4. IMPROVEMENTS OF NATIONAL STATISTICS PROCESS
Based on a review of the processes involved in data collection by the Forestry Division and the Central Statistical office on wood products it is clear that there is an urgent need to address certain issues to ensure that accurate and reliable information are collected and disseminated. The following are recommendations to deal with the issues encountered.
4.1. Issues relating to the forestry division
4.1.1. Issue 1
Monthly log returns and quarterly log return are not based on the same data set. This leads to in accurate figures.
Recommendation
The monthly and quarterly returns must be based on the same field data set. To enable this, the Forestry Division staff must ensure that the data submitted for each reporting period must be based only on all trees, which were released.
4.1.2. Issue 2
There are no systems in place for data verification from the field.
Recommendation
It is recommended that a random selection of a range be made at least once per quarter and a thorough check of all field data be made to verify for accuracy and completeness. The output data from this exercise will be used to compare the quarterly data, which were submitted for this range.
A routine validation of the data from each range must be made at the supervisory level in the Conservancy. This can be achieved by proper supervisory checks.
4.1.3. Issue 3
The field data are prepared on a monthly and quarterly basis at the range level and submitted to the FRIM section. The FRIM section summarises the production on a quarterly basis and produces only an annual report from the collated figures. There is no verification of this collated data.
Recommendation
The collated quarterly summary by FRIM should be submitted back to each Conservancy for editing and verification before submitting to any agency.
4.1.4. Issue 4
The data on sawnwood production is based on annual returns from sawmills and these returns are inaccurate based on many factors outlined in Section 3.2.
Recommendation
The Forestry Division holds two one-day workshops to explain to the sawmillers their role in the supply of accurate data for output of sawnwood. Regular field checks at sawmills by officers to enforce the law must be made. It should be made mandatory that sawmills submit a quarterly sawmill production report.
4.2. Issues relating to the CSO
4.2.1. Issue 1
The volume data collected by the CSO on trade in wood products is inaccurate based on custom declaration forms.
Recommendation
The Forestry Division holds a one-day workshop to assist custom brokers, custom officers and the CSO personnel to address inaccurate volume data. In addition to this, the Forestry Division should produce a simple conversion chart for use by Custom officers to enable accurate and consistent conversion of data.
4.3. FAO Yearbook on forest products
Since the data were not properly edited some errors have occurred and these are as follows: For example, imports of round logs from Guyana have been classified as coniferous round logs. Product codes have changed for the same product from year to year causing a problem when following the same product over the review period. Some entries were made in the wrong column e.g. the volume were entered in the value column and the value data in the volume column. Where possible average prices for a particular product were used to correct for inaccurate data. In addition, the personnel who supplied the information from the Forestry Division to FAO also got confused between roundlogs, sawlogs and sawnwood production resulting in some incorrect data. The raw uncorrected data were extracted from the FAO Yearbook and are presented in Table 7. This data was corrected with data from CSO and are presented in Table 8.
The data in the FAO yearbook on forest products as it relates to Trinidad and Tobago are as accurate as the input data. However, as pointed out in this report, there are serious inaccuracies in the gathering and verification of the data. It is therefore reasonable to conclude that the data in the FAO yearbook will contain errors. It is suggested that a base line date (2001) be set after which accurate and verifiable data will be input to the FAO yearbook.