Economic data: How are the statistics gathered? 

By Madhusha Thavapalakumar

The economic data or statistics of a country usually reflects the true picture of that country whilst allowing international institutions to understand the country’s position in comparison with its regional peers. Earlier this month, national economic data of Sri Lanka for the first quarter of 2020 was released; it showed that the economy had contracted by 1.6%.

As the data was released following a highly unusual delay, this led to intense speculation as to the reasons for said delay. Eventually, it was issued only in the third quarter, a massive delay, even when accounting for the unprecedented disruptions caused by the Covid-19 pandemic. Many opined that the data was being held back intentionally till the end of the election, which was conducted on 5 August, so as to prevent any political disadvantage. However, the Department of Census and Statistics (DCS) and the Central Bank of Sri Lanka (CBSL) both denied this allegation, and attributed the delay purely to the pandemic.

This is not the first time Sri Lanka’s economic data, or the process by which it is calculated, has come under scrutiny. Economists, politicians, and even a CBSL Governor cast doubts over the accuracy of Sri Lanka’s economic statistics in the past. Therefore, The Sunday Morning Business decided dig deep and determine how economic data is gathered in Sri Lanka while taking a look at approaches and surveys used by Sri Lanka’s national statistics agency in collecting said data.

Economic statistics/economic data

Economic statistics or data are facts and figures derived from economic surveys undertaken by national statistical agencies. Economic statistics are imperative for a government to compile national accounts. They indicate the past and future trend of an economy and generate a detailed set of accounts including economic indicators.

These statistics are concerned with changes in reactions to the economic environment to households, businesses, non-profit organisations (NGOs), and the government. The statistics are shown as values and quantities as they deal with factors of production, distribution, and the prices at which goods and services are bought and sold.

How Sri Lanka collects economic statistics

The DCS – authorised data collector

Even though private firms and think tanks may collect and report on economic statistics, usually it is a government authority that officially releases the economic statistics of a country. In Sri Lanka, the DCS is vested with the authority to conduct national surveys, collect economic statistics, and compile national accounts (NAs) since 2007.

Forty-three branch offices of the DCS are situated in various ministries, departments, and other government agencies. The DCS has established 25 district statistical offices, located in the district secretariats. Each office is in charge of a deputy director/senior statistician/statistician who is assisted by a team of statistics officers. These officers are engaged in routine data collection and provide the DCS with a footprint at the lowest level of the governmental structure.

NAs and rebasing of base year

NA statistics compiled by the DCS include production, expenditure, cost components, distribution of national income, external transactions, supply and disposition, consolidated accounts on production and expenditure, capital finance, rest-of-the-world statistics, etc. Those statistics are released with a 75 days’ time lag. Until 2015, NAs were estimated using the year 2002 as the base year.

Under the previous Government, the DCS moved the base year from 2002 and 2010 and began publishing a series of annual and quarterly estimates on their website, with technical assistance from the International Monetary Fund (IMF) and Asian Development Bank (ADB).

In the rebasing process, several improvements were made to the Sri Lankan System of National Accounts (SNA) such as the adoption of international classifications and adoption of SNA 2008 compilation methodologies. Accordingly, from the first quarter of 2015, the NAs are compiled with 2010 as the base year.

Rebasing provides more accurate estimates of the size, movement, and structure of the Sri Lankan economy, incorporating new and emerging economic activities which were not accounted for in the previous system.

The United Nations Statistical Commission (UNSC) recommends rebasing every five years, although some countries do so at intervals of less than five years. In 1975, 1990, 1998, and 2002, only the base year was changed without any improvements made to the Sri Lankan SNA.  

Surveys to collect data

The Industry, Construction, Trade, and Services Division of the DCS conducts establishment surveys to compile economic indicators. These surveys are conducted annually and quarterly. Quarterly surveys are conducted in order to give quick estimates to compile the gross domestic product (GDP) and annual surveys give detailed information about the structure, nature, and performance by activity groups classified according to the International Standard Industrial Classifications (ISIC). 

Speaking to The Sunday Morning Business, DCS Director General Dr. I.R. Bandara stated that annual statistics are far more accurate than quarterly surveys as certain businesses are hesitant to provide data four times a year. However, she stated that the Department came up with a new web-based system where they give logins for businesses.

The stages of an establishment survey operation are as follows: planning stage, preparing questionnaires and instruction manuals, updating sample frame and selection of samples, printing questionnaires and instruction manuals, activities related to sending questionnaires by post, data collection and performing field duties, data editing, coding, computerisation, verification of data, preparation of error reports, calculating weights, tabulation of data tables, verification of data tables, preparation of final reports, and publishing of final reports. 

According to Bandara, these surveys are conducted both as manual paper-based surveys and through computer-based surveys and the computers at the Department are up to date with the latest technology. 

Methods of collection


The GDP is the market value of the final goods and services produced within a country in a given period. The GDP is a tool that is used to measure the size and the growth of an economy both by local authorities and international financial institutions. Economic growth is a percentage change of real GDP, and it measures the quantitative expansion of an economy. 

According to the DCS, real GDP is derived by removing the price effect from the GDP at current prices. There are three approaches to deriving the GDP value for a given period. They are the production approach, expenditure approach, and income approach.

National Consumer Price Index (NCPI)

Since October 2015, the NCPI is considered the official measure of inflation. The NCPI is compiled covering all areas of the country and published with a time lag of 21 days through a government gazette, media release, the newspapers, and the department’s website.

Colombo Consumer Price Index (CCPI)

The Prices and Wages Division of the DCS computes and publishes the CCPI of Sri Lanka on a monthly basis. In order to obtain the price information for the purpose of computing the index, the test purchasing survey is carried out with officers of the Prices and Wages Division in urban areas of the Colombo District.

The information on selected goods and services are collected continuously from selected outlets as per exact specifications. Using this information, the CCPI is compiled and published.

Producer’s Price Index (PPI)

For compiling the PPI, price information of 36 agricultural products is collected by the department staff in the field attached to the Divisional Secretariat; for the manufacturing sector, price information is collected from the 376 institutions; and for the utility sector, price information was collected from the Ceylon Electricity Board (CEB) and the National Water Supply and Drainage Board (NWSDB).

Customs collection of data

The statistical branch of Sri Lanka Customs is responsible for obtaining computerised import and export data on a daily basis, the preparation of relevant information of external statistics, processing data files of import/export and re-export data, correcting erroneous computer data files, keeping statistical data sales, and maintaining outlet efficiency.

External trade statistics are prepared monthly and issued to the CBSL, Treasury, various ministries and departments, local and international institutions, and researchers within the first 15 days of each month.


LankaSIS is the centralised statistical information service created and maintained by the DCS which offers a wide range of official statistics collected through censuses, surveys, special enumeration activities by the DCS headquarters and its branch offices, as well as statistical data extracted from various administrative records.

Statistical data that is collected by the DCS headquarters, as well as its branch offices, are usually maintained at different locations in different formats. Lack of a coherent system for efficient retrieval of data has therefore been a problem faced by data users which has made it essential for a one-stop shop for statistical data.

LankaSIS features easy and convenient search functions, a large range of easy-to-comprehend content for general users, and referential information to facilitate users with easy and fast access to information they need.

Doubts over accuracy

The former Government and the CBSL cast serious doubt on the accuracy of Sri Lanka’s economic indicators compiled by the DCS. During a press conference in October last year, then CBSL Governor Dr. Indrajit Coomaraswamy expressed confidence in the current methodology that was adopted by the DCS two years ago, but pointed out that the lack of resources was standing in the way of the methodology being utilised successfully.

“They need resources to conduct the surveys necessary to support the methodology, in the absence of which they have to use certain proxies and assumptions. Therefore, there are some disagreements between us and the DCS,” the Governor noted at the time.

This was the second time the CBSL questioned the accuracy of statistics compiled by the DCS. In December last year, Dr. Coomaraswamy cast doubts on the GDP growth rate of Sri Lanka, saying he believed it was higher.

“I don’t think this is a 3% growth economy. My view is that the growth is somewhere between 3.5-4% in 2018. The DCS adopted a new methodology some time ago, which needs a lot of surveys to support it. We don’t have the resources to undertake lots of surveys,” the Governor said at the time.

These comments followed similar sentiments expressed by then Non-Cabinet Minister of Economic Reforms and Public Distribution Dr. Harsha de Silva just two months prior.

In October 2018, Dr. de Silva, who was the State Minister of National Policies and Economic Affairs at the time, noted that the Port City was not being included in the final records.

“I have a big issue with the growth numbers coming out. There are a lot of activities happening. Are we recording them? The numbers are not reflecting what actually happens in the sector. I will give you an example – the Port City is not in the construction samples.”

As a result, the Government set up task forces to identify weaknesses, which mainly stem from the lack of human and technological resources. The task forces were given a time period, during which they had to find out the issues and recommend surveys that should be conducted to solve these. The task forces comprised the CBSL Director of Statistics, along with private and public sector officials.

However, earlier last year, speaking to The Sunday Morning Business, DCS Director General Bandara said she stands by the current economic data compiled by the DCS.

“I would not say our statistics are wrong. We captured the data and arrived at the GDP rate using our current methodologies. There are no problems in the system.”

Criticism over delayed data

Parliamentarian Dr. de Silva in July claimed the Government was withholding the first quarter GDP data.

Dr. de Silva stated they knew for a fact that the GDP growth rate was negative for the first quarter and, according to sources, the Government was either trying to influence the department to change the figure to a positive one or was trying to delay releasing the data until the general election is over.

He added that the data pertaining to the time period from 1 January to 31 March was to be released on 28 June.

The DCS stated that they generally release quarterly GDP estimates after completion of 75 days in each quarter and accordingly, the first quarter GDP estimates need to be released on 19 June 2020. However, this year, due to the impact of the Covid-19 pandemic, the DCS had reportedly faced many challenges in gathering data and information during the months of May and June from government, semi-government, and private sector institutions in the country. 

In addition to this, the DCS in a press release stated that data availability of many institutions and the quality of data were not at satisfactory levels due to operational impediments at the ground level as a result of the restrictions imposed by the Government and regulations provided by the Ministry of Health to control the Covid-19 pandemic in the country. Therefore, the DCS had to verify the data received from these institutions for accuracy.

The first quarter data was issued on 4 August, a day before the 2020 general election. As stated by Dr. de Silva, the first quarter GDP growth hit negative levels as growth had contracted by 1.6% compared to a growth of 3.7% a year ago. 

Need for a national policy

Bandara stated that there is a dire need for a national policy to ensure the most accurate and comprehensive economic data. Furthermore, she specifically pointed out that businesses either do not register or keep registering every year under new names in order to obtain concessions given only in the first year of business.

According to Bandara, this particular issue has a severe impact on the country’s informal sector data, and as such, emphasised the importance of introducing new methodologies.