Data for policymaking in the pandemic period: Bangladesh experience

The ongoing COVID-19 pandemic has given rise to a range of adverse socio-economic and health impacts. These fallouts have exponentially increased the demand for reliable data - coming from healthcare professionals, policymakers, and the public at large. Indeed, data deficits are inhibiting the government from pursuing necessary and effective public policy actions, including vaccination drives, to combat the scourge. In this context, it is critically important to understand the process of generating basic statistics in view of COVID-19 in Bangladesh. It is also important to assess how far these statistics are translated into policy responses by the government and how data gaps have posed challenges in terms of policy actions by the government.

Since the pandemic started, global organisations such as UNStat, World Bank, Eurostat have shifted their priorities online. They have conducted surveys to assess the state of National Statistical Offices (NSOs) and developed online platforms in the form of websites and data dashboards presenting relevant COVID-19 statistics (e.g. 'COVID-19 Data Hub' by UNStat). They have also developed new tools to support the continuity of NSO activities (e.g. 'PARIS21 E-learning Academy') and facilitated peer-learning among the key stakeholders of the global data ecosystem through virtual sessions.

COVID-19 targeted data initiatives by countries around the world placed greater emphasis on health. They have favoured digital solutions to the rising data challenges. There was greater concentration on contact tracing through online based surveys and Bluetooth tracking. South Korea coordinated with relevant agencies to obtain credit card data, location data to track the movements of individuals. These methods have been controversial. Ghana, however, used alternative data sources to monitor the effectiveness of enforcing lockdowns by obtaining call detail records from telecom operators. Additionally, Ghana's focus has also been on establishing COVID-19 data initiatives that address the pandemic's impact on the country's local business, employment, and households.

THE BANGLADESH PERSPECTIVE: Since the beginning of Covid-19, the Directorate General of Health Services (DGHS) of Bangladesh, in collaboration with IEDCR, have undertaken numerous health-related data initiatives such as COVID-19 Dynamic Dashboard for Bangladesh, COVID-19 Situation Related Health Bulletin, Surokkha Web Portal. But this sheer number of data initiatives can be quite overwhelming for the uninitiated. The absence of a central data repository is a significant shortcoming of these initiatives. Besides, the majority of the health data initiatives focused on preventive and curative aspects of health care. They do not reflect the rehabilitative aspect of health care. Lack of coordination between different data initiatives is also visible. But there are positive aspects as well.

In terms of regularly providing up-to-date status of key variables of interest, the COVID-19 targeted health data initiatives exhibited considerable improvement compared to the pre-pandemic scenario. Data dissemination also took a more multipronged approach compared to the pre-COVID-19 situation. To address the COVID-19 crisis, Bangladesh used these data initiatives in implementing various public policies, notably, the "BDT 2,500 Cash Support Programme" and "Surokkha Vaccination Programme". 

However, since these developments occurred in online platforms, they have created some concern of a 'digital divide'. The requirement of internet facilities to register for vaccines left out a large number of eligible people. The requirement of NID to register also created challenges for the younger population or for those who have not obtained a NID due to prior circumstances. Transgender people also faced similar difficulties as many of the community do not have NID. As the demand for vaccination increased, the Surokkha portal could not handle the pressure of the volume of registration, resulting in frequent downtimes, which discouraged many to register.

The ongoing pandemic has adversely impacted the regular data-related activities in Bangladesh, starting from data collection to dissemination. The pandemic has forced a number of large-scale surveys like the Population and Housing Census, Labour Force Survey (LFS) to be rescheduled because they require face-to-face interaction. Activities pertaining to regularly provided data were continued despite the pandemic. Apart from these, the data collection processes of some other initiatives have been hampered, such as the literacy assessment survey, surveys on climate & tourism.

In the face of the pandemic, there have been institutional data initiatives. But they are yet to evolve into a structured and regular system. The pandemic created an opportunity to build a partnership between government and non-government entities. A number of innovative data initiatives taken in view of the COVID-19 pandemic were built upon pre-existing data generation and management systems (doing by learning). Others had to be developed from the ground up (learning by doing). The utilisation of user-generated data and telecom data also gained heightened importance.

The institutional arrangements are concentrated on health and have not spilt over to other areas of the data ecosystem. Government and non-government entities took up multiple roles and carried out the COVID-19 targeted data initiatives in a coordinated manner. The acceptance of non-government data initiatives to the government has gradually increased. While most of the initiatives had a 'government agency centric' approach, some adopted the 'whole of society' approach.

Besides, the documenting and archiving process is mostly scattered and accessing micro-level data may become an arduous task. The majority of the data initiatives, both government and non-government, have well defined and formal storage systems. But data access has become difficult due to complex and lengthy bureaucratic processes, privacy concerns on the part of the government, or the absence of any designated person or system to provide data.

Unfortunately, all data and analyses did not result in decision making. Buy-in by the policy actors was slower for many cases and often did not take place. There was an apparent lack of understanding on the part of the government as regards the data initiatives. The policymakers were not ready to accommodate the data initiatives undertaken by subordinate agencies at the beginning of the pandemic. In fact, the government started to show interest in accommodating data-driven findings from academicians and researchers in order to plan future interventions. While new data initiatives obviously contributed to policy actions, not all demands for data were met. The concerns regarding the accuracy, representativeness and quality of data put the usability of such data at serious risk. A number of COVID-19 targeted data initiatives have also been undertaken by NGOs or civil society organisations (CSOs) in Bangladesh. Regrettably, their acceptance at the policy level remained very limited.

CHALLENGES AND RECOMMENDATIONS: Evidence suggests that the improvements in the data ecosystem during the pandemic were largely concentrated in the area of health. The use of such initiatives was limited in providing public support. Even within the health issues, for the rehabilitation domain, no such initiative was found. Pre-conditions were important for the success of the initiatives. While using the NID database was possible as it covered a larger section of the population, the use of the tracing app failed due to various technical issues. Overall mindset and flexibility of the stakeholders were also found to be critical as all data and analyses did not result in decision making. Buy-in by the policy actors was slower than it should be for many cases and often did not take place. There is also ample scope for improving transparency in these initiatives. This has perhaps also constrained improving political buy-in. The scope for scaling up and involving more stakeholders could be done with improved transparency.

The future course of actions in view of the present achievements and challenges will critically hinge on three issues: the system-wide adaptation of the successful initiatives, enhanced scalability across domains, and ensuring HR-related/financial/organisational sustainability. To ensure system-wide adaptation, there is a need to establish a clear institutional architecture of the initiatives ensuring the issues related to governance and rules of business. The creation of a 'knowledge hub' containing all relevant data, statistics, research and analysis contributed by both government and non-government entities might be beneficial. Issues pertaining to data standardisation, reconciliation, disaggregation, interoperability, access, and quality assurance should receive due attention during its formation. Data privacy and confidentiality must be ensured in such efforts. Formulation of data privacy policies and the development of data sharing frameworks should also receive top priority. Scaling up of the data initiatives should focus on involving more sectors/issues as well as stakeholders both within and outside the government. Formation of a 'data community', following a whole of society approach, might be useful in this regard.

The sustainability of the data-related initiatives will require the utilisation of the learnings during the pandemic to bolster and modify the existing data initiatives. The use of administrative data and user-generated data must be expedited by the Bangladesh Bureau of Statistics (BBS). The sustainability of the data-related initiatives will also require mobilisation of additional resources- financial, human resources and technical - from both domestic and foreign sources. This is particularly pertinent for non-state actors who might lack the resources necessary to develop a robust data architecture.

Overall, political buy-in is critical for all three of the aforementioned areas. As has been evidenced, the data-driven approach towards policymaking has shown some improvement during the pandemic. However, the data generators and the knowledge community must continue their endeavour to sensitise the policymakers regarding the worth of data and its use. In addition, documenting the experience, generating evidence of the positive impacts of such initiatives, and opinion building by engaging stakeholders will also be helpful.

Dr Debapriya Bhattacharya, Distinguished Fellow, CPD; Towfiqul Islam Khan, Senior Research Fellow, CPD; Muntaseer Kamal, Senior Research Associate, CPD; Najeeba Mohammed Altaf, Research Associate, CPD; Marfia Alam, Programme Associate, CPD