CEA Arvind Subramanian on Demonetisation, UBI and Aadhaar

In a recent discussion in Washington, the chief economic adviser talked on Aadhaar failure rates, pushing Universal Basic Income forward and how demonetisation's popular response humbled him.

New Delhi: The full impact of demonetisation will only be known over the course of the next few months, according to Chief Economic Adviser Arvind Subramanian.

In a talk at the Centre for Global Development, Subramanian also pointed out that the note ban’s impact on the informal sector “will be difficult to assess”, in an allusion to how recent GDP data may not be able to how India’s informal economy performed in the months after demonetisation.

In his talk, Subramanian expands on an issue close to his heart: Universal Basic Income. According to the chief economic adviser, a number of states are currently in discussion with the Centre over how best UBI can be implemented.

In the discussion, he admits that Aadhaar authentication failure rates may have to be looked at again and that there are legitimate privacy concerns arising from the UID intiative. The Wire has collated edited excerpts of the discussion below. The full talk can be accessed here.

On demonetisation’s impact and the popular response

So I think yeah, demonetisation had a number of interesting things. We wrote about in the Survey. Firstly, there has been an impact on the informal sector which will be difficult to assess. We just aren’t going to be able to get a handle.

But I think it’s pretty much over because it was related to cash not being in the economy. Now cash is back, so those short-term costs are behind us.

That being said, I do think the headline numbers may not give a full sense of the actual impact of demonetisation. I think we will get to know over the course of the next few months what the actual impact is going to be.

Cause remember, there is going to be an impact on the downside but also an impact on the upside. Like, are we going to get more digitalisation as a consequence? Because people have been moving into credit…and digital payments are taking off. Are we going to see a situation where we are going to be more tax compliant? In some ways, demonetisation on the part of the government does signal a kind of regime shift. Which basically says, “guys if you are not going to be compliant on your tax payments, the government is going to take extra effort to make sure you do so”.

So I think a lot of these impacts of demonetisation, it’s a case of too early to tell. Except in the case of impact on informal sector which I think we don’t really are going to get a handle on.

On tax compliance, it’s something we are going to find out what the impact has been. A lot of this whole demonetisation… its part economics, part politics. I think that there are quite different lenses and perspectives with which to view demonetisation.

Could it have been done better? I think that’s something I am going to leave for the historians and not for me to discuss in any great detail.

But it does raise some really interesting questions about… The thought that crosses my mind is that is there an analogy between what’ve you seen on demonetisation, in terms of the popular response, and the kind of thing we see here [in America]. Why is it that people vote for a party that is going to deprive it of medical assistance? Kind of the ‘What’s Wrong With Kansas?’ kind of thing. I think there is a kind of counterpart here [in India, with demonetisation] as well.

I think it’s ‘What’s Wrong with Kansas’ blown up. Because… if you think there was a cost, why is it that the popular reaction to this [demonetisation] has been so overwhelming? It’s certainly humbled me in terms of my understanding of Indian politics and even Indian economics to a certain extent.

On whether Indian states can start UBI on their own..

At the moment, it’s very patchy in terms of how this has [JAM infrastructure] developed. Some of the states, where many poor people live, those are the states where last mile infrastructure is not as advanced as it should be.

But there are a lot of places where it is getting there. You can conceive of states where you can start that. [It] may be not be perfect…but [there are] places where you can reach a large percentage of the poor.

In principle, nothing prevents a state government from doing it [UBI] on its own. They can just do it. So they can do it on their own. But I think what some of the states that are thinking about this seriously do want federal money for this. One way this could work, under our constitution, the Centre transfers money to different state governments but often in the form of tied money, saying that we give you ‘X’ amount of money, but this is tied to certain schemes.

One possibility is for states to tell the Centre “don’t give us more money” but “give us this money as untied money and let us use it for say a UBI”.

That is the kind of conversation that is beginning to happen. That would make it easier for state governments to finance this. There are some states which could just say “If we can get some money from the Centre, we can go ahead and do this because this infrastructure is largely in place”.

But I should’ve said this earlier. Remember, one of the starting points for this conversation in India is the fact that existing social welfare schemes (food subsidy, employment, fertiliser subsidy) are very leaky. They don’t do a great job of reaching the poorest. In some senses, if you say well with UBI there are last mile problems… it has to be compared against the existing scheme where it is even less perfect in terms of reaching the poor.

If some states could make the case, as they are, that the existing schemes… in fact we have these calculations in the Economic Survey which shows how weak the targeting can be for these programmes [India’s existing welfare schemes]. UBI can easily improve upon that. So at least, in a relative sense, it can be shown as a substantial improvement.

On why Aadhaar can’t be used to improve targeting

The attraction of UBI in some conceptual sense is that…targeting has proved to be highly inefficient and ineffective.

If you want to say but ‘no, we have new technology so let’s use that for better targeting’, I think it’s going to kind of run into the same issues that you get with conventional targeting as well.  So maybe there are solutions where you can improve targeting, but I can give you one example why it happens that targeting will always run into problems.

Take our employment guarantee scheme. Anything that has to be accessed by the state governments for example. You have a programme and they [the states] have to implement it in some way or the other. It turns out that the very states that are better at doing that, by definition, are going to be the states that have the least amount of poor! If you have better governance, by definition, you have less poor in that state.

That’s what we find in the employment guarantee scheme [NREGA] for example. And that’s true for schemes across the board. That the states that are better off, that are better able to take advantage, are the ones who need it the least!

If you take a state like Bihar, which is probably the second poorest state in India, the employment guarantee scheme virtually doesn’t function there. Whereas in a state like Andhra Pradesh, where there are fewer poor [people] or Tamil Nadu, is where most of the money is actually taken up even under the employment guarantee scheme.

That’s always going to be there. That’s why universality… the advantage of a Universal Basic Income is that someone from ‘up there’ just sends a cheque to a bank account and you kind of bypass all the intervening stages of the bureaucracy which is where all the leakages and corruption that take place.

On confidence of overcoming last-mile infrastructure problems

So as you said, the only experiment that has taken place on UBI is in the state called Madhya Pradesh. It’s a very small thing.

What I think gives us confidence that actually a scheme that our friend here implemented here. It is the Indian cooking gas subsidy.

So think of it this way. You have cash transfers for specific subsidies. You have kerosene subsidy and so on. And then you have UBI. Basically, if you scale up cash transfers it kind of approaches an UBI.

We’ve had success with the cooking gas subsidy which has reached hundreds of millions of people. That’s the infrastructure that we want to build upon… which gives us confidence that we can actually overcome the last-mile problem. It’s not so much that the Madhya Pradesh experiment as the success with the other cash transfer programmes.

Mind you, not all of them are going great. I think the cooking gas is doing well, but the food and kerosene programmes are not doing as well. But we know we have a successful programme and we want to build upon that.

Aadhaar’s privacy concerns

It’s not about UBI. It’s about biometric identification that underpins that. I recognise that I’m in Republican America, having to take this question… So look, I don’t know a whole lot about this to be honest with you. But I think ideally what we would want to do is recognise that there are huge benefits from having this, especially in a poor country like in India to better deliver services.

That’s the starting point. That’ what underlay the bill that the Govt passed last year, to codify the Aadhaar… and there was a lot of legal uncertainty.

However, I recognize that that cannot override considerations of privacy and we need to balance the two. Where that balance should be and how that’s going to be achieved is going to be very much country-specific. And I suspect that this is going to be an on-going conversation in India. I don’t think we’ve seen the last of this convo. There are people who have legitimate concerns about privacy and I think it’s going to be an ongoing conversation.

The courts once again have to render their verdicts on this. So how it’s going to evolve is difficult for me to say.

There are always hacking issues. So the question is do we have sufficient safeguards against hacking. You can never get a perfect system. In India I think there have been some issues on hacking. But again I’m not a expert. I defer to Nandan on this and he said recently it’s a pretty secure system as systems go. Whether this prevents hacking, I don’t know. The concerns in India are less hacking and more privacy.

Fidelity and robustness of Aadhaar

The fidelity of Aadhaar. Maybe we should have had Nandan [Nilekani] here. In the last two weeks or so, I have also heard reports of things not being as… you know, the authentication rate is not as high as it should be. Because remember, this was touted as initially you would get a failure rate of even less than 0.5%!

I haven’t looked at the evidence carefully… they say in some parts the failure rate are much higher. It’s possible that we may have to reassess that. And… I think then it would be a matter of just getting the technology right and investing more in the technology. I don’t think at this stage that any of this suggests that we should abandon this scheme or anything like that.