Reform in Practice: Vishal Tripathi (CEEW) on Administrative Data and Information

What is one policy area in India that needs urgent reform?

The problem of information asymmetry is a problem that affects like all sectors in India in general is my view. There are big problems with administrative data in India. We do not capture the data that we should be capturing.

Whatever we do capture, we capture it in a manner that it's not easily accessible and used by actors, policy makers, researchers. And thirdly, sometimes we capture data with no view of knowing how that data is going to be used and maybe that data doesn't need to be captured in the first place. So why this remains a challenge across sectors is because on paper we have policies which mandate like there's a national data sharing and accessibility policy of Government of India since 2012.

So all administrative data should be shared via API access and it should be available. But do we have the people to get this done? Do we have enough funds being allocated to ministries to get this task done? I think that's where bulk of the problem is. So maybe some examples on how this creates problems in different sectors could be useful in understanding the depth of the problem.

So for instance in the power sector, I work in CNI decarbonization. So whenever commercial and industrial consumers get demand notes from the distribution companies, they often are not backed by enough data on how that bill was calculated and this creates a lot of disgruntled consumers. The regulatory commissions at the state level, they churn out tariff orders which are usually 700 to 800 page documents.

And for policy researchers like me, it's a big pain to go through those voluminous documents and extract the data points that I really need. So it just becomes a humongous task for any kind of useful research through them. So one of the better platforms in the power sector, a data platform was the web-based scheduling system run by Grid India.

And it used to be a goldmine for researchers like me. But last year the government actually put it behind controlled access because apparently people were running scripts on extracting data and that was just bringing down the platform. So we sometimes take one step forward and two steps backwards also and so whatever little data is available is now no longer available.

So these are some of the problems that I see. I have given example of power but I've worked in health before. I see similar issues there.

So in my experience, in my fieldwork and research that I've done, I've seen high school students and university students being hired as contractual staff to do the data entry work at the end of the month. So if this is how we are going to capture data, the quality of the data is going to be a challenge. And also there is like no knowledge management within these ministries.

And because you don't have the same person doing the same work over a period of time, you're not building expertise with this kind of work and all of that is being contractually given out. So these are some of the problems I think I see. 

What would a more effective alternative look like?

When we say data and data architecture, ultimately it's going to be a human that is going to make that happen.

And we all know how poor the state capacity is especially in terms of like understaffed government departments everywhere. And that is also a cause of poor data quality in India because too few people are doing too much work. And like in health, there are wonderful examples of new apps having been made and given out to frontline workers like ASHA workers and Anganwadi workers.

But they run on very poorly designed softwares, the phone and the hardware is not up to date. And this data entry burden actually takes up most of their work. It takes them away from their real work, which is delivering public health services.

And I've seen similar challenges in the power sector as well, because there is one human sitting behind a machine at every substation in India, taking notes on a register, and then later doing the data entry work to put it on the reporting platform. So I think the solution could be one, of course, lay data pipelines in a manner that is already well thought out. First decide why are you collecting this data? If the answer is yes, that yes, this is going to be useful for purpose, only then collect that data, otherwise don't.

So how you lay these data pipelines and create the data architecture should be a very well thought out process with a lot of public involvement and stakeholder consultations like we do for laws. I think we need to do something similar for while laying data pipelines as well. And secondly, there should be some sort of workforce adequacy assessment also.

Whenever you're giving extra work, data entry work to a government department, first ask the question, are there enough people there to do this? And if not, then how might we fill that gap? So in the power sector, recently, there was a workforce adequacy assessment done for state load dispatch centers, which was a good move. But I don't know if the findings of the report have been implemented yet. But we need to do this for all different government agencies, the distribution companies should also have workforce adequacy assessments.

So build a data architecture that is well thought out and through a consultative process, and then make sure you have enough people to do that as well. So I think we need to do this.