Dear Editor,
I cannot but concur fully with Prof. Thomas’ view: “… apart from poor data quality and lack of timeliness … I shall argue today and continue next week, that, data manipulation and deception by the Authorities (local and international), have been equally at fault in making Guyana’s national accounts data too weak for rigorous analysis.” I, too, have addressed the issue of data manipulation in pieces written for the Caribbean New Yorker between 1998 and 2003, but perhaps not with the rigour that Prof. Thomas deploys.
My purpose here, however, is to lend a comparative perspective to the issue of data quality, excluding manipulation. The latter is most difficult to gauge, but I believe it is very high in Guyana. Data manipulation in Guyana is largely a post-independence phenomenon begun by the administration of Forbes Burnham and has continued at least until the end of PPP administration, which was kicked – deservedly – out of office earlier this year.
The World Bank has developed recently a brave and innovative measure called the “Statistical Capacity Indicator” (SCI). Data are available for 154 developing countries and regions. The CSI “is based on a diagnostic framework developed with a view to assessing the capacity of national statistical systems using metadata information generally available for most countries, and monitoring progress in statistical capacity building over time. ¬The framework has three dimensions: statistical methodology; source data; and periodicity and timeliness.” Based on available information from the World Bank, IMF, UN Secretariat, UNESCO and WHO, a country is scored on each of the three dimensions and a dimensional composite score is derived on a scale of 0-100. A score of 100 indicates that a country meets all the criteria; a score of zero means that a country meets none of the criteria.
This multi-dimensional approach “is based on the notion that producing and disseminating reliable, relevant and timely statistics requires a certain level of capacity in all dimensions. Any imbalance would point to weaknesses in some aspects of the statistical process.” If, for example, a country reports a high level of periodicity and timeliness but low levels of statistical methodology and source data, this could suggest that indicators are not derived using recommended methodologies and timely source data. Such an assessment draws attention to data quality issues and areas that need improvements.
The statistical methodology dimension gauges a country’s ability to adhere to internationally recommended standards and methods. The score allocated is based on an assessment of guidelines and procedures employed to compile macroeconomic data, and estimation practices. According to the World Bank, a country is “evaluated against a set of criteria.” These include use of the latest Balance-of-Payments Manual, use of an updated national accounts base year, subscription to IMF’s Special Data Dissemination Standard, enrollment data reporting to UNESCO, and external debt reporting status. The second dimension, sources of data, measures whether a country “conducts data collection activities in line with internationally recommended periodicity, and whether data from administrative systems are available and reliable for statistical estimation purposes.” Specifically, the World Bank uses the periodicity of population and agricultural censuses, the completeness of vital registration system coverage, and the periodicity of poverty and health related surveys. The final dimension is that of periodicity and timeliness. It is concerned with the availability and periodicity of key socioeconomic indicators, nine of which are MDG indicators. The goal of this dimension is to “measure the extent to which data are made accessible to users through transformation of source data into timely statistical outputs. Criteria used include indicators on income poverty, child and maternal health, HIV/AIDS, primary completion, gender equality, access to water and GDP growth.”
Chart 1 shows how Guyana scores on overall statistical capacity compared to Latin America. During the eleven-year period from 2005 to 2015, Guyana’s average score was 55.76 out of 100.0 compared to 74.17 for Latin America and the Caribbean (LAC). Between 2005 and 2012, Guyana’s overall statistical capacity remained basically unchanged, while that of LAC expanded. For the next two years, 2013-14, the country’s statistical capacity seemed to have increased by about a score of 3, but declined thereafter. In 2015, Guyana’s overall statistical capacity was not different from that of 2004, while that of LAC was slightly better. During the entire period Guyana’s average capacity gap was 19.13 lower than that of LAC’s. Guyana’s overall statistical capacity, as measured by the CSI, is better than that of Haiti (41.62), Afghanistan (43.89), Angola (45.83), Antigua and Barbuda (46.67), and Dominica (53.43). It is worse than that of Bangladesh (86.94), Belize (60.20), Bolivia (70.28), Cambodia (71.67), Honduras (68.61), Nigeria (64.07), Pakistan (77.31), Nepal (69.26), India (78.06) and Zambia (59.72). More to the point, Guyana’s overall SCI was worse than 138 of the 154 countries and regions for which the indicator has been calculated.
Which of the three dimensions performed the worst? The sources of data dimension scored lowest: an average of 36.36 compared to 71.25 for LAC. Guyana’s average score on this dimension was helped by higher scores during 2007-2011, which declined thereafter. Capacity on this dimension was the same at both the beginning and ending of the period. Next comes the methodology dimension, which scored an average of 45.45 during the period compared to 65.34 for LAC. This dimension shows slight improvement with time, but that is still inadequate. About 46 percent of the 154 countries scored less than 50 on methodology dimension, with Guyana, Papua New Guinea, Kenya, Paraguay, and Zimbabwe registering a score of 45.0 to 45.83. Guyana is ranked 96th on this score during 2004-2015 on this dimension.
In terms of the periodicity and timeliness dimension, Guyana’s score was not much different from that of LAC – 85.45 and 88.07, respectively. Guyana has done well on this dimension compared to many other countries. Its score (85.45) ranks it 54th of the 154 countries. Finally, Guyana’s score on sources of data was only better that 20 of the 154 countries and regions – it was ranked 135th These averages are only meant to draw attention to gaps, and any analysis of data by countries would have be much more nuanced.
When “data manipulation and deception by the Authorities” is added to Guyana’s very limited statistical capacity, the output can only be very substandard; probably worse than I thought prior to this analysis.
Yours faithfully,
Ramesh Gampat
