Data-driven decision-making for economic prosperity and good governance – II By Dr. Ranga Prabodanie
By Dr. Ranga Prabodanie
(The first part of this article appeared yesterday (05)
The previous part of this article series explained how the great insurgence of digital data has revolutionized the institutional decision-making process in both business and governance. This latter part will look at the Sri Lankan context: Where we are currently and the way forward to a data-driven decision-making culture. Let’s first have a glance at how decisions are made in data-intensive public services.
Schools, universities and vocational training institutions, throughout the country, collect, record and report data on admissions, enrolment, completion, drop-outs, results, graduations and expenditure for various programmes, courses and subjects. The Examinations Department does have digital records of GCE O/L and A/L results. However, the data is used only for the preparation of annual statistical reports, rather than for identifying and resolving problems. Data-driven decision -making extends well beyond preparation of reports. To make informed decisions on new subject streams, curriculum revision, subject promotion, funding and resource allocation, education data should be explored for trends and associations which raise concerns. To ensure equity in resource allocation, inputs and outputs, produced by education institutions, should be compared using appropriate analytical methods. Analysis of data from industry and other stakeholders is also important to identify the skills in demand and academic disciplines with greater potential for entrepreneurship, employment and scientific innovation.
Public health records which include data on patients hospitalisations, symptoms, diagnosis, treatments, progress and side-effects, encompass valuable insights on emerging diseases, their causes, case rates, recovery rates, death rates, hospital congestion, bed occupation, treatment costs, waiting times, medication efficacies and vulnerable groups which could provide directions for healthcare management and budgeting. There is an urgent need for health records to be digitised and analysed to glean emerging trends, patterns and associations. Though some progress has been made in healthcare analytics, for example in understanding the drivers of Dengue outbreaks, the patterns, trends and socioeconomic implications of most widespread diseases are poorly understood.
Lack of competence in health data analytics was evident in the handling of the COVID-19 pandemic. There were instances where even the number of deaths resulting from COVID-19 were misreported and revised. We have seen various professionals, making claims, in the media, such as “next two weeks are critical”, “the country should be closed for three weeks” and “people are dying at home because they avoid going to hospitals”, but what we never hear is, based on what data, over which time period, analysed using which methods, such insights were derived. Sri Lanka has a well-educated population capable of differentiating facts supported by data from mere human perceptions, and perhaps, that is why they do not listen to such claims.
If you listen to the BBC news, you would often hear them reporting the status of COVID-19 in the UK directly citing the source as “according to ONS (Office for National Statistics) estimates”. Then they may present the information with relevant comparisons as, for example, “COVID-19 was the third leading cause of death in England and Wales in September 2021, accounting for 6.6 percent of registered deaths in England and 8.5 percent of deaths in Wales. The two leading causes of death in both countries were ….” In Sri Lanka, we rarely hear such alarming comparisons,based on real data, but poorly supported individual projections, based on intuition. If data-driven analytical outcomes were shared with the public, people would be compelled to listen. Given the rising healthcare costs and economic depression, it is high time to invest in professional health data analytics to understand the trends and associations, to establish the right priorities, and to inform policies, accordingly.
Agriculture is another sector in which data driven decision-making can make a revolutionary change. Some recent developments in the country, including alleged hoarding of rice, milk and sugar to create artificial shortages and to increase prices, are related to lack of reliable data on agricultural production and imports. To avoid such malpractices, particularly in times of crisis, the government authorities have to keep track of data and continuously update and analyse the data to understand the drivers of market demand and supply. Since the government provides fertiliser subsidies, the agricultural authorities should have data on the acres of food crops to be harvested each season. If there are barriers to obtain reliable data, there is technology to help. A research group in Stanford University has developed a scalable yield mapper which can predict crop yield at the field scale based on satellite data. The system has been tested not only in the US but also in Africa and India. Development or acquisition of such technologies would help authorities to monitor the production and supply of food crops and make informed decisions on subsidies and imports.
The government decision to ban agricultural chemicals came under huge criticism as a poorly informed decision. Given the global appeal for organic food, the ban on agrochemicals can have favourable impacts on our economy and wellbeing. It could have secured a competitive advantage for Sri Lankan food products in the global market. Unfortunately, the decision came as a surprise, without supporting facts derived from real data. The evidence on the associations, if any, between water pollution and agrochemicals, chronic kidney disease and agrochemicals, food prices and agrochemical imports, organic fertiliser and agricultural output, and other relevant and measurable factors, should have been elaborated together with predicted outcomes of the decision, both positive and negative. Decisions that are apparently not supported by facts indicate lack of transparency and accountability, a basic principle of good governance. Lack of data-based reasoning can create chaos irrespective of whether a decision is right or wrong.
Continued monitoring of crime data is essential for ensuring public safety. Crime data analysis can reveal spatial and temporal patterns of crime, trends, hot spots, vulnerable groups and delinquents. Such insights can inform resource allocation for crime reduction and prevention. The general public is constantly bombarded, by the media, with fresh crime data, such as “a suspect possessing X grams of ICE was arrested in Y”, which now has no significance to the general public. Instead, if the media reports crime trends as, for example, “X percent of the suspects arrested with illegal drugs in 2020 were adolescents in the Y-Z age group…”, it would immediately trigger the attention of parents, schools and other stakeholders. The former is raw data which the analyst has to work with and the average citizen has little to do with, while the latter is one of the insights derived from data which should inform decision making and policy response and thus matters to everyone.
The previous sections of this article pointed out only a few areas of business, public service and governance which can be enhanced via data-driven decision -making. There are several other sectors, such as investment, energy, transport and conservation where data-driven decision making can make a shift towards sustainable development and better living. As a viable starting point, available public service data can be digitized and made available for analysis by researchers and relevant experts. Countries like the UK, the US and Australia have made health, crime and other data available on the public domain, allowing the researchers to explore the data and inform the government. However, a strong policy framework is needed to support, promote and facilitate data-driven decision-making in all those sectors.
Barriers should be expected, and initially, it would be more difficult to change attitudes than to set-up the basic infrastructure. The biggest barriers could be institutional bureaucracy, political influences, special interest groups and disruptive intentions motivated by the fear of losing power, status, and prerogatives. Strong leadership with a sound understanding of the need for evidence-based decision making is essential. Leaders have to understand that the status reported by officers and various parties with vested interests do not always reflect the reality on the ground and hence decisions made on such advice can lead to disasters. Real data is the only dependable and reliable source of ground reality which should guide policy.
The Sri Lankan government has already taken the initial steps to digitize public service data by establishing the Information and Communication Technology Agency (ICTA), committed to implementing digital-governance in Sri Lanka using ICT to access, exchange, and utilize information efficiently. In collaboration with ICTA, some government institutions have taken progressive steps towards data-driven decision making. The Department of Immigrations and Emigration, the Department of Motor Traffic and the Election Commission of Sri Lanka have already introduced online services which autonomously collect and store data in easily analysable formats. Still we have to develop a policy framework and a culture which supports regular analyses of collected data to generate insights and integrate them into the decision-making process.
Gone were the days when institutional decision making was an exercise of sheer authority; today it’s a complex process of collecting, analysing and generating insights from data. People no longer accept mere predictions without well elaborated facts and evidence, nor do they hesitate to challenge poorly informed decisions made on sheer intuition or authority. The data revolution is on-board, demanding all policies, regulations, restrictions, grants, expenses, and all kinds of decisions to be justified by facts and science. Everyone in business, governance and public service will have to change their attitudes and come to terms with the new decision-making culture driven by data and insights.
(The writer is a Senior Lecturer at Wayamba University, Sri Lanka. However, the views and opinions expressed are those of the writer and do not reflect the policy or position of any institution.)