Next set of important information: I was born in Barisal, grew up in Dhaka, went to Bengali-medium school and college in Dhaka, did my undergraduation from the University of Dhaka. Before coming to Canada, I rarely knew economies (or anything) outside Bangladesh. So, it was highly likely that all my intellectual concepts and potential research questions are related to Bangladesh. In my masters and early years of Ph.D, I have tried hard to write these papers using my curiosities related to the Bangladesh economy.
However, I could not find a reliable micro or regional dataset to study any of these questions. There are some datasets collected by the Bangladesh Bureau of Statistics (BBS), and they rarely sell those. They rarely reply to emails too! I have applied to buy Household Income and Expenditure Survey (HIES) and the BBS agreed only to sell the data for 2016. BBS has collected HIES data in multiple other years, and they did not sell those to me. I have seen some papers using this database, so I do not know if there is a personal way to get it!
Even getting a sub-district level regional database is almost impossible on a regular basis. The only way is probably to manually extract information from paper reports for some census years. I went to the BBS office, and I got a database list that covers over 30 years. But, as you can already understand, none of these datasets were available for me to buy. Later on, for the sake of the timeline, I had to use contexts from other countries for my thesis. I had to learn historical contexts from other countries and their policies. I applied the contexts on Nepal and the USA.
Recently, out of curiosity, I started to read dissertations and current projects written by some of my contemporary people: people who finished undergraduation in economics from different universities in Bangladesh from 2000 to 2010. What I understand is that these dissertations are rarely done with Bangladeshi data and context. Some people who were/are affiliated with Bangladeshi institutions (i.e. BRAC) have papers using their data. These are also mostly randomized control experiments (RCTs). RCTs mean the idea is somewhere around an intervention, and not about any long-term economic development. Otherwise, for individuals like me, it is impossible to write a paper on Bangladesh given the current access to reliable datasets.
The reasons behind this are straightforward. Our government does not have a publicly available dataset that our economists and statisticians can use. So even with very strong applied economics training, we could not utilize our training to work on Bangladesh. Unless some of us are funded by development organizations and have to do experimental research for the dissertation. To note, these funded researches also almost always mean these dissertations have a given research question, development NGO-driven philosophy, and targeted audience. These researches are also not capable of looking at local economic growth.
In this reality, our graduate students and early-career researchers are mostly doing Ph.D. on contexts from other countries. This is why we do not have applied data work on structural transformation, local economic growth, urban formation, rural development - you name it! We have almost no good quantitative work on environmental degradation and economic loss for the same reasons.
In India and China, we see great research papers coming out from young professionals, publishing in top journals, and influencing policy debate. Even in Nepal, which normally we take as a smaller economy behind Bangladesh, we can get easy access to good micro and regional databases. My first dissertation chapter was on Nepal. I applied to buy microdata from Nepal Central Bureau of Statistics, and I did not face any issue to collect this high-quality data. I got access to the database in three weeks after application.
Where are we in that line of social science research? If we cannot support our economists and social scientists with social and economic data, what would be the return of having so many PhDs in the long run? With the current economic training, it is possible to do a cost-benefit analysis of public policies, war, historical events, and possible to figure out the total cost, and possible to use that for future policy decisions. But, we feel paralyzed given the current data infrastructure.
We are already late in building high quality data infrastructure to support policy research. It is already not possible to do quantitative research on past economic milestones. This limitation or nonexistence of data infrastructure has been highly related to our Covid-related losses and mismanagement too. I am not asking for data infrastructure just because some of us need to publish papers (well, we do!). I am asking because data infrastructure will give us information on how to target groups for subsidy, how to think about long-term regional economic growth, and how to design informative policies.