement have been complied with, it will be understood that any of the Title III wheat not used, (i.e., attributed to MR or sold through OMS), shall eventually be treated as having been sold, and that amount of the loan written off. -35­

(2) For The Next F.L.-480 Agreement

(a) A five-year Agreement involving 400-500,000 tons per year is tentatively envisioned;

(b) The new Agreement would incorporate the major goals and objectives set forth in this report, and at least most of the specific changes to be incorporated in the proposed Amendment to the current Agreement.

(c) A principal objective would be the establishment of a Stabilization Reserve to be used for OMS, as needed. To the extent the Reserve should be used for other than prescribed purposes, the Agreement would revert to a Title I basis.

(d) Agreement will be reached on a list of projects to be financed by proceeds of local currency sales made in accordance with the terms of the Agreement. The projects should be directly related to the increase of agricultural production or to the improvement of agriculture program administration.

(e) The Mission will prepare for delivery in April 1980 a Report containing detailed recommendations for the next P.L.-480 Agreement. -36-

Table VI. G - 1

Recommended Amendments as Compared with Original Provisions

Existing Title III Agreement Recommended Amendments

1. The internal procurement price of 1. The procurement price for the aman rice should be Tk. 158, and 1980-81 aman season is to be the BDG should attempt to procure announced not later than June 1, 600,000 tons of foodgrain in FY 79. 1980. The exact price and the procurement target will be deleted from the agreement.

2. The OMS sale price is fixed at 2. An initial OMS sale price of Tk. 87/- per maund. Tk. 120/- per maund. This price will move up with the market price of wheat and atta.

3. The OMS wheat sale price is based 3. The OMS wheat sale price is based on a trigger market price of on an atta trigger price. coarse rice.

4. The trigger and target price of 4. Both the trigger and target price rice were fixed at Tk. 172/- and of rice will be eliminated. A Tk. 184/- per maund, respectively; trigger price for atta will be i.e. if the price of coarse rice in established and OMS wheat will be a subdivision exceeds Tk. 172/- per sold in any subdivision when the maund OMS wheat is to be sold, and trigger price is exceeded. The the price limit within which coarse atta trigger price will move rice prices are to be contained is upward with the market, but there Tk. 184/- per maund. will be no target or ceiling price for atta.

5. OMS wheat is sold only to the 5. OMS will be opened to unlicensed licensed dealers. dealers, including atta wholesalers and retailers.

6. OMS wheat is to be sold to a 6. OMS wheat is to be sold to any licensed dealer in any quantity above dealer in lot sizes of 5 to 30 a minimum sized lot of 10 maunds to maunds. 200 maunds.

7. The retail price of OMS wheat will 7. There will be no price restrictions not exceed Tk. 100/- per maund. on the resale price of OMS wheat.

8. Up to 400 maunds of OMS wheat can 8. There will be no sales of OMS wheat be sold to flour millers in one to flour millers. transaction. Annex A 1

Procurement Impact

In a non-coercive procurement system farmers' sales to the Government are a function of the size of the crop and the procurement price. Efforts to increase procurement by raising the procurement price during poor harvest years can be expected to drive up market prices. The effect will be most pronounced after a 2 to 3 month lag as the induced shortage begins to be felt. The table below shows the estimated aman crop, aman procurement, the procurement price, and market price before, during, and after the aman procurement program during the periods 1971/72 through 1978/79:

Aman Production and Procurement and Coarse Rice Prices

1971/72 - 1978/79 (Long Tons) Proc. (Taka per Maund) Price Market Year Production Procurement (TK.md.) July-Nov. Dec. May-July

1971-72 5,695,000 5,842 NA 41.6 39.0 61.4 1972-73 5,587,000 10,066 NA 65.8 60.7 88.3 1973-74 6,699,000 70,862 72.6 81.2 77.6 132.3 1974-75 6,000,000 127,860 120.0 184.8 176.5 2(14.6 1975-76 7,045,000 340,424 120.0 144.9 104.8 lu4.8 1976-77 6,905,000 302,170 120.0 103.3 100.0 135.0 1977-78 7.423,000 500,765 134.0 135.0 124.1 135.0 1978-79 7,000,000£/ 199,567 134.0 132.0 131.9 199.0 Note: (e) Estimated.

As will be noted, the amount procured has varied with the size of the harvest but variations in procurement have been by less than the changes in crop size would seem to have indicated. In 1974-75 and 1976-77 procurement was high despite reduced crops. Procurement was sharply off in 1978-79, but not by as much as the crop shortfall itself. In 1976-77 the shortage already produced by a reduced aman crop was exacerbated by a marginal reduction in the boro crop (because of water shortages). In the early 1970's the Government occasionally resorted to coercion to get farmers and dealers to sell to the government. Where this occurred sales were higher than would otherwise have been the case, and one would expect subsequent upward pressures or prices also to have been greater. It is clear that if the policy is to maintain an acceptable price level for consumption, procurement in the face of a short crop solely for subse- quent ration distribution is a costly and ineffective measure. Every 1 percent reduction in market supplies occasioned by procurement can be expected to drive prices up by 3-4 percent (based on the estimated price elasticity of demand). The price impact is especially severe given the small proportion of the crop (20 to 30 percent) normally going into commercial channels. Illustratively, in 1978-79 the aman crop was estimated to be about 7.0 million tons and 200,000 cons were procured (2.9 percent of the total crop, or 10-15 percent of the crop available for market). Assuming a price elasticity of -0.3, the price impact A-2 would be an increase of 35 to 45 percent. Procurement of 100,000 tons less would have added (restored) 5 to 7 percent to the supply on the 'rarket, and cut the price impact by half.

To be sure, the price impact of procurement is offset to some extent when the purchased grain is used for ration distribution. However, (1) even if all of the purchased grain is added to ration distribution, the fact that it is sold at lower than market prices means that not all of the original price impact will be offset; and (2) redistribution through the primarily urban-based ration sys;em is not an equitable way in which to restore even a portion of that impact.

In 1977, it was stated as official policy (as subsequently reaffirmed in the PL 480 agreement) that the BDG henceforth would procure sufficient rice and wheat effectively to support the price at harvest (and presumably not manipulate the procurement price to increase the amount procured, in a year of short supply). Despite this general policy commitment, the 1978 Title III Agreement stated that about 550,000 tons would be procured from 1978-79 crop. Had this amount actually been procured by a combination of efforts including raising the procurement price, the market price impact could have been very severe. By procuring an additional 200,00 tons and reducing the market supply by a further 10 to 15 percent, the upward pressure on prices would have increased by an additional 35 to 45 percent. Again, compensating action in the form of increased ration distribution, which goes largely to urban recipients, would provide little relief for the rural poor. A-3

ANNEX TABLE A.1

Foodgrain Procurement, 1958/59-1976/77 (thousands of tons)

Years Ending Volun- Border Advance Mono- i/ June 30 tary Drive Levy Purchase poly Other Total

1959 180.1 12.5 - - - 24.8 196.7 1960 11.6 9.0 - 3.5 - 0.1 24.2 1961 12.2 9.1 - 5.0 - - 26.3 1962 2.1 6.0 - 2.0 - - 10.1 1963 0.1 3.9 - 0.1 - - 4.1 1964 95.3 11.4 - 4.0 14.0 - 124.8 1965 3.4 6.4 - 0.1 2.5 - 12.4 1966 7.9 - 81.8 - 2.2 0.2 92.1 1967 0.9 6.4 0.1 - - - 7.4 i9t8 12.0 9.4 - - 0.4 - 21.8 1969 2.6 6.7 - - 0.1 - 9.4 1970 2.1 4.0 - - - - 6.1 1971 n.a. n.a. - - n.a. n.a. 6.0 - 197 2 - .- 1973 0.1 - - - - 0.1 1974 48.3 16.3 - 5.3-/ 0.7 - 70.9 1975 - - - - - 137.91/ 1976 - - - - - 415.0-3/ 1977 - - - - 315.0-3 1978 - - - - 5504.4/ 1979 - -.- - 353.4-/

1/ Figures may not add up exactly due to rounding. 2/ Disposal and Acquisition Order 3/ Voluntary except for minor quantities procured under the "border drive". 4/ Voluntary

Source: a) Food Ministry data. b) Bangladesh Foodgrain Storage Project II Feasibility Study prepared for the Linistry of Food & Civil Supplies, Republic of Bangladesh, by Carl Bro International A/S and Operations Analysis Corporation A/S in association with The Economist Intelligence Unit Ltd. and Sarm Associates Ltd.., Copenhagen, October 1975. c) Preappraisal Report of the Second Foodgrain Storage Project. Annex B

Estimated Reserve and Storage Requirements

Ideally, reserve requirements are divided into two types: operating reserves which are required to operate the normal government distribution system, and strategic reserves required to deal with unusual production declines. In Bangladesh no such clear distinction has been made.

Table B-1 shows the distribution of rice and wheat by semesters, along with production in the corresponding crop (fiscal) years. Calendar year 1979 may turn out to be a peak year for distribution. During the first six months of 1979, 1,019,000 tons were distributed, slightly below the prior first semester peak of 1,149,000 tons in 1973. During July-October 1979, 949,000 tons were distributed, bringing the total for the first 10 months to 1,968,000 tons. Given the abundant supply now in Government hands, the prevailing high market prices, and the expected late arrival of this year's aman crop, the BDG is likely to continue distribution at high rates and bring the year's total to a new record of about 2.4 million tons (in per capita terms, this would approximately equal the 1973 distribution). Whether domestic production has been so low as to justify this high rate of distribution is open to question.

Table B-1 Production and Foodgrain Distribution, 1972-79

(000 Tons)

Production Distribution

(fiscal years ended June 3n) (calendar years) Rice Wheat Total 1st Sem. 2nd Sem. Year

1972 9,774 113 9,887 668 1,146 1,814 1973 9,930 90 10,020 1,149 930 2,079 1974 11,721 109 11,830 798 939 1,737 1975 11,109 115 11,224 802 897 1,699 1976 12,560 215 12,775 667 637 1,304 1977 11,567 255 11,822 835 981 1,816 1978 12,764 343 13,107 867 777 1 644 / 1979 12,288a / 600A/ 12,888-a/ 1,019 b/ 2:400!

Notes: A/Estimated k/For the period July-October, distribution totaled 949,000 tons, which compares with 542,000 tons for the corresponding period of CY 1978 and 715,000 tons in July-October 1977. B-2

It is difficult to identify a normal pattern of distribution as a basis for evaluation requirements. The crop years 1973/4, 1975/6 and 1977/8 were relatively good ones. Foodgrain offtakes in the corresponding calendar years, 1974, 1976 and 1978 were 1.737 million tons, 1,304 million tons and 1.644 million tons, respectively. In contrast, distribution following the relatively poor crop years of 1973, 1975 and 1977 was 2.079 million tons, 1.699 million tons, and 1.816 million tons respectively. This might seem to suggest that about 300,000 tons more distribution is needed following low versus high production years. However, the sharp surge in prices in 1974/5 would suggest that distribution in the latter part of 1974 was not adequate, nor was it adequate in February, March, April of 1975.

If production in any one year falls one million tons below normal, and it is desired to contain seasonal price increases within 10 to 15 per cent of the normal trend, an additional 800,000 tons must be supplied from imports and stocks.I / Even this situation presupposes an even supply distribution over the entire year. Since this is virtually impossible to achieve, the potential price increase under this scenario may be somewhat greater than the 10-15 per cent indicated.

The most frequent supply fluctuations arise from swings in the monsoon crops, aus and aman. Table B-2 shows for each crop the two largest production declines (not including the war years) occurring during the past 15 years. Taking the mean of the two, potential major crop shortfalls appear as follows (see also Table B-2):

Aus 350,000 tons Aman 850,000 tons Boro 600,000 tons Wheat 75,000 tons

.!/The 200,000 ton decline in supply would mean roughly a 4 per cent decline in combined market and PFDS availabilities. Applying the estimated demand elasticity factor of -0.3,and other things equal, prices would rise by 13 per13T cent (--0= .133). B-3

Table B-2

Production, Trend and Absolute Deviation from Trend (Data in 000 tons)

1964/65 - 1978/9.E Current Trend rate Adjustedb Crop Level Max. decline of increase decline /

Aus 3,300 (1) 396 30 426 (2) 244 274

Aman 7,500 (1) 880 60 940 (2) 699 759 Boro 2,200 (1) 636 150 786 (2) 239 389

Total Rice 13,000 Wheat 600 (1) + 6c/ 100 94

(2) + 40 c/ 60

TOTAL 13,600

Note: A/ Not including the war affected years of 1970-71 and 1971-72. b/ Decline adjusted for trend. .S Smallest year-to-year increases since 1973-74.

Clearly aman outweighs all others combined in potential seriousness, with a prospective decline of 850,000 tons. A bad aman preceded by a bad aus or followed by a bad boro would give a decline from trend of about 1.2 - 1.4 million tons. The record suggests that severe crop shortfalls can be expected to occur at least once every five years.

It is important to note that the most pronounced potential variations for a given crop year particularly affect supplies in March - July of the following year and consequently prices during those periods. Thus, the Government has until about January to arrange increased ship departures for donor imports, or until December or early January to begin purchasing abroad. Given the logistics of ordering, shipping and unloading, every month of delay in the process of crop disaster alert and commencement of arrangements to import adds to reserve requirements above normal operating requirements. Table B-3A / Table B-4

Tons Lowerb/ Normal Crop Crop One Million (000 tons) (000 tons)

Distri- Arri- Procure- Stock Change Distri- Arri- Procure- Stock Change bution vals ment Monthly Cum.c / bution vals ment Monthly Cum.­

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)

0 150 100 0 - 50 - 50 180 100 0 - 80 - 80 N 120 100 0 - 20 - 70 160 100 0 - 60 -140 D 105 100 250 +245 +175 110 100 100 + 90 - 50 J 105 100 100 + 95 +270 130 100 0 - 30 - 80 F 120 100 0 - 20 +250 180 100 0 - 80 -160 M 140 100 0 - 40 +210 200 100 0 -100 -260 A 150 100 0 - 50 +150 210 200 0 - 10 -270 M 140 100 50 + 10 +170 200 300 50 +150 -120 J 140 100 0 - 40 +130 200 300 0 +100 - 20 m 140 100 0 - 40 + 90 200 300 0 +100 + 80 A 140 100 0 - 40 + 50 190 200 0 + 10 + 90 S 150 100 0 - 50 0 190 100 0 - 90 0

a/ Assumes normal distribution of 1.6 million tons and imports of 1.2 million tons. b/ Crop one million tons less for aus and aman; procurement drops to 100,000 in Dec./Jan. from 350,000. S/ Since tables show only cumulative stock changes, starting balance of 500,000 tons should be added to the monthly figures to arrive at monthly stock positions. B-5

In a normal year the Government should have about 500,000 tons of stock ashore at any particular time for routine operations. When stocks fall below this level additional, often expensive, measures are required to move stocks from surplus to deficit areas.

An analysis has been made of the additional stocks required to meet both normal conditions and a one million ton shortfall, assuming as a starting point the normal level of imports, average monthly offtakes, and the approximate average level of procurement of recent years. Table B-3 shows the situation for a normal crop year. For the sake of simplicity, imports are scheduled to arrive evenly throughout the year. In this case, assuming a 500,000 ton minimum stock level (as of Sept. 30), stocks would vary from 500,000 tons to a low of 430,000 tons on Nov. 30 and to a high of 770,000 on Dec. 31.

The second situation is for a poor year in which the production shortfall is one million tons. Procurement falls by 250,000 tons (to a level of 150,000) and distribution rises by 550,000 tons (to 2.15 million). To hold the supply decline to 200,000 tons, imports must rise by 800,000 tons. The magnitude of the crop shortfall is established by December so that increased purchases begin to arrive in April. The effects of the situation are shown in Table B-4. Since the maximum stock decline is 270,000 tons, this much additional stock would be needed to keep from going below the minimum required for operating purposes.

If the alert and initiation of increased purchase were delayed by 3 months as happened in 1978/9, the situation would be the same as in Table B-4 up to March. The result of the delay from that point on is shown in Table B-5.

Table B-5

Crop Decline of One Million Ton and Delayed Alert (000 tons)

Normal Stock Pro- Cum. Distri- cure- Stock Change Change-a / bution Arrivals ment Monthly Cum.

(5) (6) (7) (8) (9) (10) M +210 200 100 0 -100 -260 A +160 210 100 0 -110 -370 M +170 200 100 50 - 50 -420 J +130 200 100 0 -100 -520 J + 90 200 300 0 +100 -420 A + 50 190 400 0 +210 -210 S 0 190 400 0 +210 0

Note: - From Table B-3, Col. 5 B-6 To deal with this situation the Government would need an additional reserve )f 520,000 tons above the normal 500,000 tons working stock level. The consequences of the delay in responding to the crop shortfall are reality of a serious costly in terms of port congestion, costs and higher foodgrain higher administrative prices as a result of having portion of the supply to buy in the rising pericd. Diring the past between May year prices rose $40 per ton and June. Prices at that time were higher than those prevailing in January or February.

Additional Storage to Support Unusually Large Procurement Besides having to provide for the contingency of a large the Government must crop shortfall, be prepared also to deal with unusually large the contingency of an crop and hence to procure heavily announced levels. to support prices at the This may seem paradoxical inasmuch shortfall and unusually as a large crop heavy procurement cannot same time. be expected to occur at The point is, however, that the since firm news of aman crop cannot be expected before January results (with shipments arranged well the beforehand), BDG must be prepared for both.

The BDG needs then to be prepared for the following, as of January: A. Normal Requirements (tons) 3 months Normal Distribution, approximately 500,000 B. Contingencies

1. One million ton shortfall and delayed alert 520,000 Minimum January stock position 1,020,000 2. Large crop in aman/aus and procurement at 800,000 tons + 400 000 / Possible January Stock Position 1,420,000 MT Note: 1/ Procurement in excess of normal.Storage requirement for normal procurement is assumed within the 500,000 ton minimum. If peak storage could reach capacity 90 per cent of capacity, total required for handling these storage contingencies would be 1,580,000 Capacity now tons. is 1.2 million tons or less. B-7

Presumably, if the BDG could develop a dependable, improved crop reporting and alert system, enabling it to take action on additional purchases by late December, and if it could be sure purchase and deliveries could be made quickly, the additional reserve requirement might be cut to 270,000 tons, as shown in the Table B case above. In that case, the tgbl+s 6a~e4fi8pacity requirement would be reduced to about 1.3 million tons ( = 1,300). It would be unrealistic, however, to expect actual development of such a system within the next few years. Meanwhile, the combination of an increasing population and (hopefully) rising real per capita income will have added further impetus to the stock level and storage space requirements.

As an interim measure, heavier arrivals should be scheduled for the March to July period. Heretofore, the focus has been on the August-October lean period. However, in years of major aman production shortfalls, the March-Jul) lean season becomes the more critical period. Earlier scheduling of normal imports can help to provide the needed stability.

Current Situation and Scheduling of Required Imports

The BDG uses the period July 1 - June 30 as its fiscal year as well as its crop year. This corresponds to our marketing year concept, and based on our previous analysis serves as a good planning year basis. Annex Table B-6 provides historical data and projections to June 30, 1980 regarding crop production, PFDS offtakes and PFDS stock positions. Two scenarios have been considered for the 1979/80 crop year ending June 30, 1980.

The first scenario assumes poor crops, only 400,000 tons of imports arriving between January and June of 1980 (1,943,000 tons are expected to have arrived between July 1 and December 31, 1979), and offtakes at 2.4 million tons. Under these circumstances, the PFDS June 30 stock level would be below 500,000 tons. On the other hand, if crop results are somewhat better and all imports currently scheduled (i.e. 622,000 tons) arrive, then a more comfortable position of over one million tons would result. Combined close monitoring of the crop situation, procurement, government offtakes and import scheduling is necessary. If poor weather continues through April of 1980, the import requirement may be as high in the latter half of CY 1980 as it was in 1979. Whatever scenario is used, however, it seems clear that if 400,000 tons of Title III wheat do not arrive before June 30, 1980, BDG stock levels are likely to be at fairly low levels, even if domestic production is good. B-8

Table B-6

Foodgrain Stock Projection, 1979-80 (000 tons)

Low Projection High Projection

Opening balance 214 214 Imports 2,343 2,565 Domestic procurement 300 600

Availabilities 2,857 3,379

Distribution 2,400 2,300

Closing Balance 457 1,079 1

Annex C

Food Distribution by Income Status; Urban vs. Rural Distribution

Mission policy with respect to Public Food Distribution has consistently favored distribution to the poorest groups located mainly in the rural areas. However, except for Food-for-Work (FFW) and Gratuitous Relief (GR), the existing PFDS categories are not clearly divided either as to the relative income status of recipients or to rural vs. urban status. Modified Rationing (MR) serves the rural population, and by the terms of the Title III Agreement is supposed to go only to A ration card holders, i.e. those so poor that they pay no income tax. However, from results of the Mission's sorghum studies and numerous other indications from field sources, it is known that B and even C and D card holders have been receiving MR rations. OMS coverage is mainly rural, but it does serve urban as well as rural areas. As for Statutory Rationing (SR), while its recip­ ients are by definition urban and, therefore, better off on the average than the rural population, many SR recipients are truly poor.

The establishment of policy guidelines and judgment of BDG performance must, therefore, be based on the admittedly imprecise nature Qf the PFDS cate­ gories. Mission policy, as set forth in the Title III and previous Title I Agreements, may be summarized as follows:

- preference is accorded to OMS and MR (A card holders only);

- overall PFDS distribution to categories other than OMS, MR, FFW and GR should not increase over time;

- SR, in particular, should decline over time.

FFW, it may be added is covered under a separate PL 480 Agreement (Title II), and is therefore not explicitly a Title III issue.

Table C-1 shows the offtake trend during fiscal years, 1973-80, with the generally "rural/poor" categories - OMS, MR, FFW and GR - identified separately from the PFDS total:

Table C-1 PFDS OFFTAKES, TOTAL AND RURAL/POOR (000 LT)

(1) (2) (3) (4) (5) (6) (7) TOTAL (6) as % FY TOTAL MR OMS/OMO FFW 1EL.LIEF (2) to (5) TOTAL

1973 2,295 1,592 207 1,799 78.4 74 1,728 528 21 549 31.8 75 1,741 535 161 696 40.0 76 1,655 496 11 47 180 733 44.3 77 1,473 288 59 16o 33 546 37.1 78 1,874 353 6 255 18 631 33.7 79 1,797 311 30a/ 412 60 813 45.2 July-Oct.,1980 949 224 77a/ 131 20 452 47.6

Note: a/ Adjusted to exclude distribution to flour mills. C-2

As Table C-i clearly shows, 1972-73 was a highly atypical year. In the wake of the disruption caused by the war, MR and GR were at levels not to be approached thereafter. From that year onward, GR distributions have held at a basic 1,500 to 3,000 tons per month (mainly for camps and centers for the destitute), except for crisis periods such as 1974-75 and 1975-76 (crop pro­ duction recovered in the latter year, but purchasing power among the popula­ tion remained low enough to warrant high GR distribution). By July - October, 1979, BDG policy had been reoriented, with Rainy Season Food-for-Work becoming the preferred relief channel.

FFW provided an upward push to overall rural/poor distribution from its inception in April 1976. The same cannot be said of OMS, which had been preceded by OMO_! (at an even higher level in FY 1977), and which tended in part to replace MR. For most of the period under review, distribution to the combined rural/poor categories as a percentage of total PFDS rose only during crisis years. FY 79 constitutes something of an exception to this point be­ cause while the drought was well underway during the second half of 1978-79, emergency allocations as a result of the drought did not get well underway until FY 1980 (i.e. after June, 1979).

Table C-2 shows total rural/poor distribution on an index basis expressed both in total and per capita terms:

Table C-2

Rural/Poor Food Distribution, 1973-79

TOTAL INDICES (1974-79 = 100) FY (000 LT) TOTAL PER CAPITA

1973 1,799 272.0 299 74 549 83.0 89 75 696 105.2 110 76 733 110.8 113 77 546 82.6 82 78 631 95.4 92 79a/ 813 122.9 115

NOTE: a/ Adjusted to exclude OMS distribution to flour mills.

Table C-3 shows distribution in the other categories, which may be termed, for convenience, "urban/realtively affluent". SR, identified separately, is by definition entirely to urban recipients. Included in the "other" category are Government Employees, employees of large companies (those with over 50

1/ For Open Market Operations; See Section IV.D(l) C-3 employees), Priorities (the military, hospitals, etc.)!I and flour mills. Recipients under these categories are mainly urban, or - as in the case of the flour mills - serve a relatively well-to-do portion of the population.

Table C-3

PFDS Distribution to Urban/Relatively Affluent FY 73-79 Index (000 LT) (1974-79 = 100)

a/ FY SR Other Total Total Per Capita

1973 465 30 495 47.3 52 74 494 685 1,179 112.8 120 75 471 574 1,045 100.0 104 76 335 586 921 88.1 89 77 377 550 927 88.7 87 78 451 766 1,217 116.4 i1 79 417 567 984 94.1 88

NOTE: a/ Government employees, employees of large companies (over 50 employees), Priorities (military, school teachers, hospitals, etc.) and flour mills.

As Table C-3 shows, the BDG has been making tairly steady progress in reducing these categories of offtakes, especially when measured on a per capita basis. FY 1978 constituted an aberration with offtakes under the urban/relatively affluent categories rising sharply, a phenomenon that can probably be attirbuted to the pre-election climate of that year. By 1978-79, per capita distribution to these categories was back to FY 1976 and 1977 levels.

In short, the BDG has made fairly good progress in reducing both the amount and share of total offtakes going to the urban/relatively affluent categories. The progress made should not, however, obscure the fact that this group still accounts for a share of total PFDS offtakes that far exceeds its share of the total population (only 10 percent of the population are classified as urban).

1/ To be sure, this category includes recipients who can legitimately be considered "essential". 1

Annex D

The Crop Reporting System

Accurate and timely crop forecasts and estimates are needed for monitoring of the production outlook and projecting total foodgran needs; all the more so in a country where the subsistance margin is so narrow. Delays in the informa­ tion system of 2 to 3 months can necessitate much larger storage facilities and reserves (at high cost). The basic operation of the existing statistical re­ porting system has changed little from that described in Glen Fisher's July 1977 report "Report on the Bangladesh Crop Survey Systems Study".

Agricultural crop estimates incorporate both the objective and subjective methods. The objective method employs an area sampling technique originally developed by the British based on cadastral surveys of basi. units called Mauzas. The technique involves a two stage procedure resulting in a 1/1000 sample. Sub samples are observed duri'ng the year and crop cuts taken. In recent years there has been little training or ground control exercised over the data collected, resulting in reduced reliability. The subjective method involves estimates of area and production by union agricultural assistants. These in turn are aggregated at higher levels. They are generally considered to have an upward bias. All estimates, objective and subjective, are reviewed by the Bureau of Statistics. Final official estimates for rice and jute are set by a Reconciliation Board chaired by the Bureau of Statistics and composed of representatives of the Ministries of Jute, Agriculture, Planning and Food. Other crop estimates are made using the objective method only.

Final estimates are not made in a timely manner and few forecasts of crops are available. For example, the final estimate for the last boro crop is not yet available some eight months after completion of the harvest. To avoid crisis situations, it is vital to have early forecasts of crop conditions. Only when crisis conditions can be accurately forecast and handled can concentration be focused on long run problems.

FAO is providing for U.S. Census Bureau technicians to the Bangladesh Bureau of Statistics. Additional support in the agricultural statistics area is to be provided by FAO. There is currently underway an effort to upgrade the system starting with the Dacca District. This has involved the training of personnel in a revised crop cutting technique which reduces the cut to 100 sq. ft., relocates sample plots, measures their area, and improves the monitoring and control of data submitted. Progress has been made, but satis­ factory control has not yet been obtained. Work is shortly to be expanded to another district. It will be some time until the system can be upgraded to cover the entire country.

In short, there is a pronounced need for an objective crop reporting system that can produce reliable statistics in a timely manner. An agricultural statistics project for funcing with Title III local currencies should be con­ sidered for inclusion in the Amendment to the current Agreement. 1

Annex E

OMS and Flour Mill Allocations

OMS was used by the government to substitute for and/or augment distribution to flour mills which were already being supplied under the regular flour mills categories of the PFDS. This is clear from the offtake patterns.

Regular flour mill allocations, which had been over 18,700 tons per month during January-Augu&t, 1978, fell to 14,000 tons per month during September and October, the first two months of OMS. The 9,400 ton reduction was more than compensated for by 24,000 tons supplied to the mills under ONS. In November and December, when OMS operations were nil or virtually nil, regular flour mill allocations rebounded to an average 18,400 tons a month. As OMS resumed during January - May, 1979 regular flour mill allocations again fell to an average of about 14,000 tons per month, but the reduction was mcre than offset by a distribution under OMS averaging about 5,000 tons a month. In June and July, reflecting the abnornially low stock levels at that time, both regular flour mill and OMS allocations were far below normal. During this time, the June Amendment to the Title III Agree­ ment was signed, stipulating among other things that henceforth sales to flour mills under OMS would not f;xceed 20 percent of total OMS in any subdivision. In fact, sales to flour mills under OMS were 9 per­ cent of overall OMS durin- the period July - October. Thus, the BDG has adhered to this restriction. TABLE 1

FOODGRAIN PRODUCTION AND IMPORTS 1964-65 to 1978-79

(In Thousand LT)

All Aus rice Aman rice Boro rice rice Wheat Total Imports

1964-5 NA NA NA 10,334 34 10,368 345

1965-6 2,918 6,799 618 10,335 35 10,370 923

1966-7 2,674 5,919 831 9,424 58 9,482 1,100

1967-8 3,069 6,812 1,114 10,995 58 11,053 1,019

1968-9 2,673 6,870 1,612 11,165 92 11,257 1,119

1969-70 2,963 6,950 1,903 11,816 103 11,919 1,547

1970-1 2,344 5,912 2,192 10,967 110 11,077 1,146

1971-2 2,341 5,695 1,738 9,774 113 9,887 1,688

1972-3 2,273 5,586 2,070 9,930 90 10,020 2,825

1973-4 2,802 6,699 2,220 11,721 109 11,830 1,667

1974-5 2,859 6,000 2,250 11,109 115 11,224 2,558

1975-6 3,230 7,045 2,286 12,560 215 12,775 1,445

1976-7 3,011 6,906 1,650 11,567 255 11,822 798

1977-8 3,103 7,422 2,239 12,764 343 13,107 1,609 E 1978-9 3,288 7,000 2,000 12,288 600 12,888 1,147

E = Estimated

Sources: BDG reports for 1964-5 to 1977-78; BDG, FAO and USAID for 197E-79. Table 2

AREA UNDER MAIN CROPS (Million Acres)

Crops 1969/70 1970/71 1971/72 1972/73 1973/74 1974/75 1975/76 1976/77 1977/78

Rice 25.48 24.50 22.98 23.80 24.41 24.20 25.53 24.42 24.78 Aus ( 8.46) (7.89) (7.42) (7.24) (7.68) (7.86) ( 8.45) (7.95) (7.82) Aman (14.84) (14.18) (13.37) (14.12) (14.13) (13.47) (14.24) (14.36) (14.26) Boro (2. 18) (2.43) (2.19) (2.43) (2.60) (2.87) (2.84) (2.11) (2.70) Wheat 0.30 0.31 0.31 0.30 0.30 0.31 0.37 0.40 0.47 Other Cereals 0.28 0.30 0.25 0.18 0.24 0.24 0.23 0.22 0.22 Pulses 0.91 0.92 0.89 0. 78 0.70 0.76 0.75 0.82 Oilseeds 0.78 0.76 0.69 0.68 0.63 0.69 0.70 0.69 Rape and Mustard (0.54) (0.53) (0.47) (0.47) (0.44) (0.48) (0.48) (0.48) (0.51) Til and Linseed (0.16) (0.15) (0.15) (0.15) (0.14) (0.16) (0.17) (0.16) ( .. ) Groundnut (0.08) (0.08) (0.07) (0.06) (0.05) (0.05) (0.05) (0.05) (0.06) Jute 2.46 2.20 1.68 2.21 2.02 1.42 1.28 1.60 1.81 Spices 0.42 0.40 0.39 0.38 0.35 0.38 0.38 0.37 Sugarcane 0.40 0.40 0.35 0.32 0.36 0.38 0.33 0.36 0.38 Potato 0.21 0.21 0.19 0.19 0.20 0.23 0.24 0.19 0.22 Sweet Potato 0.18 0.18 0.17 0.16 0.15 0.16 0.18 0.17 0.18 Fruits and Vegetables 0.64 0.63 0.58 0.58 0.59 0.61 0.63 0.63 Cotton 0.03 0.02 0.03 0.04 0.02 0.02 0.02 0.01 0.01 Tea 0.11 0.11 0.11 0.11 0.11 0.11 0.11 0.10 Tobacco 0.11 0.11 0.10 0.12 0.11 0.11 0.12 0.17 0.14 Others 0.53 0.50 0.39 0.40 0.31 0.50 0.40 0.28a

Total,all Crops 32.84 31.53 29.11 30.24 30.50 30.12 31.27 30.43a/ .

Net Cropped Area Inclu­ ding Currcnt Fallow 22.49 22.48 22.47 22.47 22.48 22.7_Zfl_ 22.56a/ .. Cropping Intensity(%) 146 140 130 135 136 134 139 .. . = Not available. a/ Provisional estimate Source: Bangladesh Bureau of Statistics. Table 3

Rice Yields, 1965/66-77/78

Average Yield Per Acre in Maunds

Aus Ama n Boro Average

1965/66 10.8 12.6 14.8 12.1 1966/67 10.4 11.5 16.2 11.4 1967/68 10.2 12.6 19.7 12.2 1968/69 9.5 13.0 21.8 12.6 1969/70 9.5 12.7 23.7 12.6

5-Year Average 10.1 12.5 19.9 12.2

1970/71 9.9 11.4 24.6 12.2 1971/72 8.6 11.9 22.6 11.6 1972/73 8.5 10.8 23.2 11.4 1973/74 9.9 12.9 23.3 13.1 1974/75 9.9 12.1 21.3 12.5

5-Year Average 9.4 11.8 23.0 12.2

1975/76 10.4 13.5 22.0 13.4 1976/77 10.3 13.1 21.3 12.9 1977/78 10.8 14.2 24.9 14.3

Source: Agricultural Statistics Wing, Bangladesh Bureau of Statistics. Table 4

Fertilizer Offtake, Fertilizer/Rice Price Ratio and Grain Production 1965/6 to 1979/80 (1) (2) (3) (4) (5) Fertilizer Offtake % Change Fertilizer/rice Grain Production (in thousands of tons) from prev. Price Ratio (in millions of tons) year Nutrient Production Tons Tons

1965/6 50 106 1.04 10.37

6/7 77 160 +51 1.04 9.48

7/8 98 209 31 1.21 11.05

8/9 105 223 7 1.01 11.26

9/70 130 277 28 1.05 11.92

70/1 143 306 11 1.14 11.08

1/2 114 244 -17 .83 9.89

2/3 179 384 15 .91 10.02

3/4 177 379 - 1 .88 11.83

4/5 130 270 -29 .83 11.22

5/6 2i2 439 60 1.66_a/ 12.76 6/7 235 507 14 2.1 11.82

7/8 330 711 40 2.2 13.11

8/9 345 741 4 2.00 12.88

Notes: (a) Estimated (b) Relative to procurement price. Actual ratio was more favourable. Table 5

BADC LOWLIFT PUMP PROGRAM (1960-1979)

Average Area Irrigated Total Area Number of Pumps Cusecs per Cusec Irrigated Years in Operation per Pump (Acres) (Acres)

1960/61 1,267 . . 64,528 1961/62 1,543 .... 98,163 1962/63 2,024 1.70 38.5 133,043 1963/64 2,456 1.85 34.1 156,281 1964/65 2,238 1.95 30.0 131,129 1965/66 3,420 2.05 24.6 173,553 1966/67 3,990 2.14 26.4 225,105 1967/68 6,558 2.08 23.3 317,903 1968/69 10,582 1.89 20.7 424,799 1969/70 17,844 1.84 19.6 639,000 1970/71 24,454 1.85 18.3 820,074 1971/72 24,254 1.84 19.4 864,427 1972/73 32,924 1.84 20.3 1,230,468 1973/74 35,343 1.84 20.1 1,330,810 1974/75 35,427 1.86 20.0 1,296,960 1975/76 36,376 .. 1,431,000 1976/77 28,224 .. 1,034,328 1977/78 36,735 .. 1,355,138 1978/79(target) 40,000 .. 1,460,000

Source: Bangladesh Agricultural Development Corporation.

.. = Not available. Table 6

BADC SHALLOW TUBEWELLS PROGRAM

Avera ge Number in Area Irrigated Irrigated Area Year Sunk Operation a! (Acres) (ac. per well)

1972/73 1,808 994 3,560 3.6

1973/74 1,142 2,293 11,854 5.2

1974/75 911 3,605 23,470 6.5

195/76 1,083 2,162 12,894 6.0

1976/77 1,627 3,045 17,707 5.8

1977/78 3,452 6,517 58,850 9.0

TOTAL 10,023

a/ Under a subsidy scheme which was outside the usual program of BADC but implemented under the supervision of BADC technical personnel, 793 tubewells were sunk and 672 commissioned by June of 1972. These tubewells are included here.

Source: Bangladesh Agricultural Development Corporation. Table 7

BADC DEEP TUBEWELLS PROGRAM

Average Number in Area Irrigated Irrigated Area Year Sunk Operation (Acres) (ac. per Well)

1967/68 106 .. 4,117

1968/69 292 .. 16,080

1969/70 662 .. 32,119

1970/71 51 734 32,070 43.7

1971/72 221 906 29,330 32.4

1972/73 872 1,237 37,776 30.5

1973/74 1,778 1,493 61,456 41.2

1974/75 3,178 2,699 117,664 43.7

1975/76 1,656 3,828 153,747 40.1

1976/77 634 4,455 161,263 36.2

1977/78 1,271 7,453 338,474 45.4

TOTAL 10,721

= Not available.

Source: Bangladesh Agricultural Development Corporation. Table 8

FOODGRAIN PRODUCTION, IMPORTS AND PER CAPITA CONSUMPTION Selected Years, 1966/67 - 1977/78 (Data in millions)

Ave rage 1966/67-1969/70 1975/76 1076/77 1977/78

Gross Production (LT) 10.93 12.77 11.82 13.11 Less: 10% for feed, seed and wastage -1.09 -1.28 -1.18 -1.31

Net Production 9.84 11.49 10.64 11.80

Imports 1.20 1.44 .80 1.61

Change in Stocks (- = addition) NA -. 09 .45 -.20

Less: losses in distribution -. 07 -. 07 -. 07 -. 07

Total available for consumption 10.97 12.77 11.82 13.14

Population 67.90 80.40 82.70 85.00

Apparent per capita consumption: (oz. per day) 15.9 15.6 14.9 15.2

Sources: See Table 1; also IBRD (for population) Table 9

WAGES AND PRICES 1961/62 to 1978/79

CLI Unskilled of Medium Industrial Agricultural Industrial Rice Wages Wages Workers Prices Year (1969/70=100) TK/day 1969/70=100) (TK/md)

1961/62 80 2.Z1 87 30 6Z/63 89 Z.45 89 32 63/64 97 2.41 96 Z9 64/65 91 2.47 91 30

1965/66 9Z Z.81 97 36 66/67 92 2.90 95 46 67/68 100 2.99 10Z 42 68/69 94 2.88 9z 46 69/70 100 2.97 100 45

1970/71 105 3.06 104 45 71/72 106 2.76 108 57 72/73 130 3.75 193 89 73/74 173 5.53 268 120 74/75 221 7.24 430 244

1975/76 Z44 7.28 364 154 76/77 252 7.49 354 133 77/78 280 9.35 419 169 78/79 346 11.26 457 184

Sources: Statistical Digest of East Pakistan, No. 5, 1968, p. 230 Statistical Yearbook of Bangladesh, 1975 and 1979 Economic Indicators of Bangladesh, August 1979 Mohiuddin Allmigar, "Some Analysis of Income, Consumption Savings & Poverty, " The Bangladesh Development Studies, VII, No.4, October 1974, p. 7 5 8 Table 10

REAL INDICES OF WAGE & RICE PRICES 1961 /62 - 78/79

Industrial Agricultural Medium Year Wages Wages Rice Prices

1961/62 92 86 77 1962/63 100 93 80 1963/64 101 85 67 1964/65 100 91 73

1965/66 95 98 82 1966/67 97 103 108 1967/68 98 99 91 1968/69 102 105 i1 1969/70 100 100 100

1970/71 101 99 96 1971/72 98 86 117 197Z/73 67 65 102 1973/74 65 70 100 1974/75 51 57 136

1975/76 67 67 94 1976/77 71 71 84 1977/78 67 75 90 1978/79 76 83 89

Source: See Table 9 Table 11

PFDS OFFTAKES, 197Z-79 (Calendar Years - Long Tons) Jan. -Oct. Category 1972 1973 1974 1975 1976 1977 1978 1979 (10 months) Statutory 353,866 478,254 532,783 407,018 283,232 468,019 411,756 391,960

Modified 1,118,741 1,138,308 556,078 625,663 251,944 395,228 228,175 403,405

Priority - - 90,613 108,651 85,189 113,627 114,306 70,094

G. Relief 466,229 91,307 116,390 90,4Z3 172,338 25,543 39,842 40,146

Others 283,429 370,852 - - - - - Govt. Employee 223,675 293,795 177,029 304,803 339,525 381,694

Large Employee - - 79,938 106,472 58,373 82,187 76,520 77,507

Flour Mills - - 133,939 157,221 160,356 196,273 202,909 146,616

Food For Work - - - - 68,308 203,360 205,352 337,930

OMO - - 48,749 26,571 1,124 17,752

OMS ------24,564 101,669

Total Rice: 729,455 152,974 131,193 385,811 563,261 771,237 573,679 539,503

Total Wheat: 1,808,881 1,925,747 1,606,223 1,403,432 742,257 1,044,374 1,070,394 1,429,270

Grand Total: 2,538,336 2,078,721 1,737,416 1,789,243 1,305,518 1,815,611 1,644,073 1,968,773

(%Rice) (28.7) (7.4) (7.6) (21.6) (43.1) (42.5) (34.9) (27.4) Table 12

PRODUCTION AND CFFTAKES By Calendar Year (000 tons)

Foodgrain Production 1/ PFDS Offtakes

1972 9,815 1,804

1973 9,490 Z,079

1974 11,780 1,737

1975 11,990 1,789

1976 12,550 1,305

1977 11,830 1,806

1978 12,930 1,644 E 1979 12,900 2,400

Note: 1/ The aman crop is included in the calendar year in which the harvest is completed; e. g., the 1972-73 crop appears in the data for CY 1973. Table 13

END OF MONTH BDG FOODGRAIN STOCK LEVELS

( 1972/73 - 79/80 )

(Figures in thousand tons )

Months 1972-73 1973-74 1974-75 1975-76 1976-77 1977-78 1978-79 1979 -80

July 434 218 340 752 806 472 605 419

August 409 213 363 787 764 607 592 645

September 351 296 196 684 705 730 602 892

October 203 388 106 704 565 663 716 776

November 208 294 120 655 453 624 877 757

December 136 267 176 926 464 815 910 905

January 205 203 279 1,050 550 874 855 8J5*

February 225 171 252 1,046 559 821 775 845*

March 501 144 274 1,060 471 752 608 765*

April 491 149 311 844 371 713 485 780*

May 401 238 479 792 370 736 385 820*

June 297 214 749 823 376 667 209 830*

Source : WFP Foodgrain Forecast

* USAID Projections Table 14

Official Prices of Paddy, Rice and Wheat

(Taka per Maund)

Paddy Rice Wheat (a) (b) (c) (a) (b) (c) (d) (e) (c)

Procurement Price:

Before transpor- 80 96 105 130 154 165 80 86 106 tation bonus

Aft !rtranspor- 84 100 110 134 158 170 84 90 110 tation bcnus

OMS Prices: (f) Trigger price - - - 140 172 -

Target price - -- 150 184 - (a) (b)

OMS selling price ------85 87

Ration prices - -- 100 lz0 - 80 95

Notes: (a) In effect during 1978-79. (b) Changed as a result of the Amendment of June 29, 1979 (N. B.: Actual changes took place some time later). (c) Effected Nov. 12, 1979. (d) Prior to March 1979 (e) From March, 1979 to Nov., 1979. (f) Established at TIk 145/md. in the original Title III Agreement but ±uwered to Tk. 140/md. at BDG request in October, 1978. FIGURE I FOODGRAIN PRODUCTION AND OFFTAKES, 1971-72 TO 1978-79 (PRODUCTION AND OFFTAKES ON FISCAL YEAR BASIS)

PRODUCTION (Millions of Tons) OFFTAKES (Milions of Tons) 14 4

13- 3 PRODUCTION

12 -2

12~~~~~% ~ ~ -snf ~ A 0*'---OFFTAKES 2

11 1

10

1971-72 1972-73 1973-74 1974-75 1975-76 1976-77 1977-78 1978-79 FIGURE 2 FOODGRAIN . RODUCTION AND OFFTAKES,1971/72-78/79 (PRODUCTION FOR ADJUSTED CROP YEARS ENDING IN SEPTEMBER, OFFTAKESEBY CALENDER YEAR) PRODUCTION (Millions of Tons) OFFTAKES (Miltions-of Tons) 13 3

12 '0, 2

11 OFFTAKES 1

PRODUCTION 10 N0 0

9

1972 1973 1974 1975 1976 1977 1978 1979 FIGURE 3 FOODGRAIN PRICES AND OFFTAKES, 1972-79 OFFTAKES (Millions ot Tons) PRICE (Taka per Maund) 3.0 300

2.5 250 OFF TAKE S

2.0 200

1.5 5 I

1.0 100 RICE PRICE

0.5 50

1972 1973 1974 1975 1976 1977 1978 1979 FIGURESEASONAL INDICES OF OFFTAKES AND COARSE RICE PRICE "

INDEX (Ave. price for each year=100) 120 FOODGRAIN OFFTAKES (1972-77)

110

100 .1977-78 PRICE INDEX

90 1972-77 PRICE INDEX

80

JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC

_/ RICE PRICE INDICES ADJUSTED FOR INFLATION I' INDEXES OF REAL WAGES AND RICE PRICES, 1962-79 / FIGURE 5

(1969/70=100) A 110 1 I

/ INDUSTRIAL WAGBS I / "I III

I I/ I AGRICULTURAL WAGES , 100

I RICE PRICE

// S [I l / V

60

700 1962 '63 '64. 65 '6 6 '67 68 69 '70 '71 '72 '73 '74, '75 76 _7 7 '78 '79 FIGURE 6 BANGLADESH RICE (Take per Seer) RETAIL WHEAT, ATTA AND RICE PRICES, 1971-79 WHEAT & ATTA (Tk./Md.) 7.50 300

6.25 250

5.00 200 - ~3COARSERICE

3.75 150 2.50 100 -

WEATC '

1.25 50 1

1971 1972 1973 1974 1975 1976 1977 1978 1979 FIGURE 7 RETAIL PRICES AND GRAIN OFFTAKE, BRAHMANBARIA (Sept. 1978 - Oct., 1979)

PRICE (Tk./Md.) OFFTAKE (000 Mds.) 300 300

250 250

NEW AMAN RICEC--­ 200 -- 200 " -AUS RICE

OLD AMAN RICE

15 150

lOO - 100

so0- 50 '1WHEAT i OFFTAKE - 50Is

RICE OFFTAKE OMS 0 SEP OCT NOV DEC JAN FEB MAR APR MAY JUN JUL AUG SEP OCT 1978 1979 FIGURE 8 RETAIL PRICE AND OFFTAKES, SUNAMGANJ PRICE (Tk./Md.) (Sept. 1978 - Oct., 1979) 300

250 --­ 4 V

)AUS 200 O(-FTAKE (000 200Md.)

AT TA

150 A150 _BORO/­

10A US NEW AMAN A A . -ll

-- May be wheat not atta 100 W -- WHEAT OFFTAKE Unclear [abet 50 W sW w -50 5 W 1 - -W W 0° R -

0 - -- O=OMS ..... R. R R SEP OCT NOV DEC JAN FEB MAR APR MAY JUN JUL AUG SEP OCT 0 1978 1979 FIGURE 9 RICE & WHEAT PRICES AND OFFTAKES, KURIGRAM PRICE (Tk./Md.) JUNE-DECEMBER, 1979 300

OFFIAKE (000 Mds.) 250 250

200 ATTA - 200 WHEAT

TOTAL OFFTAKE '-", COARSE RICE (Aus) 150

100 - 100

50 - 50

0 0 JUN JUL AUG SEP OCT NOV DEC 1979