Why is data transparency crucial for economic decision-making, risk management and policy formulation, as well as the government’s accountability to its citizens?

Data is vital for determining how the economic actors will function and ensuring the government’s accountability to the people. Let me draw an example. One of Professor Mainul Islam’s research projects in the year 2000 brought an important issue to light. While Bangladesh was considered self-sufficient in food, a lot of food was being imported simultaneously. The question arose as to whether there was an error in per capita food consumption data, total food production data, or data regarding food imports. 

The country’s food security was under scrutiny as it is crucial for its economic and social stability. If we measure the amount of food needed in a year based on incorrect information, whether we have that amount of food grains in stock or the production process, then it will create a major crisis. So, accurate data is essential because errors are exposed in the end, leading to a crisis.

Another important aspect is the government’s accountability to the people as their representatives. People can only hold the government accountable when they know what steps it is taking and how those will impact various economic actors and the country’s citizens. 

What role can data play in solving national issues like inflation? For instance, in Ethiopia, farmers are provided with real-time price data, which allows them to improve their decision-making in product sales. Some countries use GPS tracking devices and mobile data to track late-night logistics routes, which ensures an efficient supply chain. Can we follow suit in Bangladesh to fight extortion and optimize transportation costs?

Inefficiency is one of the manifold problems in our supply chain, which can be solved using technology. However, issues will persist if the stakeholders are corrupted, be it the producers, middlemen, transportation authorities, law enforcement agencies, wholesalers, or retailers. Our problem is that we are excessively profit-driven. Only a digital solution will not solve the problem of deprivation of producers and extortion of consumers. 

Many in the supply chain even consider rent-seeking a right practice. Hence, I do not think it is possible to increase supply chain efficiency or check the factors leading to price hikes, i.e., greed, middlemen’s share, and rent-seeking, through technology only. Technology will bear fruit when the supply chain actors reach a consensus among themselves regarding the transparency of production costs and pricing. Moreover, the government’s oversight system for breaking monopolies or oligopolies works properly. 

So, if we can increase data flow through technology, can it be effective in curbing issues like inflation? 

Theoretically, it can. For instance, today, everyone, including farmers, middlemen, retailers, or wholesalers, can track the prices of products in different markets through their mobile phones. There is no integrated system for this, but pricing data is shared through an extensive organic network. However, more transparency can be ensured if we arrange digital boards in every market containing today’s product prices in other markets. The Department of Agricultural Marketing can do this. 

However, there are deeply rooted problems in the system. Take the case of farmers selling products at a much lower price than their retail market price. Even after knowing their product will be priced at BDT 80 in Dhaka, they settle for BDT 10. Because they do not have a choice but to sell their product at a price fixed by the mohajons (local wholesale buyers), as they sell their products at their doorsteps. The monopsony behavior in the market (farmers to middlemen) does not allow farmers to bargain prices. So, the effectiveness of transparency is in question here. As far as I know, none of the projects conducted worldwide to control market prices with data transparency have been successful. 

When it comes to selling products to consumers, our markets have an oligopolistic structure. A few cartels fix the final prices of products in Karwan Bazar, Khatunganj, or any big market. Technology cannot be useful in such a market structure unless we can introduce an alternative system where the producers have end-to-end control.

Experts have long been advocating for data-driven policymaking. For that, strong, credible data is required. While discussions about strengthening the Bangladesh Bureau of Statistics (BBS) have been in the air for a long time, some are proposing autonomous statistics offices (like that in South Africa) and data commissions. How do you see the possibility?

This is very important. Recently, initiatives have been undertaken to establish a separate and independent data management commission, aiming to free Bangladesh’s whole statistical system, which is now under a ministry’s supervision. If it were directly accountable to the parliament, it would be even better. 

Another aspect is BBS’s competence. BBS’s data management standards are not up to global standards. Let me mention a case in point. Two specific criteria are supposed to be used to maintain the data of Bangladeshi workers working abroad or in Bangladesh: Industrial Classification and Occupations Classification. 

Occupations classification is ISCO-8 (International Standard Classification of Occupations), a globally recognized system introduced by the International Labor Organization (ILO). Neither BBS nor BMET (Bureau of Manpower, Employment and Training) stores this data (following ISCO-8) in Bangladesh. Hence, we are unable to compare our data with others. 

Therefore, an independent commission for data is indeed essential and possible. We must also strive to follow the international data collection and management standards. 

There’s an IMF-prescribed voluntary Central Bank Transparency Code. Adopting this, the Bank of Canada releases summaries of policy deliberations, the Central Bank of Chile created a special section on its website to provide additional information on its operations, and the central banks of Morocco, Seychelles, and Uganda use review findings to boost the effectiveness of their communications by developing institutional strategies and strengthening communication units. Do you believe such a voluntary transparency code can be effective for the Bangladesh Bank? 

We need to examine the IMF (International Monetary Fund) standard and BIS (Bank for International Settlements) and then decide about developing our own framework, which is compliant and fulfills our needs. Following BIS standards is voluntary, but following them instills confidence among the global financial actors. In general, there is no harm in following IMF standards; however, the extent and form of disclosure vary. 

It needs to be mentioned that BB publishes more data on its website than any other government agency in Bangladesh. Monetary policy and other relevant policies are regularly updated there. 

The issue is that the economic and policy analysis departments are now relatively weak. These departments have limited ability to make real-time decisions by analyzing financial data. Hence, a strong department for research and analysis and timely decision-making can be created at BB. 

Also, the data BB releases on its website are not in a searchable format, i.e., Excel sheets. Thus, researchers cannot import and use the data smoothly. BB usually publishes its data in PDF format, which needs to change. 

To me, more important is ensuring Bangladesh Bank’s autonomy first and then deciding to adopt the voluntary code. Bangladesh Bank had greater autonomy in the 1990s. In 2009, the government introduced the Financial Institution Division, which coerced BB’s autonomy and enabled ex-bureaucrats to join as governors. This was the beginning of the downfall. Data standards would help in finding anomalies in a more structured manner, but manipulation happens when decisions are made politically. 

In the past, data protection laws in our country have been politically influenced and used for political gain. On the contrary, we also failed to stop data breaches, as we saw NID data leaked. Now, in Bangladesh, for an ideal data protection framework or law, what elements would you expect to ensure data transparency while ensuring that data privacy is not breached?

This is an important issue. First, a country’s data protection and cybersecurity legal frameworks depend on the political and philosophical standing of the government. Ideally, data protection legal frameworks should be citizen-centric. In legal terms, we are talking about two entities: natural persons or individuals and legal persons or organizations. So, a data protection legal framework should also work for individuals’ data held by the privately owned companies they work for. 

However, the data protection law drafts we have seen so far are not concerned with protecting citizens’ data. They are mainly designed to store citizens’ data digitally and let intelligence agencies and other forces access it without accountability. This is a lopsided approach that is not designed to focus on citizens’ welfare but rather on citizen surveillance and the interests of government agencies. 

To discuss data breaches, let us start with issues regarding data interoperability. Different government organizations hold citizens’ data. An information exchange system must be in place for citizen service purposes, which we do not have. Each organization keeps data without any interconnectivity among them. Fortunately, the interim government has made data interoperability a priority. 

On the other hand, we saw incidents like selling citizens’ information to various institutions for money, which is a big threat to our data protection system. Information exchange is a regular practice, but if money is involved, it becomes questionable. 

The strongest data protection framework in the world is the General Data Protection Regulation (GDPR) in Europe. If we can follow this framework to formulate our data protection law, it will be better for citizens’ privacy protection. Once the standard is set, we can reach the highest level gradually. We need to be vocal to ensure that data protection law, in the name of citizens’ information protection, is not made for surveillance purposes. 

Bangladesh’s data ecosystem is complex. Companies collect data from various sources, making tracking and disclosing data usage difficult. In recent years, some large groups have siphoned people’s money abroad illegally by manipulating data. How can we hold big companies and business entities accountable to the people in such an ecosystem? 

First, we need to discuss data manipulations in terms of money laundering. Trade channels are one way to siphon money out of the country by manipulating export and import data. Launderers under-invoice the amounts while exporting to keep the extra amount in their foreign bank accounts. Say you sold a hundred units of products for USD 1000, but brought back home USD 500 and kept the rest in your foreign account. However, customs can identify such fraud by tracking market prices if they feel suspicious about their on-paper prices. 

Product prices are over-invoiced in the case of imports, for instance, sending USD 2000 for a USD 1000 product. This is also traceable by having a database to compare product prices with the international market. An integrated information system can solve this invoicing issue—both over- and under-invoicing. 

Again, there are many other ways of money laundering apart from trade, like remittance, service-related purchases, etc. The solution, however, is not complex. Through an efficient system, the government can track every product that crosses the border or comes in and identify instances of such fraud. Whatever amount of export-import the businesses declare must be monitored to ensure it actually happened. There is a trade facilitation agreement of the World Trade Organization (WTO) and the International Trade Center (ITC) for this monitoring purpose. We can follow this framework to bring transparency to our trade. 

In 2003, the Partnership in Statistics for Development in the 21st Century (PARIS21) formed a task team to examine ways to improve the statistical systems of developing countries. The task team found two key issues: underfunding and weak administrative data. Do similar issues prevail in developing nations today after two decades? 

Investment in data generation and analysis is inadequate. This segment needs to see a big rise in investments, some of which can be utilized to develop data cadres. They will work with data, develop talents at the university level, and work to develop the capacity of organizations like BBS in data management and transparency. We need this because our data generation and management system is outdated and cannot assist us in our journey to a middle-income country. 

Apart from investing in this sector, we must also follow internationally recognized data frameworks. Most importantly, we must stop political interventions in the data management system. Establishing a powerful data commission accountable to the parliament can bring transparency. At least, it can theoretically lead us to the path to progress. 

There is a 3-tier IMF data standard framework. Member countries voluntarily participate in the initiative and publish timely economic and financial data. In our country, however, data is not usually published promptly. It’s March 2025, but we’ll still see 2022 or 2023 data on government websites. Clearly, voluntary initiatives are not enough for Bangladesh. What’s your take on the matter?

Following standards is indeed necessary for timely data. We may use an IMF-prescribed framework complemented by BIS-prescribed standards. 

Now, the issue is that we do not collect the transactional data properly. Data collection is not automatic in all cases. Data of bank transactions, in particular, is automated. However, different types of data are required to measure GDP, such as consumption data, production data, export-import data, inflation data, etc. This data collection is not yet automated because we do not have an integrated digital system at the transaction level. With an automated transaction data collection system, a real-time summary of the overall economy and interpretation is possible. 

Take labor statistics, for example. We do not have an updated system for tracking the latest labor data, such as the number of people unemployed right now and the number who lost their jobs last month. However, a national labor registry for real-time labor updates is already in use in many countries. 

In this regard, Bangladesh’s priority should be a system that records all transactional data for all public and private sector transactions. A central watchdog entity like Bangladesh Bank can have access to this data through interoperability. To materialize such systems, we need significant upgradation in terms of skills and globally recognized standards. 

Bangladesh is far behind in data credibility compared to some of the most trusted countries globally. Users do not feel confident using Bangladeshi data like they do when using data from the USA or European countries. They tend to doubt whether the data sampling was correct or whether the survey followed standard procedures. What can we do to establish this credibility? 

I think ensuring transparency is a political decision (in our country). We have to strive not to suppress or manipulate any data. We need strong political commitment and promises to the people of Bangladesh about not manipulating data and intervening in the data system. Civil society and business communities must work as pressure groups so that the government is obliged to provide correct data. We also need to repeal the ‘Official Secrets Act 1923’ because this act enables hiding necessary data in the name of secrecy. Also, the Right to Information Act 2009 needs to be revised to increase transparency and accountability regarding data. 

Lastly, how do you think the perception of data transparency and credibility will evolve as AI advances rapidly?

AI can change many things radically, and the speed of data collection is one of them. A survey that took 2 years to complete in the past can now be done in a mere 3 months using AI. It can also do a tremendous job of creating attractive and easily intelligible infographics and visuals to present data. The best part is that if we can aggregate data from various organizations through data interoperability, AI can sort the data and quickly create an overall economic report for policymakers to enable them to make data-driven decisions. So, how AI will be used should be a part of our policy discussion.