conclusions and recommendations

Based on the information collected during this study, this final section of the report presents some conclusions about the quality of the existing economic and general management information available about the forestry sector in Suriname and makes recommendations about how this information can be improved and regularly updated in the future.

The quality of existing cost information

A considerable amount of detailed information about forestry production costs was collected during this study. This information covered most of the inputs required for all of the main forestry activities. The only serious deficiency was information about inventory and planning costs, which is difficult to collect because very few forest operations are currently carrying-out such activities in Suriname. The information collected showed a reasonable degree of internal consistency and validity. It also seemed reasonable in comparison with similar information collected by the author in other countries (see, for example: Whiteman and Scotland, 1998).

Information about forest industry costs was much more difficult to obtain. It is believed that this was not due to a reluctance of the industry to divulge this information, but rather that many sawmill owners do not appear to keep very good records of such information themselves. The little amount of information that was collected was not very detailed, but seemed to be reasonably consistent. However, it was suspected that this information might have been biased by the currently low rates of capacity utilisation in Suriname. In particular, the labour cost element of current sawnwood production seems rather high compared with other developing countries. This may, in part, be due to the small average size of sawmills present in Suriname.

In conclusion therefore, the forestry cost information presented in this report is probably the best that can be reasonably expected. It should be possible to collect better information on processing costs, but this would require a great deal of time and effort, which might not be a high priority at the moment.

The quality of existing productivity and product conversion rate information

Information about productivity rates and product conversion factors is necessary to convert the raw cost data collected in this study into unit cost data (i.e. cost per m³ of roundwood or product produced). Almost no reliable information was available about productivity rates in the forestry sector in Suriname. What was available all seemed to be rather low, for a variety of reasons, including: the small scale of many operations; weak planning and management; and other institutional factors (e.g. overmanning in parastatal companies).

It will be necessary therefore, to rely, to a certain extent, on assumptions about likely productivity in the calculation of economic rent. These can be based on the author's own experience of other countries and information collected from international sources, such as Caterpillar (1996). A critical factor in this analysis will be whether current working practices or more efficient practices are assumed in the analysis. There is likely to be considerable divergence between currently achieved productivity rates and what could be achieved with a more efficient forestry sector. Thus, it would be useful to carry-out the analysis under both sets of assumptions to examine the benefits of removing some of the obstacles to forestry development currently present in the sector.

The quality of existing price information

A significant amount of detailed domestic roundwood and forest product price information was collected during this study. The information showed a high degree of internal consistency and validity. The only potential problem with the data was that it was collected from a small number of producers (five sawmillers and four logging operations). However, the information collected is probably reasonable enough for the purpose of calculating economic rent.

LBB currently collect export volume and value information for all forest products legally exported from Suriname and export prices can be calculated from this information. This information contains a great amount of detail about types of product, prices and market destination. However, due to the current foreign exchange restrictions, imposition of ad valorem export tariffs and minimum export prices imposed upon forest products exporters, it is believed that most of this information is significantly biased downwards. This was confirmed by comparison with export prices achieved in neighbouring countries such as Guyana, where forest levies are structured differently. It is likely that it might be necessary therefore, to use price information collected from international sources to calculate what the real level of economic rent is in the parts of the forestry sector currently exporting roundwood and forest products.

The quality of general information about production and trade in the sector

Generally, the information available about forestry production in Suriname is quite weak. Starting with the forest resources, the potential yield of commercial species is unknown and serious questions remain concerning the appropriate harvesting regime (i.e. cutting diameter, growth and cutting cycle) which should be followed. The collection of roundwood production statistics largely relies on forest concessionaires submitting their production information to LBB, for the purpose of royalty assessment. It is suspected that the current figures probably underestimate production by at least 20% and almost no information is collected about production by species or quality, other than what is required for the calculation of royalty payments (i.e. production is divided into three broad species groups). Only slightly better information about the production of forest products is available. The 1990 sawmill survey collected some rough information about production from about half of the sawmills in Suriname and an attempt has been made to update some of this. However, as with roundwood production, this information does not collect much detail about species and quality. It also need to be systematically updated (and measured properly) on a regular basis.

Trade statistics are probably somewhat better. With the exception of illegal trade with neighbouring countries, the export volume information currently collected by LBB is probably reasonably accurate and sufficiently detailed to get a reasonably good impression about forest product trade. However, as noted above, the value information is probably quite unreliable.

Recommendations to improve and update cost information

The main weaknesses in the calculation of production costs is likely to be the lack of product recovery rate and productivity information and information about the costs of processing roundwood into forest products. In order to rectify this situation, it is recommended that the following activities should be considered:

Recommendation 1: A regular census of the sawmilling sector should be carried-out every three years. Every sawmill owner should be visited and information should be collected about: estimated capacity and utilisation (sawmillers' own estimates); the number and type of all major pieces of sawmilling equipment they have (by observation - to check on capacity estimates); and employment (sawmillers' own estimates - checked by observation). Any new mills established in between censuses should also be visited and added to the database of results when they start operations (and need not be visited during the next census). It is estimated that it would take approximately three months for two members of staff to design this census, collect the data and analyse the results. Priority: HIGH.

Recommendation 2: Using the sawmill census as a sampling frame, information about product conversion rates should be collected from a stratified random sample of sawmillers. Information should be collected by direct measurement of roundwood entering the mill and product production over a number of days at each mill. Five one-day visits to 15 mills would probably provide a sufficiently large enough sample to accurately estimate recovery rates by log size, species and type and quality of product. This information should be collected once, then be updated on a less intensive cycle than the sawmill capacity information. Updating could take place every six years or on a rolling programme of, say, three mills per year. It is estimated that the initial survey would take approximately five months for three members of staff (one data analyst and two data collectors) to design the survey, collect the data and analyse the results. Priority: HIGH.

Recommendation 3: Sawmills should be required to provide information on the annual volume of sales every year. It is unlikely that a survey at the end of each year would provide reliable results. Therefore, a postal questionnaire should be sent to a sample of mills every month asking them to record their volume of sales over the coming month. Information collected should include sales volume by species, type of product, and quality, but should not ask about sales revenue. By sending this to six mills each month, it should be possible to survey every mill once per year (a different month each year for each mill). The capacity and capacity utilisation census can be used to convert the monthly results into an annual estimate of production. It is estimated that this survey would take approximately two months for one data analyst to design the survey and analyse the results and three days per month to collect the data. Priority: HIGH.

Recommendation 4: Forest machinery productivity should be estimated using a small stratified random sample of forest concessionaires and independent loggers. General information about productivity should be collected using interviews with forest managers, then detailed productivity information should be collected using standard work-study techniques. Ten days at ten logging operations working in different types of forest and terrain and at different scale of operations, would give an adequate sample to produce work-study tables for most of the conditions currently encountered in Suriname. It is estimated that this survey would take approximately two months for one data analyst to design the survey, three months for four data collectors to collect the information (two teams of two) and three months to analyse the results. Priority: LOW.

Recommendation 5: Sawmill processing cost information should be collected from a small number of sawmills (say, five) and be used to model the costs of sawnwood production for a range of different mill sizes, capacity utilisation and product recovery rates and product types. This information should ideally be collected by a qualified independent expert (e.g. an accountant or economist with some forestry experience) examining the records of sawmills. It is estimated that that this survey would take approximately one month for one data analyst to design the survey, three months for one data collector to collect this information (two weeks at five or six mills) and three months to analyse the results. Priority: MEDIUM.

In addition to the above surveys to improve areas where existing information is poor, the cost information, which is reasonably good, should also be updated on a regular basis as follows:

Recommendation 6: Existing information on the cost of labour, consumable and capital inputs used in the forest harvesting and management sector should be revised every year by collecting such information from forest concessionaires, equipment suppliers and other suppliers of consumable items. It is estimated that that this revision would take approximately one month for one staff member to collect and analyse this information each year. Priority: HIGH.

Recommendations to improve and update price information

Information about prices is likely to be the most sensitive information to collect and also the information most likely to be deliberately mis-stated. Therefore, reliable information about prices can only be collected where this is given freely and without fear of later repercussions. It is recommended that price information should be collected and updated in the following three ways:

Recommendation 7: LBB should keep a file on domestic product prices containing published list prices obtained from sawmillers and advertisements in local newspapers. It is estimated that that this will be an ongoing activity and will not require significant inputs. Priority: HIGH.

Recommendation 8: Information about domestic roundwood prices should be obtained from casual discussions with individuals operating in the sector during the course of the other surveys listed above. It is estimated that that this will be an ongoing activity and will not require significant inputs. Priority: HIGH.

Recommendation 9: For export price information, LBB should supplement the information collected locally from exporters, with information from international sources. Any new price information from ITTO, FAO and MTC should be downloaded from their Internet sites, which should be checked every month. In addition to this, LBB should attempt to establish a simple library of this information (and any other relevant international price information that comes their way). It is estimated that that this will be an ongoing activity and will not require significant inputs. Priority: HIGH.

Although the above recommendations do not appear to be systematic, following them should ensure that recent reliable international product price information will be available when it is needed

Recommendations to improve general information about the forestry sector

Information about forest stocking will be collected when the National Forest Inventory is implemented. Similarly, information about roundwood production and forest stocking in individual concessions should improve when the SBB becomes operational. The detailed recommendations for information collection during both of these activities can be found in Cox (1998) and Mitchell (1998). Recommendations number one to three above should increase the quality of general information available about the processing sector. The only new suggestion (for general data collection) that arose out of the information collected during this study concerns estimation of growth and yield.

As noted in Section 3.5.1 Timber yields and cutting cycle, there is considerable uncertainty about what the level of growth and yield might be in Surinam's natural forest. Such information is usually collected from permanent sample plots, but this takes a considerable amount of time, can suffer from measurement difficulties and often reflects research conditions rather than typical operating practices in the forest. An alternative way in which growth any yield can be estimated however, is to examine stocking across a cross-section of forest areas which have been harvested at different times. This provides a rapid way of estimating regrowth and is more representative of the growth that might be expected after normal forest operations.

Estimating regrowth in such a way is, of course subject to another set of possible biases and measurement difficulties, the most common of which are being able to guarantee that the forest has not been re-logged since it was first harvested and being able to control for differences in original stocking and logging intensity at each site. However, if such factors can be identified and measured, then statistical techniques such as multiple linear regression analysis can be used to extract a simple estimate of growth. Discussions with the logging manager at Bruynzeel suggested that, in some of their areas, this might be possible. A preliminary estimate of growth might be estimated therefore, by attempting such an approach.

Recommendation 10: The possibility of identifying past logging areas which have not been re-entered and collecting information about the original harvesting intensity of these areas should be examined with Bruynzeel. If this information is available, consideration should be given to collecting stocking information from these areas and analysing this information with a view to estimating forest regrowth after harvesting. The inputs required for this sp ecial study can not currently be estimated with any great accuracy. However, based on past experience, such a study would probably require two to three months of forest inventory work for two to three teams, plus around three months analysis. Priority: LOW.

Information storage and dissemination

In order for much of the information collected above to be of greatest value, it should be systematically catalogued and stored for a reasonably long period of time. Much of it should also be disseminated widely throughout the industry in order to help with the management and development of the sector and show companies the results of their collaborative efforts. With respect to this, the following recommendations are suggested for further consideration.

Recommendation 11: Information about sawmilling capacity, capacity utilisation and product recovery rates should be stored on a simple spreadsheet database, which should be continuously updated as new information is collected. The database developed as part of this study could be taken as a starting point for this and could be modified as necessary, to include other information collected in the census and surveys. This can be done as part of the ongoing survey work recommended above and would not require any additional input. Priority: HIGH.

Recommendation 12: Other cost and price information should be stored on paper in referenced files for use as and when required. Reports on costs and prices should be produced from this information every time forest levies are revised (ideally every year) and these reports should be archived in a forestry library (see below). This can be done as part of the ongoing survey work recommended above and would not require any additional input. Priority: HIGH.

Recommendation 13: The FAO Project has collected a considerable amount of project reports and other publications. In addition to this, various sections of LBB also have their own reports and publications on a variety of subjects and the surveys recommended above should result in the production of several reports each year. Consideration should be given to establishing a properly referenced forestry library, which should be handed over to LBB when the Project finishes. With respect to this, an internationally recognised classification system for forestry publications is presented in Appendix 3. Responsibility for the maintenance of this library should be gradually handed over to LBB or SBB during the final year of the Project. In order that maintenance of this library is guaranteed, it will be important to ensure that one member of staff is given responsibility for the management of this resource. It is estimated that that the establishment of the library might take four months for one staff member and one month per year on average to maintain the library. Priority: MEDIUM.

Recommendation 14: Reports of the surveys recommended above should be distributed to all those taking part in the surveys. This information may be of use for the development of their companies and disseminating the information is likely to increase their willingness to collaborate. Such information may also generally promote the development of the sector and such reports should be distributed to other interested parties at LBB's discretion. Individual copies of these reports can be produced as part of the ongoing survey work recommended above, but additional resources will probably be needed to produce multiple copies and disseminate these reports. Priority: HIGH.

Recommendation 15: Consideration should be given to producing a simple publicly-available annual report on forestry in Suriname. This should contain summaries of the information collected in the surveys recommended above plus other readily available information about forest resources, the forest industry, production and trade, forestry policy and legislation and the Forest Service. Production of such a report should not involve much additional work (say, one month) but additional resources will probably be needed to produce multiple copies of the report. Priority: LOW.

The role of LBB, industry associations and NGOs

In many countries, several of the surveys recommended above would be carried-out either by private-sector industry associations or by the government in collaboration with such associations. In Suriname however, the forest industry associations do not currently appear to have the capacity to undertake such activities. It would be advisable therefore, to involve them wherever possible in the activities recommended above. This will not only develop their capacity to be of more use to their members, but might also increase collaboration in the collection of this information.

Another factor, which needs to be considered, is the commercial sensitivity of the information being collected. During interviews conducted as part of this study, most respondents appeared to be quite willing to provide information about their operations. If LBB were to start collecting this information on a regular basis though, respondents might start to raise concerns about what this information will eventually be used for. Careful thought needs to be given therefore, as to whether independent institutions such as the university should be used to collect some of this information.

These institutional issues should be addressed in the following ways:

Recommendation 16: The private-sector industrial associations involved in the forestry sector in Suriname should be involved in the development and implementation of the surveys recommended above. This does not, of course, mean that the Forest Service has to surrender any control over policymaking to these organisations, but rather that they should be involved in some way in the collection (and possibly the interpretation) of this information. This should not require any additional resources. Priority: HIGH.

Recommendation 17: Consideration should also be given to re-establishing regular round-table meetings with the industrial associations and NGOs. These meetings could be used to disseminate the results of these surveys and discuss additional qualitative information about general issues affecting the sector. This should require only minimal additional resources. Priority: MEDIUM.

Recommendation 18: Confidentiality may be an important issue in several of the recommended surveys. This should be discussed during the above discussion in order to identify areas of concern and potential solutions. If confidentiality does become a problem, LBB may have to consider employing independent institutions to collect some of this data. This may have serious resource implications. Priority: HIGH.

Training requirements

The staff of LBB Planning Bureau already generally appear to have most of the skills necessary to carry-out many of the activities recommended above. However, in view of the general level of staff resources in the Bureau, it may be necessary to utilise other staff members of LBB to collect some of this information. This will require that any survey materials produced are clear and easy to use and that LBB Planning Bureau staff have the capacity to train data collectors to help with the collection of this information. The following training is recommended therefore:

Recommendation 19: Staff of LBB who will be directly involved in collecting and analysing the information collected in the surveys recommended above should receive additional training in the following areas: survey planning and design; questionnaire design; data analysis and report writing. Much of this training can probably be accomplished through continued regular contact with the FAO Project and a little technical backstopping from Rome. This should be achieved as part of the ongoing survey work recommended above, which should be started during the remaining year of the Project. Priority: HIGH.

Capacity to implement these activities

The last (but probably most important) consideration is the question of whether LBB have sufficient resources to implement all of these recommendations. The total estimated staff resources required to establish and maintain the ongoing surveys recommended above come to 29 person months in the first year and eight person months thereafter. The additional special studies on forest productivity rates, forest processing costs and yield estimation might require an additional 42 person months to implement.

The initial requirement is clearly well above LBB Planning Bureau's current capacity (but could be achieved if additional resources within LBB were used to assist with data collection). However, the ongoing staff requirements are much lower and may be within the capacity of LBB Planning Bureau to sustain. This leads to the final recommendation below.

Recommendation 20: LBB and the FAO Project should consider which of the above recommendations should receive the highest priority and are within the capacity of LBB to sustain beyond the end of the project. LBB should then start to plan the implementation of these activities and institutionalise them within their annual programme of activities. Priority: HIGH.