Dear Editor,
I am delighted that the renowned economist Professor C.Y. Thomas has decided to address “matters concerning Guyana’s development” (SN, 18 October). No one in the world is more qualified to do this than Dr. Thomas, who taught me at UG and a man for whom my respect only grows with time.
In the referenced piece, Prof. Thomas raised the issue of the quality of Guyana’s data. Perhaps all data (economic social, health) in the country’s national accounts are poor in quality, unreliable and untimely. As someone who has a quantitative background, I am keenly aware of data methodologies and data quality. In my PhD thesis, which was done in 1994, I raised the issue of data quality and timeliness (Annex 4, p. 382-83). I returned to the issue of data quality in a 2000 paper on inflation in Guyana, where I wrote: “it is well known that the economic (and even more so, social) database for the Guyana economy is inadequate and unreliable …” (Transition 2000:68). I also address the issue of data quality in several articles published in the Caribbean New Yorker between 1998 and 2003. And in April this year I returned once again to the issue in my two-volume book on the health system and major diseases in Guyana. Here’s what I wrote in Volume 2, pages 688-89.
Concern about the quality of Guyana’s socio-economic data has been a long-standing issue. The World Bank, for example, noted in 1986 that “since 1981, the quality of national accounts … has deteriorated considerably.” (World Bank 1986: 73). In a comprehensive review of Guyana’s public sector, published in 1993, the World Bank returned to the issue: “Demographic and health indicators in Guyana are limited, often dated, and of dubious quality.” A few sentences later, it adds “… it is widely recognized that a large percentage of the births and deaths go unrecorded, particularly in the rural and hinterland areas” (World Bank 1993: 238). More than a hundred years since vital statistics began to be collected in the country (1869), a portion of births and deaths continued to escape the official net and thus even births and death rates are not entirely accurate.
Even today, more than two decades after the World Bank recorded its dismal observations, the quality of the country’s data has not improved much. Recently, the Pan American Health Organization (PAHO) pointed to the “very serious problem” with Guyana’s health data: “A further challenge for Guyana lies in obtaining up-to-date, accurate, and consistent health data on the same indicators from both in-country and out-of-country data sources; there is need for strengthening the efficient and accurate reporting of public health and related data, and in consistently and objectively evaluating health interventions. Various national and international sources may issue differing data for the same indicators in the same time period. International agencies often rely on estimated values where there are perceived data gaps, non-standardized indicator definitions, or uncertain methods of data collection and analysis at national and subnational levels” (Pan American Health Organization (PAHO) 2012a: 375).
An example serves to illustrate the point. In its 2003 report titled Health Sector Analysis. Guyana (draft), PAHO notes that the infant mortality rate for 2007 was officially estimated at 25.5 per 1,000 live births. However, data obtained through vital registration yielded an estimate of 54 per 1,000 live births. Two years earlier UNICEF arrived at the same conclusion: “The measurement of infant and under-5 mortality, using vital registration data in Guyana has produced quite different results. For 1997, the same reference year as the estimates from Guyana MICS, Infant Mortality was calculated at 25.5 per 1000 live births and under-5 mortality at 31.8 per 1000. This underscores the perception that the reported levels of infant mortality derived from civil registration data was too low because of the under-registration of infant deaths and inadequacies of the administrative and institutional arrangements of the current vital registration system” (United Nations Children’s Fund (UNICEF) 2001: 19).
According to a 2011 USAID health assessment report, the Ministry of Health regularly collects a large amount of data at the national and regional levels. “However, there are challenges with the reliability of the data: there are few horizontal linkages, and data validation and feedback is lacking” (United States Agency for International Development (USAID) 2011: 91). Data quality could be improved by ensuring that data sets are consistent and conform to requirements for decision-making.
Aside from the quality of the data, timeliness is also a concern. The latest statistical bulletin, a compendium of health data, was issued by the Ministry of Health in December 2012, but it related to data for 2009. Indeed, published health data are difficult to come by, especially if one is looking for a time series. Fortunately, PAHO’s website contains data on several health indicators and individual diseases over a relatively long period of time. Nor are there regularly published analyses of health conditions by the Ministry of Health. Once again, PAHO comes to the rescue with its Annual Health In the Americas, which, contains, among other things, a county analysis of each country in the Americas.
Despite the above, it remains true that, while the quality of some health data are not up to accepted standard the broad trends these data suggest are probably accurate. But trends are not entirely sufficient. To be useful for informed decision-making, reliable, timely and relevant data are of cardinal importance. Some of these issues are being addressed: “significant steps have been taken to advance HIS [Health Information Systems] in the ministry; data quality has improved and will continue to improve,” according to USAID (United States Agency for International Development (USAID) 2011: 81). As to what extent the quality of data has improved and whether they provide an adequate base for decision-making, I cannot say.
In view of the above, I fully agree with Prof. Thomas that Guyana’s data is too weak to be subjected to rigorous analysis. However, the available data that we have is all we have, and some data is better than none, if only because they draw attention to glaring inadequacies. But if the available data allow for broad trends, that is a good thing; precision to the last decimal point is not necessary. However, I hope the discussion of the data and what can be inferred from them will motivate the authorities to allocate the necessary resources to improve the quality and reach (disaggregation by region and ethnicity, for example) of Guyana’s data.
Yours faithfully,
Ramesh Gampat