Bangladesh's Data Dilemma: What lies beneath the advertised success?

In 2020, as the world faced the worst pandemic of the century, Bangladesh was thrust into immense challenges. The loss of lives and livelihoods, coupled with the disruption of normalcy, meant that policymakers had to respond urgently to the crisis. It also meant that the havoc wreaked left a lasting impact on the country's well-being and economy. Overall, there was a well-planned and data-driven response to the crisis, but this is precisely where the problems began.

The data presented by the Bangladesh Bureau of Statistics (BBS) did not align with that from numerous research organisations. Specifically, there were concerns about discrepancies in the reported poverty rates. The poverty rates provided by research organisations were significantly higher than the official figures. While this discrepancy could stem from various factors, what was particularly problematic was the government's response. There were no proper discussions with the research organisations, as would be the norm in scientific research. Instead, the government chose to disregard all the evidence put forward by the social scientists in these organisations. Furthermore, the methods the government used to generate the data were not made public.

Even today, questions remain about the official poverty target rate, inflation rate, and other key statistics. Any questions from the scientific community and civil society, however, have been labelled as rumours or strategic manoeuvres, even equated to acts of treason. Thus, this practice of disregarding any evidence that does not align with official data, without proper engagement, has resulted in the country being unaware of the true state of these key statistical indicators. This means that any policies currently being pursued will be completely misguided, as they are not based on the true circumstances. As the country begins its rebuilding, only data-driven policymaking can ensure it is not derailed.

There have been vested political interests in presenting a perfect picture through several key economic indicators. In particular, when it comes to GDP growth rate, inflation rate, poverty rate, and export data, there are still questions about their authenticity. The government's desire to show how successful they have been in promoting progress and growth has led to a vicious cycle of distorting data when presenting it to the public. Such incentive structures should be dismantled, and mechanisms need to be put in place to keep such actions in check. Only then can we be sure that the data presented by the government is authentic and not merely political theatre.

It is also evident that governments often shy away from accountability and transparency in data generation, management, and dissemination. The data concerning the social safety net is a prime example. There has been insufficient discussion about the indicators or criteria used for inclusion and exclusion in these programmes, or how effective they have been in improving the lives of the recipients. Given the importance of this data in guiding government actions, especially those involving monetary and other resources, it is vital that data accessibility is ensured and that it is disseminated in a way that encourages discourse and feedback.

Well-integrated and established datasets are crucial for policymakers and researchers to understand the internal dynamics of socioeconomic issues and to devise solutions for the problems that may arise. Unfortunately, in Bangladesh, there is a persistent gap in acquiring and disseminating proper datasets across several sectors. This creates a significant gap in the country's understanding of its own issues.

In Bangladesh, data is managed by the Finance Division of the Ministry of Finance, the Bangladesh Bureau of Statistics (BBS), the Bangladesh Bureau of Educational Information and Statistics, and other organisations. These government-owned bodies conduct various surveys to produce datasets, such as census data, education data, labour force surveys, and income surveys. Despite this, major challenges remain in Bangladesh's data management.

Upon reviewing the available data on key socioeconomic aspects, we conclude that its quality is insufficient. In most cases, the data is not up-to-date, and often it dates back several years, which may not reflect the current situation. This is particularly crucial given the crises the country is facing at present. The Ukraine-Russia war, the global rise in prices of essential commodities, and distortions in local markets have impacted the realities on the ground. This means that the values of specific indicators might have changed significantly. Additionally, annual or even quarterly data is essential to identify trends and patterns that require attention at policymaking levels. The regularity of data thus impacts the evaluation and decision-making process needed to achieve holistic development.

Despite the targeted priorities for development, it is important to recognise that progress towards any goal is not uniform across socioeconomic strata. The lack of disaggregated data, such as by gender and age, hampers the identification of specific gaps in promoting the current circumstances to the targeted levels. Problems persist on several layers, and the discrepancies in performance levels provide insight into these multifaceted issues.

Volume, accuracy, and completeness are the key components of data quality. The size and accuracy of data are critical to ensuring that it is truly representative. Intrinsic and contextual features—objective and dependent on usage, respectively—are vital to maintaining good quality. Reputation, accessibility, and relevance are key elements of contextuality. Therefore, the data frameworks in place to evaluate indicators mustn't be devoid of local contextualisation, particularly when addressing equity concerns.

Regarding collection methods, there needs to be better transparency and updates. It is important to recognise that globally, several indicators are scrutinised daily, so it is crucial to incorporate such updates into national accounting, calculation, and collection methods. Open discussions regarding the automation of updates or calculations are necessary. For example, when discussing the gender budget, after a certain point, it becomes a black box, and subjective opinions are used to conclude the final data. It is essential to be transparent about the considerations incorporated into such criteria and what truly constitutes the gender budget.

Capacity building and political will are two critical issues in data management. Often, there is a significant lack of coordination and cohesion among the agencies responsible for data management. Many individuals in charge of data are not competent or skilled enough to handle the volume and precision required. 

Moreover, there is insufficient political will to train relevant personnel and build the infrastructure and human resources necessary to ensure clear and accurate data generation. The subordination of key governmental data agencies to political pressures undermines their professionalism, which can be harmful when building a system of data accountability.

Regarding data availability in Bangladesh, it is crucial to examine the legal and financial dimensions of accessibility. There must be clear, distinct, and transparent guidelines regarding who can request data, how much it costs, and how long it takes to access it. These conditions should favour a freer environment where data accessibility leads to better analysis of specific issues.

Bangladesh adopted the Sustainable Development Goals (SDGs) in 2015, alongside other United Nations member states. From the goal of zero poverty to ensuring decent employment and climate action, the SDGs form the blueprint for sustainable development. To make the well-being of the nation resilient to persistent economic shocks, Bangladesh must achieve the SDGs. However, from 2000 to 2017, the data availability for the SDGs was only 39% of the total disaggregated data series in Bangladesh.

The General Economics Division (GED) of the Bangladesh Planning Commission reported in 2018 that, out of 232 SDG indicators, data is readily available for 64 indicators (27.6%), partially available for 58 indicators (25%) and unavailable for 110 indicators (47.4%). The gaps are often in the environmental dimension. 

Given that Bangladesh is one of the most vulnerable countries to climate change, this is particularly alarming. A comprehensive and disaggregated database is crucial to tackle the monumental challenge of climate change.

It is already 2024, and there is no point in denying the current realities if the country aims to achieve the SDGs by 2030. Bangladesh must acknowledge the true state of its SDG indicators and determine the next steps. There must be discussions on whether the country should modify the SDG targets, identify gaps in the strategies, or prioritise specific indicators. 

Any move forward without clear discussions with the relevant stakeholders will only worsen the situation by removing transparency and accountability from the process of progressing towards the SDGs.

With the advent of new technologies, it is essential to integrate new methods into data analysis and presentation. Traditional forms of analysis will fail to provide a data system that meets international standards. Incorporating big data and using sophisticated tools can help make large and complex datasets more understandable.

On the other hand, efficiency and cost are the most important measures of data management systems. Adequate allocation and innovative implementation of resources can mobilise better-coordinated and more advanced data systems.

Lubaba Mahjabin Prima is a research assistant at South Asian Network on Economic Modeling (SANEM), Dhaka, Bangladesh.

 
Eshrat Sharmin is a senior research associate, at South Asian Network on Economic Modeling (SANEM).