Data revolution and Sri Lanka

For decades, the Sri Lankan Government didn’t have a data strategy to obtain actionable data. Even after periods of reform, data is still largely collected on paper or called in and entered into spreadsheets. Visits to several Government agencies would show you that most workers had impossibly large stacks of paper on their desks instead of computers

By Yehen Wijedoru

Functioning governments require large and growing amounts of data. Birth and death rates. Crime and weather patterns. Demographic and traffic changes. Governments since at least Roman times have always sought more and better data. 

Civil society organisations, the private sector, media outlets and governments look to official statistics for accurate and actionable information. Governments that are democratically responsive to their people want increasingly better information and will act upon the stories they tell. 

In an age of extreme weather, famine and food insecurity, rapid urbanisation, population growth, water scarcity, and protracted conflict, data innovation from every sector is needed to bring the latest expertise to bear on the world’s most serious problems. 

As they become wealthier and freer, societies invest more money in tracking important things. For example, in Myanmar there are approximately 500 people carrying out basic statistics countrywide, compared with approximately 5,300 and 16,000 in Vietnam (a bit richer) and Indonesia (middle income), respectively. 

Naturally, as new technologies emerge, the number of people may not be an accurate way to measure a country’s commitment to gathering accurate and actionable data, however it is important to ascertain Sri Lanka’s current manpower in this sector as a first step. 

For now, the collection of accurate and timely data, especially in the developing world, is often logistically difficult and expensive. In launching its second round of global goals – Sustainable Development Goals – the UN also formed the Global Partnership for Sustainable Development Data (GPSDD). Though noble, its call for a ‘revolution in data’ by addressing the “crisis of non-existent, inaccessible or unreliable data” is easier said than done. 

For decades, the Sri Lankan Government didn’t have a data strategy to obtain actionable data. Even after periods of reform, data is still largely collected on paper or called in and entered into spreadsheets. Visits to several Government agencies would show you that most workers had impossibly large stacks of paper on their desks instead of computers. 

Steps are being taken to have Government servants’ access to a computer and necessary training to use it, however there is more to be done in formulating a blueprint for an effective data strategy for Sri Lanka. 

Concepts that are used in the US or the developed world such as ‘data revolution’ and ‘big data’ are still largely foreign notions in Sri Lanka. Though we do indeed globally live in an era of ‘big data’ – according to IBM we create 2.5 quintillion bytes of data every day – Sri Lankan bureaucrats are more interested in acquiring and disseminating basic information that is readily available and relevant.  

Even when data are successfully collected and utilised, two major pitfalls await. The first is the potential for politicisation of the data. It is no surprise that data are used for political gain. The second pitfall relates to privacy

Adequate training for the senior government officials on the concepts of ‘big data’ and its importance will support the efforts in wide acceptance of a data strategy. Sri Lanka must first get ‘small data’ right i.e. the basic data collection skills across ministries and functioning Government departments need to effectively deliver services. Accurate population data in Sabaragamuwa Province. The number of third graders in most populous district of Gampaha. Dhengi rates in Dehiwala Galkissa municipality, etc. 

Certainly, some ‘small data’ can come from modern technologies such as remote sensors, cell phones, satellite information and digitising information. A recent proliferation of mobile data acquisition efforts makes direct citizen and customer interaction easier and cheaper than it was even five years ago. Such efforts are already producing important insights in hard-to-reach, data-impoverished areas, often before traditional statistical agencies. 

Though in a more nascent stage of development, satellite data has potential. From measuring the size of urban slums to tracking the ownership of rural farmland to disaster impacts and response, satellites have the ability to produce accurate data, as is the case with a joint NASA/USAID system SERVIR. More satellites are launched every year by an expanded group of countries and companies, which will inevitably lower the cost of data acquisition and thus increase its accessibility. 

But for these data to be useful, there first must be computer and internet access and basic numeracy skills for the folks who are going to interpret and package this data in forms that are actionable by public and private sector. A sound training mechanism in basic statistic skills to Government servants must be formulated; without these skills, even the simplest SMS surveys, satellite data, and new collection technologies will be futile.  Prime objective of a data strategy is to provide quality, accurate, timely and comprehensive data that can be translated into better policy by all agencies; public and private. However, to accomplish just that, Sri Lanka needs more than data itself. It needs better writing, research, and analytical skills for the staff that produces data. These are all achievable, but it will take time, effort and money. 

Even when data are successfully collected and utilised, two major pitfalls await. The first is the potential for politicisation of the data. It is no surprise that data are used for political gain. The second pitfall relates to privacy. Household level health, economic, and demographic data are crucial when determining future Government programs. This is also personal information where serious privacy concerns abound. 

In the US we are preoccupied with companies ‘data mining’ to sell us more targeted products. Imagine if information on your HIV status or annual income was publicly available; how closely would you want your privacy guarded? 

Efforts like Institutional Review Boards address some of these concerns, but they are typically only available to researchers at elite universities even in the US. Given that cyber-attacks are also an increasing threat – even to US security clearance data – Sri Lanka needs to put in place effective data governance framework to enhance Government official’s capacity to properly protect privacy. 

Responsible uses of data involve balancing political considerations and privacy concerns. Groups like the Responsible Data Forum offer excellent resources on how and when to use data. Ultimately, data should not only measure progress, it should inspire it. Collectors must balance the potential value of data versus the risks of misuse. 

On a global scale neither the UN, World Bank nor the US Government has a full grasp on how much is spent on improving data collection and utilisation in developing countries – a seemingly obscure but actually very important topic. 

The Global Partnership for Sustainable Development Data estimates around $ 650 million per year is needed to collect data, only $ 250 million of which is currently funded. Any future Sri Lankan Government has to ensure that the country allocates adequate funding for basic skills first and then look at context and technology fixes or techniques that increase the speed and ease with which governments, civil society, and the private sector can collect – and use – relevant data. 

New technologies and increasing political will for acquiring and using better data are signs of progress. But without the skills and infrastructure needed to utilise data, the ‘data revolution’ will remain an aspirational notion for many developing countries, including Sri Lanka.  

(The writer is Consultant Data Strategy and Operation – KPMG US.)