Credibility of Sri Lankan data questioned due to lack of transparency
View(s):Data on inflation, GDP and per capita income would be credible only if Sri Lanka’s Department of Census and Statistics (DCS) operates independent of political interference, two top statisticians have said.
S.S. Colombage and Dr. Anila Dias Bandaranaike, who both served as Director of the Central Bank’s Statistics Department before retiring at a higher level, shared this view during a recent, vibrant discussion on the ‘quality of data’ produced by Sri Lankan agencies like the DCS and the Central Bank (CB).
It was organised by the Centre for Poverty Analysis (CEPA) as its 50th Open Forum titled “Ensuring quality of survey data” and held at the Lakshman Kadirgamar Institute for International Relations and Strategic Studies in Colombo last week.
Dr. Indrajit Coomaraswamy, also a former Central Banker and Finance Ministry official, chaired the forum which unfortunately drew a small audience or about 30-35 people conspicuous in a large hall. Discussants and the audience were involved in a vibrant discussion on a crucial topic that would have interested many corporate CEOs, public sector officials and academics. Heavy rain during the day also resulted in many absentees.
“In the context of recent debates relating to national statistics, questions have been raised about the quality of data. These questions and resulting reactions from the general public have served to undermine the reliability of survey data, including survey based poverty data, such as those generated through the Household Income and Expenditure Survey (HIES), which is the basis for a substantial body of research in Sri Lanka. When the credibility of national data is called into question there is a knock-on effect on other research, which can lead to a general sense of distrust pertaining to research evidence. This unfortunately, can undermine the work of the entire research community in the country,” CEPA in a pre-session, briefing note to the media.
Prof. Colombage said a major drawback in the data (collecting) sector is the absence of a National Quality Assurance framework.
The DSC and the CB are silent on quality assurance and make no mention about the authenticity of data. Quality is not given a prominent place in these two institutions. Hence the credibility of data is questionable,” he argued.At the outset, Dr. Coomaraswamy said that data is important for companies and their business plans. If plans are charted on faulty data it would affect shareholders, workers, stakeholders and consumers, he said.
The financial crises in Greece and Argentina were partly due to manipulating data. “In both these countries no one knew how bad things were until it was too late; people didn’t know, business didn’t know, policymakers didn’t know, international financial institutions didn’t know and rating agencies didn’t know … and suddenly (all because of the data) the whole thing blew up.Unfortunately the people who suffer the most from (policy) adjustments to these crises are the poor,” he said.
Prof. Colombage said that quality assurance was essentially for statistics as there are doubts about GDP and inflation not only in Sri Lanka but most countries.
One example of data being questionable is the case of Sri Lanka’s poverty index which has fallen drastically in the past several years, he said. In 2002 it was 23 per cent; then it fell to 15 per cent in 2007, 8.9 per cent in 2010 and 6.5 per cent in 2012.
He noted that poverty reduction should be supported by economic growth but in this there was no evidence to show a substantial increase in growth between 2002 and 2007.
Growth accelerated to 7.3 per cent only after the war ended but the growth in the earlier years against poverty level figures showed unexplained patterns. “I am not saying this data is wrong but there are some inconsistencies,” he noted.
These inconsistencies could also be applied to inflation and unemployment data, he said.
While traditionally data was all about accuracy, the trends have changed and now relevance, timeliness, acceptability of data, efficiency, etc are equally important.He lamented the lack of quality assessment/reporting saying that Sri Lankan statistical surveys need to have quality assurance mechanisms (with internal and external input)
In her presentation, Dr. Bandaranaike referred to a old World Bank saying that “without good statistics the development process is blind, policymakers cannot learn from their mistakes, and the public cannot hold them accountable” to stress the need for transparent mechanisms with accountability.“As users we need credible data for decision-making. The world is turning more and more towards evidence-based, decision-making. We need the information to assess current conditions critically whether it is domestic violence, income, levels of poverty. We need good data to better manage our resources,” she said, adding that there are best practices and systems that Sri Lanka should follow.
“Data needs to be accessible and accessible at the same time. It should not be ‘contaminated’ (doctored or altered) by anyone else (outsiders including government officials) before being available to all (public),” she stressed.
The former CB statistician, often outspoken in public forums on data issues, said while users of data (public, academics, others) have rights, they also have responsibilities.
While the rights deal with the transparency of the methodology, receiving independent and unbiased data and a predictable system of dissemination; responsibilities of users mean they need to be aware of the best practices, compare the data with these practices, ensure checks and balances on data and lobby for access to the data.
She said globally there is an information revolution which has led to an information- overkill: a massive load of data being available on TV, interviews, Internet, blogs, etc. “Because there is so much information, it is very difficult to check the sources of the information. For example there are miracle cures for cancer, diabetes, etc which are not necessarily supported by hard evidence. Who checks this? Where are the sources? We need to be careful (about believing such information),” she said.
Dr. Bandaranaike said private companies that crashed during the financial crisis had beautiful, glossy annual reports and paid more attention to “form (beauty) than substance (quality of information)”.
She said research papers often don’t have facts to support their theories.
“There are vested interests (in such research). Pharmaceutical companies get research reports supporting their ‘cause’ issued by paying huge fees,” she said urging the need for ‘data with integrity’.
She said the DCS should not come under the Ministry of Finance. “That is not good; they cannot be (fiercely) independent. They should be outside (the political structure). Independent institutions should not work in the interests of the policy makers,” she said.
She said ‘users’ should demand accountability and governance structures from ‘producers’ (government and their data agencies)
“There is no need for polarisation. Users and producers must work together. Objective criticism should be allowed. Unfortunately the Government is defensive and doesn’t create the space for dialogue,” she added.
During the discussion, Wimal Nanayakkara, retired DCS Director General and presently Senior Visiting Fellow of the Institute of Policy Studies, sprang to the defence of his former department saying during his time they had independence in data collection and processing even though it was under the Ministry of Finance.
He said he believed the same position applies today.
A CB staff official, present in the audience, responding to a question, said officials could freely express views on any issue.
“We have prepared staff studies where we have said that the fertiliser subsidy is not sustainable and explained why. But whether the government takes action is a different policy scenario which we don’t have much control of. We have also discussed the losses in the electricity sector and the need to raise prices (rather than provide huge subsidies),” he said.
| Recent fiasco on GDP figures was due to a misunderstanding
The recent debate on supposedly ‘doctored’ GDP figures which led to the suspension of a senior DCS official who refused to obey a directive by the Director General, was due to a lack of misunderstanding and communication, according to former CB Statistics Director Anila Dias Bandaranaike. She told the CEPA Forum on the ‘credibility of data’ that there were genuine reasons for changes in the data but this message was not properly communicated to the public. “Sri Lanka is going through a structural change economically and there is a need for compilers of GDP to re-assess the baseline which takes time and till such time the baseline survey is done, estimating GDP on the basis on the old structure which is no longer valid, is difficult. Thus there are problems. So when the GDP estimates come out there are certain issues and changes. Unfortunately the DCS didn’t explain this to the public, and when changes were noticeable, the public made a hue and cry and there was no proper explanation offered,” she said. |