Recovery from the pandemic will depend on data-driven policy-making

The finance minister has recently rejected the survey findings of multiple organisations on the number of "new poor" created due to the Covid-19 pandemic. A number of think tanks including the non-government ones have undertaken representative household surveys on the employment and income loss due to the pandemic. Numbers vary across organisations. However, each of them has shown that the pandemic has eroded income of the poor and low-income families as they lost employment. This has pushed them into poverty. So, the number of the poor has increased compared to the pre-Covid period.

Surprisingly, the recently announced national budget for fiscal year (FY) 2022 has not mentioned anything about the new poor. Hence, there has not been any fiscal measures for them. The absence of the new poor in an important government document such as the national budget is equivalent to non-recognition of this hard-hit population. The minister is said to have mentioned that survey data by the non-government think tanks on the Covid-induced poor are not acceptable to him.

Unfortunately, despite having more financial and human resource capacity than the think tanks, no survey has been undertaken by the official data supplying organisation—Bangladesh Bureau of Statistics (BBS). Even after about 16 months since the outbreak of the pandemic in the country, BBS did not feel the urge to conduct a survey to have a thorough understanding of the impact of the pandemic on the economy and society. In the absence of such an initiative, several think tanks conducted both online and field-level surveys to estimate the socio-economic impact of the pandemic on various sections of the people.

These surveys are crucial not only for researchers but also for the government. Researchers analyse data and suggest policies. The government has to take decisions based on the data and analysis. The Centre for Policy Dialogue has been urging for timely, up-to-date and credible data for sound and effective decision making by the government for long. In view of the current debate on the poverty numbers suggested by the think tanks, a few data related problems are reiterated here.

First, there is a lack of adequate and timely data in Bangladesh.A number of important surveys are conducted by BBS after a certain time gap. BBS conducts important surveys such as the Household Income and Expenditure Survey (HIES) once in every five years. and the Labour Force Survey (LFS) once in every three years. Then the release of the survey results takes a while. So, researchers have to use this dated data for their analyses and policy recommendations. The government takes decisions based on these data for a couple of years. These surveys are conducted at large scale as these have to be representative. Therefore, these are expensive. But to address the data gap and make data relevant, BBS can conduct a smaller survey biennially to complement the larger surveys. This, of course, is not a full-proof method, but at least these can provide some idea on the trend of major indicators. In the absence of current surveys, when the Ministry of Finance (MoF) uses data to prepare the national budget, it cannot reflect the real situation.

Second, lack of disaggregated data hampers effective policy measures.  Aggregate data on many socio-economic indicators cannot provide much information. Therefore, disaggregated data at the sub-district and union levels are essential. Such information is useful not only for robust analysis but also for designing emergency response at times of crisis. The availability of such disaggregated data could have helped in better targeting of social safety measures for the affected people during the ongoing pandemic. For better targeting and selection of beneficiaries under social safety net programmes, BBS has been preparing a National Household Database. This is long overdue since this was to be completed by 2017. Needless to say, this could have been extremely useful for distribution of the cash support by the government during the ongoing pandemic. To help the pandemic-affected poor people, the government had announced to distribute Tk 2,500 among 50 lakh poor households in 2020. However, the support could be distributed among 35 lakh households only due to lack of proper listing of the poor beneficiaries. This has been an example of how data limitation can constrain government efforts to implement its decision.

Third, the quality of available data is poor due to lack of reliability, standardisation and consistency. We often find two different numbers in case of revenue data from the MoF and from the National Board of Revenue. In case of data on Annual Development Programme, information provided by the MoF are sometimes different from the ones supplied by the Implementation Monitoring and Evaluation Division of the Planning Commission. Also, at times, budget deficit and financing data vary between the MoF and Bangladesh Bank. Numbers are also found to be different at times for similar types of data from various sources. For example, information generated by a few surveys such as Sample Vital Registration System, Multiple Indicator Cluster Surveys, Bangladesh Demographic and Health Survey sometimes indicate different figures for the same type of information. The estimate on the growth of gross domestic product (GDP) is widely discussed since the numbers often do not correspond with supporting indicators, such as private investment and employment generation. The solution to such incoherence requires harmonisation of data collected from different surveys through better inter-agency coordination.

Fourth, data are also inaccessible at times. Since we have to rely on government sources for data, access to data is becoming difficult day by day. Sharing of data online is yet to be practiced widely in Bangladesh. Often it takes a long time to receive data from the organisations we request data from. At times, some organisations do not reply to our requests at all. Some would even decline on your face to provide data. Such non-cooperation could be due to fear among officials. There could be political pressure on them. Some officials would tell us that their data are confidential. It is difficult to understand what confidentiality would be there in publishing and sharing those data which are publicly available in other countries. The hesitation to face public scrutiny probably hinders the responsible organisations from publishing data. We recall, BBS published quarterly LFS data a couple of years ago. But this was discontinued following analysis of those data by experts.

Fifth, BBS must initiate data collection in order to set realistic goals during the pandemic. In the absence of official nationally representative data, several non-government think tanks have collected some critical socio-economic data. Some are still collecting such data. To ensure consistency, credibility and representativeness of data collected by non-government think tanks, BBS should extend cooperation to those institutions by sharing its survey methodologies.

Sixth, there is no alternative to a strong government institution such as BBS for data collection. The organisation needs to enhance its capacity by recruiting more statisticians. It should also reinvigorate its technical committees by engaging more experts. More financial resources are also needed to conduct surveys annually.

Additionally, BBS should be more independent and transparent in data generation and their dissemination to improve its credibility. The institution, as the prime source of official data, should have regular consultations with major users of data including the planning commission, think tanks, the private sector and non-government organisations.

Finally, the recovery from the pandemic will depend on the policy measures by the government. The effectiveness of these measures will depend on proper understanding of the state of the economy based on field-level data on key indicators including investment, production, employment, poverty and inequality. Authentic data on health indicators are also required. There has been an indifference towards the importance of quality data. It is time that the policymakers not only ensure the availability of reliable data but also use these data for effective decision making at this critical juncture of history.

Dr Fahmida Khatun is the Executive Director at the Centre for Policy Dialogue.

Views expressed in this article are personal.