Document Sample




                      OF THE




                   December 1997

                                   EXECUTIVE SUMMARY

       The Rubber Rehabilitation and Expansion Project was conceived in the context of the
Government’s policy to produce rubber in the Chittagong and Sylhet regions on land which is
unsuitable for food crops. Rubber was promoted for its potential to generate employment and
save foreign exchange. Through the rehabilitation of 5,100 hectares (ha) of existing plantations
and new planting of 5,048 ha (of which 3,790 ha would be on four new estates), the Project was
expected to yield 11,500 tons of dry rubber annually during peak years of the plantations.
        The Project was supported by a Bank loan of $20 million, approved in August
1980. The Government-owned Bangladesh Forest Industries Development Corporation
(BFIDC) was the Executing Agency for the Project. Implementation of the Project was
completed in June 1989, with a delay of two years, and at a cost of $25 million, or 32
percent less than the appraisal estimate of $36.6 million. A Project Completion Report
(PCR) was prepared in September 1989. Although none of the new rubber plantings
had reached maturity for tapping at the time of PCR preparation, the PCR projected that
the expected increase in rubber production would yet be achieved. The PCR concluded
that, as a pilot undertaking to determine the commercial viability of large-scale rubber
production, the Project could be viewed as a success. No Project Performance Audit
Report was prepared for this Project because it was not selected for independent
        In view of the long gestation period of the Project and its uncertain outcome at
the time of PCR preparation, a reevaluation of the Project was undertaken in 1996 to
assess its actual operational performance and the achievement of its intended
objectives. Contrary to the PCR’s conclusions, the reevaluation findings indicate that the
Project has not performed up to expectations. The rehabilitation works did not contribute
to any increase in rubber yields as envisaged. The average yields of the replanted and
newly planted areas have been substantially less than the appraisal projections; the
incremental annual rubber production from all Project estates during the peak period is
now expected to reach only 55 percent of the appraisal target. The shortfalls in rubber
yields and production were a result of the low tree density, reflecting the effect of natural
calamities and partly inefficient management of BFIDC’s rubber estates. Because of the
low rubber output, the processing facilities provided under the Project have been
        As a result of the substantial shortfalls in rubber production, both the economic
internal rate of return and financial rate of return for the Project are reestimated to be
negative. Investment in the Project appears too costly to save foreign exchange through
domestic production of rubber to substitute imports. The benefits of strengthening the
capability of BFIDC staff on rubber plantations cannot be sustained because effective
training was not provided after Project completion. Overall, the Project is considered
unsuccessful; however, it has demonstrated that high-yielding rubber trees can be
cultivated in Bangladesh. Recent private sector involvement in rubber production has

indicated that rubber plantations, if properly and professionally managed, could be a
  financially viable venture.
         Experience of the Project suggests the need to carry out a domestic resource
  cost analysis to determine the conditions that would make it worthwhile for Bangladesh
to promote rubber production to substitute imports. The promotion of rubber production
should be built upon the results of research and scientific experiments on various
agronomic aspects of rubber cultivation under local conditions. The reevaluation
findings also indicate that a conducive policy environment needs to be created to
stimulate active private sector involvement in development of the rubber sector. The
Government should (i) develop an efficient marketing system for rubber products, (ii)
adopt land use policies allowing long-term leases of unutilized public lands for rubber
plantations, (iii) improve basic infrastructure in rubber growing areas, (iv) institute a
regulatory framework to strengthen Bangladesh’s competitiveness in rubber production,
and (v) review the role of BFIDC and initiate measures to privatize BFIDC’s rubber

                                        I.       INTRODUCTION

A.     Background

1.             The idea of cultivating rubber on hilly, degraded, Government-owned lands
goes back to the 1960s. The rationale was that the climatic and soil conditions are similar
to those of neighboring countries, which have long cultivated rubber. Furthermore, despite
the overall shortage of arable land in Bangladesh, there are hilly areas that have been
denuded and degraded but are still suitable for rubber cultivation. If these otherwise
underutilized resources could be brought into production to satisfy part of the country’s
demand for rubber, there could be additional benefits to the national economy. Following
a 1964 feasibility study,i which confirmed the economic and financial feasibility of large-
scale planting, the Government established some 10,400 hectares (ha) of rubber
plantations in the Chittagong and Sylhet regions between 1965 and 1975.

2.            By the late 1970s, however, the results in terms of tree survival and yields
were disappointing, and the plantations had generally deteriorated. In 1979, the Bank
approved technical assistance to the Government of Bangladesh to prepare a feasibility
study on a rubber rehabilitation and expansion project.1 The Project was appraised by a
Bank mission in January/February 1980. In August 1980, the Bank approved a loan of
$20 million from its Asian Development Fund to finance the Project's entire foreign
exchange cost of $10.4 million and a substantial portion of its local currency cost of $26.2
million equivalent. The Executing Agency for the Project was the Bangladesh Forest
Industries Development Corporation (BFIDC), a semiautonomous Government

B.     Objectives and Scope

3.              The Project's principal objective was to increase raw rubber (latex)
production. Other objectives were to promote the productive use of steep and degraded
lands that are unusable for other crops, substitute for the import of raw rubber, create
employment and generate incomes, and prevent further environmental degradation of
hilly lands. To achieve these objectives, the Project comprised the following
components: (i) rehabilitation of 5,100 ha of seven existing plantations through
replanting with improved varieties, infilling of thinned-out plantations, repairing terraces,
improving ground cover, applying fertilizer; (ii) new planting on 5,048 ha to expand
existing plantations and to establish four new estates (one in the Sylhet region and
three in the Chittagong region); (iii) construction of buildings, roads, and water supply;
and provision of equipment, vehicles, and machinery; (iv) establishment of processing
facilities to produce smoked rubber sheets; (v) staff development through out-of-country
training, and establishment of a Technical Training Unit; and (vi) consulting services for
Project implementation.

      TA No. 286-BAN: Rubber Rehabilitation and Expansion Project, for $60,000, approved on 27 March 1979. The
      feasibility study was carried out by the Food and Agriculture Organization's Investment Centre.

 C.     Findings at Project Completion

        1.     Implementation Performance

4.             Implementation of the Project was completed in June 1989, two years
behind the original completion date of June 1987. Most of the delay occurred during the
initial two years because of slow transfer of land from the Forest Department, delays in
purchasing the initial planting material for the nurseries, and subsequent difficulties in
producing sufficient quantities of cloned planting material. The actual cost of the Project
was $25.03 million, which was about 32 percent less than the appraisal estimate of $36.6
million, because of a reduction in the scope of some components (paras. 5-6),
depreciation of the taka against the US dollar, and overestimation of contingencies.

5.            A Project Completion Report (PCR) was prepared by the then Agriculture
Department in 1989. The PCR found that the appraisal targets of new planting of 5,048
ha and replanting of 2,487 ha under the rehabilitation program for existing plantations had
been achieved. However, other rehabilitation works fell short of expectations: terrace
repair reached 47 percent of appraisal target; infilling with new plants on vacant points, 32
percent; and ground cover improvement, only 20 percent. Fertilizer application fared
better, and was reported to have been performed on 90 percent of the targeted area.

6.             Implementation performance of other project components was varied.
Vehicles and construction equipment were supplied on time. Tractors, however, were
soon immobilized due to lack of repair capability and/or spare parts. Construction of
residential buildings on the plantations was largely underachieved; only about half of the
targeted numbers were built, and the standard of completed buildings turned out to be
lower than envisaged. On the other hand, the number of offices, garages, workshops,
schools, mosques, regional centers, resthouses, etc., exceeded the targets. The
provision of processing facilities for producing ribbed smoked sheet (RSS) was delayed;
only one sheeting battery unit was reported completed, while work on four other units was
ongoing at the time of the Project Completion Mission. Rudimentary training facilities were
constructed. Foreign training was provided to 34 staff for a total of 30 person-months as
compared with 90 person-months envisaged at appraisal. The Consultants were judged
to have satisfactorily imparted skills in rubber nursery, planting, and latex processing to
workers of BFIDC's rubber estates. They provided technical training to BFIDC staff, and
also introduced an accounting system and plantation management procedures.

       2.     Early Operational Performance

7.            The PCR observed that none of the new tree plantings had reached
maturity, and that cultivation efforts and maintenance of trees and facilities varied among
plantations, ranging from marginal to excellent. For those locations with newly planted
trees visited by the PCR Mission, the planting material and care of trees were up to
standards, and tree growth was satisfactory. These observations led to the PCR's
prediction that, after a seven-year maturing period, all newly planted areas under the
Project would achieve at least 90 percent of the yield estimated at appraisal, i.e., 1.1 tons
(t) per ha.

8.               Regarding the rehabilitation component, the PCR predicted that the
 incomplete rehabilitation of the existing plantations would not be severely detrimental to
 the overall outcome of the Project. It was argued that the yields of the old trees in the
 existing plantations would not increase much in response to terracing, ground cover, or
fertilizer application.

9.            In assessing the Project benefits, the PCR estimated that the rubber trees
planted under the Project would be fully tapped and processed to produce about 10,000 t
of RSS per annum starting from FY2006, which was close to the appraisal's peak
production projection of 11,500 t by FY2004. However, neither the financial internal rate
of return (FIRR) nor the economic internal rate of return (EIRR) was calculated in the

       3.      Assessment of Project Performance

10.           The PCR concluded that, while some Project components were not fully
implemented, the basic requirements for commercial rubber estates had been
established. As a pilot endeavor to demonstrate the commercial viability of large-scale
rubber production, the Project could be viewed as a success. The PCR noted that further
improvement in the efficiency of rubber production in Bangladesh would require changes
in the ownership and management structure of BFIDC and in the heavily protected
domestic market for raw rubber products.

D.     Purpose of Reevaluation

11.           Following procedures for selective evaluation of completed projects, a
project performance audit report was not prepared for this Project. Based on current
practice, impact and reevaluation studies to be undertaken may be selected from two
groups of completed projects: those that were postevaluated at least five years earlier,
and those that were not selected for postevaluation but were completed at least eight
years ago. This Project is the first to be selected for reevaluation from the second group.

12.            Undertaking a reevaluation of the Project is considered appropriate given its
long gestation period. Reevaluation enables the Bank to examine the operations and
impact of the Project several years after its implementation, and permits a more realistic
assessment of its direct and indirect social and economic impact, potential for further
development, and sustainability. Major issues relevant to rubber cultivation can be
identified and discussed to provide a basis for the Government to review and redefine its
involvement and policies in the sector. The Bank could also draw on this experience for
considering its own future involvement.

13.              The Reevaluation Study is based on the findings of a Reevaluation Mission (REM)
that visited the country 13-26 May 1996. The REM held discussions with various concerned
agencies of the Borrower, specifically with BFIDC and its field office personnel. Apart from visiting
BFIDC's rubber plantations, the REM consulted with private small- and large-scale rubber tree
planters, traders, and manufacturers as well as other knowledgeable persons in the rubber sector.
The REM also drew on the experience of concerned Bank staff, the data base of the PCR, and
other documentation in the Bank files.

                            II.    REEVALUATION FINDINGS

 A.     Rubber Sector Development

14.            Even with the prevailing low-income level in Bangladesh, the increasing
population exerts a steadily growing demand for rubber products, such as footwear,
bicycle tires, foam rubber, tubes, conveyor belts, and sanitary and medical articles. In
FY1993 overall consumption of simply processed raw rubber, i.e., RSS, amounted to
5,076 t, valued at $36 million. Future demand for natural rubber might be constrained
somewhat by the possibility of synthetic substitutes. However, traders and manufacturers
of rubber products feel that any such substitution is likely to be compensated for by the
growth in demand created by increasing population and incomes.

15.            About 90 percent of the domestically produced rubber is traded through
BFIDC. Privately owned rubber estates typically have their own arrangements with users.
Traders as well as manufacturers of rubber products directly import natural rubber as well
as semifinished or finished rubber products. Most traders and manufacturers perceive
quality differences between locally produced and imported RSS, although this is disputed
by BFIDC. There is general agreement, however, that BFIDC rubber sheets are better
processed than those processed by small private rubber growers.

16.             Approximately 70 percent of Bangladesh's demand for natural rubber has
been met through imports. Time series data show that the percentage of imports to total
consumption does not vary with the availability of locally produced rubber. One reason for
this is that local production as well as the volumes marketed through BFIDC tend to vary
greatly from year to year. The uncertainties created thus discourage rubber users from
depending on domestic output. The users would rather import and adjust their production
and stocks than rely on the vagaries of locally available raw material.

17.            The price of imported RSS, inclusive of tariff and taxes, is slightly higher
than that of locally produced rubber. Import tariffs on rubber have gone down by more
than 60 percent in recent years, and tariff liberalization is expected to continue. Demand
for rubber will continue to be met primarily through imports, as reduction of imports could
be expected only if local productivity increases significantly and marketing becomes more

B.    Operational Performance

      1.     Plantation Operation and Cultivation Practice

18.           The Project has contributed to the expansion of BFIDC's rubber plantation
areas from the pre-Project level of 5,100 ha to 9,926 ha, which is slightly less than the
appraisal target of 10,150 ha. A total new area of 4,855 ha, as against the planned target
of 5,048 ha, was established under the Project, comprising 3,677 ha in four new estates
and 1,178 ha in six existing estates (Appendix 1). In addition, replanting was
accomplished on a total of 2,488 ha in the seven existing rubber estates as planned,
although with some delays (Appendix 2).

19.            The REM field visits as well as the data supplied by BFIDC show vastly
  differing conditions and operational practices at the various plantations. The good
  plantations have uniform tree girth and canopy and minimal weeds, whereas the poor
  ones have big gaps in the planted areas with undesirable undergrowth dominating the
spaces between trees. The new estates established under the Project are not necessarily
in better condition than the plantations existing prior to the Project.

20.            The rubber estates in the Sylhet region are generally in better condition and
perform better than those in the Chittagong region (see the profile of BFIDC's rubber
estates in Appendix 3). Data from FY1995 show that the average density of all plantations
in the Sylhet region was 286 trees per ha, compared with 188 per ha in the Chittagong
region. The overall average density of all trees for both regions was 221 per ha. In terms
of density of tappable trees, the averages were 231 per ha and 81 per ha for plantations
in the Sylhet region and the Chittagong region, respectively, while the overall average for
both regions was 130 per ha. These figures are far below the projection of 370 tappable
trees per ha made at appraisal. Two cyclones in 1991 and 1992 severely affected tree
stands in the Chittagong region, destroying as much as 60 percent of the trees in some of
the plantations. More damage to the plantations was caused by pilferage of lumber and
firewood by people rendered homeless by the catastrophes. Natural mortality and
brushfire further reduced plant density.

21.            Why have the gaps in some plantations remained for so long? Although
some replanting and infilling are being done at small annual increments, establishing new
trees among the existing large and matured trees with overextended canopies is
becoming more difficult. In the long run, uniformity of trees will not be achievable, and the
benefits of related economies of scale will not be realized.

22.            Prior to the Project, low-yielding varieties were planted in BFIDC's rubber
estates. Under the Project's new planting and rehabilitation components, only high-
yielding varieties originating from Malaysia were used. The plant material is produced by
seeding and then cloning, i.e.,c implanting buds taken from stands of imported well-
performing trees (bud banks) into the seedlings. The cloning is done by specially trained
plantation staff. At present, high-yielding varieties are planted in new areas as well as to
replace immature, damaged, or diseased trees. Of the total area of 9,926 ha of rubber
plantations owned by BFIDC, some 7,665 ha (77 percent) was planted to high-yielding
varieties (Appendix 3). However, no research has been carried out to test the adaptability
of the imported high-yielding varieties to Bangladesh's topography and climate, nor has
research been undertaken to determine optimal plant spacing.

23.             The condition and the productivity of nurseries vary among the plantations
but generally leave much to be desired. The REM noted that several of the plantations did
not have stands of trees to supply buds for cloning. Generally, the nursery operations
were inadequate either to meet the need for a systematic, continuous, and substantial
planting effort, or to provide large numbers of quality plants to the private sector.
Ironically, vital nursery activities were allowed to deteriorate even though the need for
them was obvious and intensified over the years.

24.           The cultivation of beneficial ground cover with nitrogen-fixing legumes
  serves as another indicator of efficiency in operations. The level of effort in clearing
  undesirable weeds and replacing them with beneficial ground cover varies, but appears
  to be low in most plantations. Encroachment by settlers to graze cattle and to collect
firewood is a major problem at some plantations. It is important to avoid conflict with
impoverished people who crowd around some of the plantations.

       2.     Yields and Production

25.             It was assumed at appraisal that, under Bangladesh's climatic regime and
soil types, the newly established plantations would start producing latex about seven
years after planting. The yield measured in the form of dry rubber was expected to be 225
kilograms (kg) per ha (based on 150 trees tapped per ha) in the first year of tapping, and
to gradually increase to 674 kg per ha, and 1,010 kg per ha during the third and fifth year,
respectively. The yield would stabilize at around 1,235 kg per ha during the period from
year 7 to year 21, and then taper off until the end of the productive life of the rubber tree
at year 30, after which the tree should be felled and replaced. The assumed yields were
based on the experience of other rubber-producing countries. As for the existing
plantations, it was predicted that the yield would increase from 370 kg per ha to 865 kg
per ha in the mature areas as a result of rehabilitation through terrace repair, ground
cover improvement, replanting, and fertilizer application.

26.           Rubber yields vary widely among the plantations. The rubber estates in the
Sylhet region have generally achieved higher yields than their counterparts in the
Chittagong region. However, none of the plantations, except the Rupaichara estate in the
Sylhet region, has achieved the yield targets anticipated at appraisal (Appendix 3). The
yield of dry rubber for the newly established estates averaged about 192 kg per ha in
FY1995, while that of the existing plantations averaged about 321 kg per ha in the same
year. These yield levels compare unfavorably with the appraisal projections.

27.             The major reasons for low yields in the newly established plantations are (i)
significantly lower tree density; (ii) poor survival rate of the initial planting; (iii) missed
opportunities for early infilling and/or replanting of perished trees; and (iv) unsystematic
cultivation works such as ground cover planting, terrace repair, and fertilizer application.
The yields in the rehabilitated plantations are impeded by the mix of low- with high-
yielding species, the mix of mature and young trees, and by the absence of a system to
care for the trees and monitor their growth. The low yields also reflect the lack of
continuous guidance and supervision by management to motivate workers to use proper
tapping technique. Well-managed private plantations in Bangladesh were observed to
label trees with numbers, possess maps of blocks of trees, and to maintain records by
blocks to monitor tree growth and yields as well as the performance of tappers. None of
these management practices could be observed at BFIDC's plantations.

28.           The rubber production in the Project estates grew steadily from 899 t in
FY1984 to 1,313 t in FY1991, with an average annual growth rate of 5.5 percent. In
FY1992, production dropped by about 12 percent to 1,153 t as a result of damage to
several estates caused by a devastating cyclone that hit the Chittagong region. Rubber

production registered a big jump to 2,713 t in FY1995, when the four estates established
 under the Project reached mature age for tapping. However, this level of production was
 only one third of the appraisal's estimated output for that year. The production shortfall
 reflects the low yields at the Project estates, which was caused by several factors (para.
27). Statistics of rubber production from BFIDC's estates are provided in Appendix 4.

29.           The danger of cyclone damage to the rubber trees was not sufficiently
considered in the site selection for new rubber plantations. As a result, some plantations
in the Chittagong area are located in sites prone to cyclones. Also, the detrimental impact
of an unstable law and order situation in the Chittagong hill tracts on the estates with their
many indigenous personnel was underestimated. These two factors, together with
generally unrealistic expectations of production from a newly introduced tree crop,
contributed to the large discrepancies between forecast and actual rubber production.

       3.     Processing and Quality of Output

30.          All five automatic sheeting batteries for processing latex into RSS were
installed as originally planned. These units are located at the Dabua, Ramu, and
Dantmara estates in the Chittagong region; and at the Satagaon and Rupaichara estates
in the Sylhet Region. Except for the one in Dabua, all units were completed after
preparation of the PCR. Because of the shortfall in the anticipated increase in latex
outputs, none of these units was operating at full capacity at the time of the REM.

31.             Firewood is the main source of fuel for processing latex into RSS. For
quality control, adequate smoking time and uniformity in heat distribution are required.
During its field visits, the REM observed that proper supervision of rubber processing was
lacking, and that there was either insufficient or excessive smoking, which impaired the
quality of the RSS produced.

       4.     Other Support Facilities

32.            The civil works for support facilities constructed under the Project are of
varying quality. The regional offices, guest houses, and plantation managers' houses
were well built. In contrast, the quality of smokehouses and stores was mostly
unsatisfactory. In particular, the staff housing and related structures, as well as the roads
leading to and within estates, were of poor design and construction. The poor condition of
infrastructure and housing have adversely affected the morale of estate staff. Poor
communication facilities hinder supervision at the plantation level by staff from regional
offices and the headquarters of BFIDC. Another unfortunate development was that the
training activities at the Technical Training Unit tapered off after the departure of the
consultant in 1986 and were altogether suspended from 1988 until 1994.

 C.     Marketing and Distribution of Raw Rubber

33.            BFIDC controls nearly 90 percent of the domestic rubber supply. To sell its
rubber, BFIDC periodically seeks bids from its enlisted customers, taking into account the
status of its inventory and management's perception of demand for local rubber. The
prevailing world price of rubber along with import activities by traders and large rubber
goods manufacturers are taken into consideration in assessing the demand conditions.
Usually, BFIDC calls for a tender at least four times a year, although sometimes the gap
between two tenders may be more than six months. The bidders submit their quotes
either at BFIDC's headquarters in Dhaka or at its Chittagong regional office. The highest
bidder is guaranteed a minimum quantity, and the remaining inventory is offered to a
selected number of the next highest bidders at the price quoted by the highest bidder. In
the event BFIDC perceives the highest bid price to be too low, it reserves the right to
ignore the quotations and solicit fresh bids at a later date. Appendix 5 shows the average
price of BFIDC rubber for the period FY1984-1995.

34.           The arbitrariness in BFIDC's decision to make its rubber supply available to
users and the delays and uncertainties in timing the auctions tend to distort the market
and increase speculation among buyers. It is alleged that BFIDC can artificially raise the
price by including a single buyer expected to bid a high price. Collusion among enlisted
buyers has been attempted to reduce BFIDC's influence. The rubber marketing system of
BFIDC lacks transparency and reliability, and gives rise to both speculative and collusive
behavior among the buyers.

35.            There a number of reasons why traders and rubber goods manufacturers
continue to seek BFIDC's product in spite of the difficulties encountered. The alternative
domestic source of rubber comes from private sector producers, but the supply from this
source is still small, and its distribution system is not well developed. Most private
plantations sell dry rubber at the farm gate to selected customers. End users consider the
quality of rubber from small private producers inferior to that of BFIDC. Small buyers of
rubber find the transaction costs, including custom formalities, and delays, too high. The
intermediate traders and the large rubber goods manufacturers usually expect the highest
bid price to be somewhat less than the world price plus taxes and transportation costs.
However, as the tariff rates decline, customs clearance procedures are simplified, and
supply from private domestic growers increases, it will be more difficult for BFIDC to set
target prices and successfully clear its inventory.

 D.     Institutional Development

36.             BFIDC was established in 1959 under the East Pakistan Forest Industries
Development Corporation Ordinance No. 67 by the then Provincial Government of East
Pakistan. It adopted its present name after the independence of the People's Republic of
Bangladesh in 1971. BFIDC was a division under various ministries. BFIDC is now a
semi-Government corporation under the jurisdiction of the Ministry of Environment and
Forests. The organization is headed by a Board of Directors and a Chairperson, who
reports to the Secretary of the Ministry. The organizational structure and functional
divisions within the Ministry as well as within BFIDC have changed repeatedly. According
to its charter, BFIDC's tasks are to encourage, sponsor, or establish forest-related private-
sector manufacturing. In reality, most of its activities have concentrated on expanding
public involvement in the sector. It has established and operated 14 production units, one
of which deals with rubber production.ii The rubber production unit is BFIDC's largest in
terms of employment, second in terms of capital investment, and third in terms of sales.

37.            During appraisal, it was assured that the Director as well as key staff would
be changed less frequently than in the past, and that they would remain in their positions
for at least three years. It was further suggested that more managerial responsibilities
would be delegated to the estate management level. At the time of the PCR, it was noted
that there had still been a fast turnover in management. Five directors were appointed
during the eight-year implementation period. Also, BFIDC's management system
remained highly centralized. Little authority for day-to-day decisions was delegated to the
estate management. The PCR recommended an increase in autonomy for estate
managers and implementation of a commercial accounting system. It was also suggested
that a separate Rubber Estate Directorate be established within BFIDC to manage all
production and marketing activities.

38.           The REM observed some positive changes that have taken placed since
the time of the PCR. One Directorate has been exclusively entrusted with responsibility
for rubber activities, and more authority has been delegated to the regional and estate
managers. Under normal circumstances, key staff are expected to remain in the same
post for three years. However, financial accounting has not significantly improved. In
addition, BFIDC has continued to expand its operations. Its employment and
administrative overhead costs have grown out of proportion to the increase in rubber
production from its plantations. Consequently, its productivity and cost efficiency are low.

39.           Currently, the private sector is becoming more involved in rubber

production. This began in the late 1970s, when Government-owned fallow lands in the
Chittagong hill tracts were leased to over 1,200 allottees on the condition that rubber
trees be planted on a portion of the land. Commonly known as Block Rubber Planters,
many of the active growers have expanded their ownership by buying land from inactive
or failed allottees. The Bank supported private sector involvement in rubber production by
including a rubber smallholder component in the Chittagong Hill Tracts Development
Project,2 which was implemented 1979-1992. Under this Project, newly resettled
smallholders were allotted 1.6 ha per family, some 1,060 ha were planted to rubber, and
two processing facilities were established. Tea estates, foreign as well as locally owned,
have shown increasing interest in converting some of their lands to rubber production.
Finally, home-based smallholder rubber planting is practiced in selected areas of
Chittagong hill tracts and in the Modhupur area. Appendix 6 provides a summary of
BFIDC estates and various types of private sector estates.

E.    Economic and Financial Reevaluation

40.            The Project was expected to contribute to increased rubber output from the
rehabilitation of existing plantations covering an area of 5,100 ha and from 5,050 ha of
area under new rubber planting. It was estimated during appraisal that the Project would
yield an EIRR of 14 percent and an FIRR of 18 percent. Sensitivity analysis indicated that
both EIRR and FIRR would be sensitive to changes in yield and price of rubber: a 1
percent reduction in yield would bring down the EIRR by 1.24 percent and the FIRR by
1.09 percent; similarly, a 1 percent decline in the price of rubber would reduce the EIRR
by 1 percent and the FIRR by 0.9 percent.

41.            The actual incremental rubber output resulting from the Project was
substantially less than the appraisal projections. The average yield of the seven existing
rubber estates was 321 kg per ha in FY1995, which was about 63 percent lower than the
appraisal estimate of 865 kg per ha, whereas the average yield of the four newly planted
estates during the first two tapping years was 192 kg per ha, some 41 percent less than
the appraisal projection of 336 kg per ha. The rubber production from all Project estates
amounted to 2,712 t in FY1995, which was 68 percent lower than the appraisal projection.
Rubber yields are expected to increase as trees in the newly established areas reach
maturity. However, because of the failure to maintain the projected density of trees per
ha, the total production at peak performance is expected to fall below the maximum target
of 11,500 t by about 45 percent. Furthermore, the appraisal's projected increase in world
market prices for rubber did not materialize; in fact, prices declined by about 20 percent
over the 1988-1995 period.

42.          The REM reestimated the EIRR and FIRR for the Project based on the
actual operational data on benefits and costs for the period FY1983-1995 and the

      Loan No. 404-BAN(SF), for $28.5 million, approved on 28 June 1979.

projections of benefits and costs for the remaining life of the Project (Appendix 7). The
 results show that both EIRR and FIRR for the Project are negative. This is mainly
 because of the substantial shortfall in the incremental rubber output and the decline in
 rubber prices (para. 41). The negative EIRR and FIRR also reflect BFIDC's high
operating costs. The unit cost of producing RSS by BFIDC's rubber estates was generally
higher than that of the private sector rubber estates by about 30-35 percent.

F.     Socioeconomic Impact

43.           Rehabilitation and expansion of the rubber plantations under the Project
have generated direct employment for about 2,660 jobs, which is below the appraisal
target of 6,500. However, the Project has had positive demonstration effects in promoting
rubber production in Bangladesh; it has generated some interest in the private sector to
invest in rubber plantations. The Project and BFIDC helped pioneer the use of high-
yielding species, nursery operations, and the cloning technique. Furthermore, a core
group of trained and experienced planters, tappers, and managers has been formed. A
number of these managers and tappers are now working for privately owned rubber
plantations. These indirect, nonquantifiable impacts could be the Project's most important
benefits in the long run. Unfortunately, the positive demonstration effects are partly
countered by an unfavorable policy environment (paras. 50-52).

44.             BFIDC expanded its pioneering role six years ago by venturing into rubber
planting in the centrally situated Modhupur area. The REM observed during a short field
visit that the plantation was more densely planted and had more equally grown stands of
trees. This particular plantation could demonstrate higher profitability when the trees
mature. Also observed were numerous private smallholdings in the vicinity emulating the
comparatively well-tended BFIDC plantation in the Modhupur area. The private rubber
estates have to rely on BFIDC for technical advice.

G.     Women in Development

45.            Rubber plantations in other Asian countries have demonstrated that women
can be employed especially in tapping and nursery activities. Women are generally
believed to be more adept in handling bud grafting, bud-bank maintenance, and tapping.
In Bangladesh, however, women's work in plantations is discouraged by social and
cultural tenets. Of the 2,660 jobs created by the Project, REM estimates that not more
than 3 percent were filled by women, mostly in bud-bank and grafting activities, and
usually on day-to-day contracts. In view of the high unemployment and underemployment
rates among men, it could not be realistically expected that many women would be drawn
into the plantation labor force. Women have benefited indirectly when families were
accommodated in plantation housing. Even if the accommodation remains below the
envisaged standards, there are benefits from reliable water supply, availability of firewood
and electricity, provision of transport in medical emergencies, and facilities for schooling
on or near the plantations.

 H.     Environmental Impact

  46.             If large contiguous tracts of land are converted into monoculture rubber
  plantation, the natural biodiversity could be disturbed or destroyed, with long-term negative
environmental effects. Damaging effects could also result from the unbalanced use of fertilizers,
eventual exhaustion of the natural soil condition, and possible contamination of runoff water.
Furthermore, the substantial amounts of water used in the processing of raw latex are
contaminated with acidic chemicals. This water is neither treated nor is its disposal controlled.
The Government agency concerned needs to monitor closely the environmental impact of
rubber plantations. The smoking of treated and pressed rubber sheets requires substantial
amounts of wood, most of which has to be purchased by the plantations. Replanting of trees for
firewood should be promoted.
47.             Nevertheless, there are positive environmental effects when barren or grass-
covered slopes, which cannot revegetate and would be vulnerable to periodic grass fires, are
brought under rubber cultivation. Terracing and planting of rubber trees under the Project
proved to be technically feasible and environmentally acceptable measure to prevent further
degradation. The soil is protected from erosion, and the growing trees shed leaves and
branches, which, if left to decompose naturally return nutrients to the soil. If nitrogen-fixing
ground cover is cultivated in the plantations, the environmental benefits can be further
increased. In Bangladesh, the environmental impact of rubber production can be
overwhelmingly positive if the plantations maintain beneficial ground cover, take responsibility
for the safe treatment of contaminated water in holding ponds, and plant and maintain trees for
their own firewood needs.

                                      III.   KEY ISSUES

A.     Domestic Resource Cost of Rubber Development Project

48.           Experience of the Project raises a question as to whether it is worthwhile for
Bangladesh to meet its rubber demand through domestic production rather than imports.
While import substitution helps to improve the country's balance of payments, it is
important to consider the cost of saving foreign exchange through domestic production of
rubber. This can be done by calculating the domestic resource cost (DRC) of earning or
saving foreign exchange as a result of an investment project. The DRC can then be
compared with the shadow prices for foreign exchange, and such a comparison could be
used as a basis for evaluating an investment in rubber production. Analysis of the DRC
would also help determine whether Bangladesh has the potential to produce enough
rubber to make it an economically desirable alternative to imported rubber. The
preliminary analysis of the DRC undertaken by the REM indicated that selected private
rubber plantations, due to their lower operating and overhead costs as compared with
those of BFIDC's plantations, could generate added value in excess of the opportunity
costs of production. However, more detailed analysis is required to confirm that
Bangladesh would have a clear comparative advantage in producing rubber.

 B.     Policy Environment

49.             The Government's policies in the rubber sector need to be reexamined to
induce more dynamic development of rubber production. With the continuing import
liberalization, the import tariff on rubber has decreased by more than 60 percent in the
past three years. This decline, coupled with the simplifying of customs clearing
procedures, would tend to bring down the prices of imported rubber. It will become
increasingly difficult for domestically produced rubber to compete unless a high level of
productivity is achieved.

50.           Authorities controlling large tracts of underutilized hilly lands are not eager
to make them available for rubber planting. BFIDC and the association of private sector
rubber planters have been repeatedly stymied in their efforts to persuade the landholding
agencies. Since the implementation of a Bank-supported scheme to allocate public land
for rubber planting by hill tribe people,iii and a ten-year-old Government program of
leasing lands to individuals for rubber planting, no further initiatives have been taken by
the Government to use idle public lands for rubber plantation in the Sylhet or Chittagong
regions. This might be due to the fact that past programs were criticized for lack of
transparency in their land allocation procedures.

51.            Strategies should be developed to promote private sector rubber production
and make it more competitive with imports. The Government's land-use policies should
allow the lease of idle public lands for rubber plantations by the private sector. Although
private sector initiatives in rubber plantings are more in the nature of pilot undertakings,
some of the private plantations have recently reached the tapping stage, and early results
indicate that they have achieved higher productivity than BFIDC's plantations. The private
sector estates have planted high-yielding varieties of rubber and maintained a greater
density of trees compared with BFIDC's estates. They also enjoy a lower operating cost
per unit of output because of low overhead costs and greater flexibility in adjusting their
labor requirements. The share of private estates in local rubber production will increase in
the near future. BFIDC will be able to maintain its market share only through Government
policies that allow it to obligate traders and users to purchase its output at prices upheld
through withholding of local supplies, import restrictions, and tariffs. Private smallholders
are in a better position to cope with further decline in rubber prices, whereas BFIDC
estates will find it difficult to face increased competition unless they can substantially
reduce their operating and overhead costs.

52.              Recent developments in the rubber sector suggest that the Government's
policies to protect BFIDC's market share need to be reviewed and reformed.
Uncertainties and confusion surrounding BFIDC's auction mechanism have encouraged
speculative bidding as well as collusion, to the detriment of rubber product producers and,
ultimately, of the consumers. The continuation of such policies will be at the expense of
private initiatives and will result in unnecessarily high rubber prices for end users.

53.              The role of BFIDC in the rubber sector also needs to be reviewed. If the private
  sector is encouraged to increase its involvement in rubber production, privatization of BFIDC's
  rubber estates should be considered. With over 30 years of experience in pioneering rubber
  production in Bangladesh, BFIDC can play an important role in two major areas. It can conduct
research and provide extension services to further increase the productivity of private sector
rubber plantations. It can also serve as a regulatory and advisory body to provide guidance and
institutional support towards strengthening Bangladesh's competitiveness in growing rubber, and
towards developing an orderly rubber marketing and pricing system that would benefit both
producers and consumers.

                                     IV.     CONCLUSIONS

A.     Overall Assessment

54.            The Project helped to rehabilitate about 5,100 ha of existing rubber
plantations and to expand rubber planting over 4,855 ha, including the establishment of
four new rubber estates. However, the rehabilitation comprising terrace repair, ground
cover improvement, infilling with new plants on vacant points, and fertilizer application, did
not contribute to any increase in rubber yields. The average yield of the four newly
established estates has been substantially less than the appraisal projection. The annual
rubber production from all Project estates at the peak period is now expected to reach
only 55 percent of the appraisal projection of 11,500 t. The main reasons for the shortfalls
in rubber yields and production are low levels of fertilizer application and the significantly
low tree density, reflecting the effect of natural calamities, and partly the inefficient
management of BFIDC's rubber estates. Because of the low rubber output, the
processing facilities provided under the Project have been underutilized.

55.            The Project has not performed up to expectations. The substantial shortfalls
in rubber production are expected to bring the EIRR and FIRR down to negative levels.
Experience indicates that the Project does not save foreign exchange through domestic
production of rubber for import substitution. The assumptions used in projecting benefits
at appraisal were overly optimistic, especially when considering that high-yielding
varieties of rubber were introduced to Bangladesh, and the rubber estates were to be
managed by a public sector agency. The benefits of strengthening the capability of
BFIDC staff in rubber plantations cannot be sustained, because the training unit
established under the Project has been inactive. Viewed overall, the Project is considered

56.             Nevertheless, the Project has had a positive impact in its demonstration
effect, which shows that high-yielding rubber trees can be cultivated in Bangladesh.
Recent private sector involvement in rubber production has indicated that rubber
plantations, if properly and professionally managed, could be a financially viable venture.
Most of the private sector's sustained interest in rubber production springs from the belief
that BFIDC's shortcomings and inefficiencies can be avoided under a private
management system.

 B.     Lessons Learned

57.            The expansion of still unfamiliar large-scale rubber cultivation, the
introduction of new high-yielding varieties and rubber processing technology, and the
determination of required management capabilities need to be based on thorough
investigations. Particularly, it is important that research and scientific experimentation be
carried out to test the adaptability of imported high-yielding varieties under local
topography and climate, and to determine the optimal spacing and density of plants for
rubber plantation. Similarly, the rehabilitation of existing rubber plantations should be
based on a clear understanding of the factors causing their deterioration, so that proper
corrective measures can be identified for incorporation in project design.

58.            Assessment of the economic soundness and financial viability of an
investment in rubber production should be based on realistic assumptions about rubber
yields. Projections at appraisal should not be based on the assumption that the same
high level of productivity of the world's leading rubber producers can be achieved
immediately. It is more realistic to anticipate slower buildup of production and lower cost
efficiency than those reached in countries with longer experience with the crop.

59.           For countries prone to natural calamities such as cyclones, site selection for
rubber plantations should take into consideration their potential effects. The risk of
damage from cyclones should be taken into account in the estimation of EIRR and FIRR
for investments in rubber production.

60.           While a Government agency can play a pioneering role in introducing a new
productive activity into a country, it is important to stimulate private sector involvement.
Private enterprises generally have greater flexibility and capacity to carry out the
expansion of productive activity more efficiently than does a Government agency. In this
respect, the Government should create a policy environment conducive to investment by
the private sector.

C.     Follow-up Actions

       1.     For the Bank

61.           Any further Bank or external donor involvement in the rubber sector in
Bangladesh should be based on a comprehensive analysis of the domestic resource cost
of producing rubber to substitute imports, the potential for future development, and the
policy and institutional reforms that need to be implemented to stimulate private
investment in the sector.

       2.     For the Borrower

62.            The major actions recommended to be taken by the Government include
the following:

                         (i)     to undertake scientific research (by the Bangladesh Forestry
             Research Institute and/or the Bangladesh Agricultural Research Council in collaboration
             with BFIDC) on various agronomic aspects of rubber production to determine ways to
             increase productivity;

                            (ii)   to support the expansion of private sector investment in rubber
            plantations through measures such as (a) development of an orderly, fair, and reliable
            marketing system for rubber products; (b) adoption of land use policies to allow long-term
            lease of unutilized public lands for rubber plantations; (c) improvement of basic
            infrastructure, e.g., roads, power, and water supply, in rubber growing areas; and (d)
            institution of a regulatory framework to provide guidance and institutional support towards
            strengthening Bangladesh's competitiveness in rubber production; and

                           (iii) to review the role of BFIDC and initiate measures towards the
            privatization of BFIDC's rubber estates.

  Conducted by EBS Management Consultants, Inc.
Other production units are grouped into timber extraction, timber production, wood preservation, veneer, and plywood manufacturing.
Loan No. 404-BAN(SF): Chittagong Hill Tracts Development Project, for $28.5 million, approved on 28 June 1979, contained one component for the
introduction of rubber planting by tribal smallholders.
Staff of the Agriculture and Social Sectors Department (West) suggest that the Project be classified as partly successful because (i) it was successful
in accelerating and expanding rubber plantation in lands unsuitable for food crop production; (ii) it generated private sector interest in rubber plantation;
(iii) it provided training to planters, tappers, and estate managers, who have provided significant help to the private sector operations; and (iv) it
assisted institutional and operational reorientation of BFIDC including decentralization of estate management. However, in the view of the REM, these
positive aspects were not commensurate with the high investment costs as evidenced by the negative EIRR and FIRR.