Karthik Muralidharan (Tata Chancellor's Professor of Economics, University of California, San Diego) recommends targeting the 20% of poorest blocks in the country under NYAY, and making the cash transfers universal or quasi-universal in those areas. 

Can you explain why NYAY is necessary when we already have so many other poverty alleviation schemes? Why should we not just increase the budget for the existing schemes? 

I think the announcement of NYAY mainly serves a political purpose to remind voters that the UPA (United Progressive Alliance) was responsible for creating landmark anti-poverty programmes like the MNREGS (Mahatma Gandhi National Rural Employment Guarantee Scheme) and to reinforce a political commitment to the most under-privileged voters.  

Given the confusion regarding the details in the various statements made about the scheme, I do not think that the exercise has been accompanied by much attention to details of budgets, implementation, and most importantly – the opportunity costs of this specific programme. This is seen in the criticisms of the programme from both the left and the right. So, my sense is that the details are still very much up for debate if the UPA were to have an opportunity to implement a scheme like this. 

My personal view on new programmes is that any party that wishes to announce a major new national programme has the responsibility to first try it out in states that are directly administered by that party to demonstrate viability and support before trying it out nationally.  

The MNREGS was a good programme in part because the model had been validated in several state-level efforts over the years. So, if the UPA is serious about NYAY, I would recommend that they first try it out in states like Chhattisgarh, Madhya Pradesh, Punjab, and Rajasthan to fine-tune the details before attempting a national version. 

Do you think that the money would be better utilised if instead it is used to improve the existing public education and health services? Or, have we simply given up on trying to make public education and health delivery functional? 

I have mixed views on this question. On the one hand, the case for improved public services is very clear. On the other hand, the quality of expenditure on public services under the status quo is really poor.  

Take the example of schooling where the public spending per student in government schools is considerably higher than the total cost per-student in affordable private schools. Yet, millions of parents are choosing to pay out-of-pocket for these private schools (which are only slightly better than the government schools, but typically have better accountability – which is visible to parents).  

The situation in health is even more striking where over 70% of primary healthcare visits in rural India are to fee-charging private providers, even when the village has a public health clinic. These private providers are less qualified, but exert more effort and evidence suggests that the quality of care is not lower despite their being less qualified. 

So one striking way of thinking about the quality of public services is: 

“What does it say about the quality of your product, that you can’t even give it away for free?” 

In other words, revealed preferences suggests that there is a lot of ‘value destruction’ in public service delivery because the government incurs large costs in providing a service that is of such low quality that most people do not want it. This is mainly because government salaries are too high and accountability of frontline workers is low. 

I still strongly believe that we need to improve the quality of government services, but also do not think that this will happen by simply increasing spending. Evidence from multiple sectors (NREGS, education) suggests that improving governance can be ~10 times more cost effective at achieving the same level of effective increase in the presence of a programme on the ground than spending more on the programme itself. So, one way of forcing the public sector to improve its efficiency is by not providing it with a captive market of poor households, but to increase the options that the poorest have to avail of services from both public and private providers. 

My bottom-line here is that providing some income support should be seen as a complement to public services and not a substitute. Over time, once the State shows that it can credibly reach the poorest through income transfers, it opens up a set of policy options whereby the poor can choose between status quo public services and an equivalent cash transfer (an approach I have outlined for the PDS with Paul Niehaus and Sandip Sukhtankar).    

Thus, the role of an income transfer would be to empower beneficiaries to have a stronger voice in how social sector funds (spent in their name) are actually spent. In other words, the public sector would need to show that it can deliver more value than an equivalent income transfer, and compete for the business of the poor over time. 

Would you prefer an alternate scheme that distributed the same amount of money through cash transfers over a larger number of households by reducing the amount to each recipient? If yes, why? 

Yes. While there has been a fair bit of debate in India on rolling up existing subsidies and welfare programmes and replacing them with a Universal Basic Income (UBI), I am less in favour of this. 

Rather, my preferred way of making income transfers a part of the portfolio of solutions used to reduce extreme poverty is to implement what I call an Inclusive Growth Dividend or IGD (pegged at 1% of GDP (gross domestic product)) that is paid to all citizens as a supplement as opposed to a substitute to existing welfare programmes.  

The logic of such an approach is spelt out in detail a recent op-ed I wrote along with Paul Niehaus and Sandip Sukhtankar – but the main reasons to do so include built-in progressivity, low targeting cost, no disincentive effects to work and earn because there is no phasing out of the benefit, sociologically more acceptable because the programme does not change ranking of households by income on the ground (which NYAY would do), demonstrating that the government can credibly reach every citizen in a consistent and reliable manner, last-mile financial inclusion, female empowerment (because the allowance for children would go to mothers), and setting up income transfers as an attainable benchmark against which other government spending can be held accountable. 

If fiscal constraints are binding, I would strongly recommend targeting by location – by identifying the most disadvantaged districts/blocks and making the transfer universal within those jurisdictions.  

This is an approach that the 15^th Finance Commission should strongly consider because an important goal of the Finance Commission is equity across the country.  Typically, this translates into larger allocations to more disadvantaged states.  But these states also typically have weaker governance and so the translation of public expenditure into outcomes is much weaker in these settings. Thus, the value of having at least some portion of spending earmarked for ‘equity’ going directly to the beneficiary is especially high in the poorest states.  

Since an approach that targets the poorest districts/blocks will automatically over-weight the poorest states, such an approach can let the 15^th Finance Commission achieve its equity mandate while also creating all the positive effects discussed above. 

Do you have any concerns about the fiscal burden of the extra expenditure associated with NYAY? Do you think that the existing tax rates would have to be changed or the existing subsidies eliminated to accommodate the scheme? 

Fiscal space is always scarce and a central question is how much we allocate to public goods versus redistribution. To the extent that politicians and voters will continue to demand ways of alleviating current hardship of voters, having income transfers constitute one component of a portfolio of anti-poverty policies makes a lot of sense. 

What would you say to people who are worried that an infusion of cash would increase demand but not the supply and cause a price rise? 

I think this is a second-order concern because any price increases are likely to be seen only in the short-term because in the medium term, supply is more elastic.  Further, connectivity and market integration continue to grow in India reducing both search and access frictions and so I am not really worried about this issue.  

In fact, recent evidence from a large-scale income transfer programme in Kenya finds considerable positive effects on non-participants through multiplier effects of additional economic activity induced by the demand increase resulting from the income transfers. 

Targeting the bottom 20% seems to be a herculean task.  Which data can the governments use to identify the bottom 20%? What sort of targeting mechanism would you suggest? 

I am not a fan of this approach for the reasons mentioned above. I would much rather target the 20% of poorest blocks in the country (for perspective, per capita income in Bihar is around one-fifth that in Maharashtra) and make the transfers universal in these areas (or perhaps quasi-universal – by excluding those meeting clearly verifiable criteria of not being needy such as having a government job, or having an income above the income tax exemption limit). 

The proposed scheme is bound to generate perverse incentives for a significant part of the population. They would try to show that they are a lot poorer than they are or even to lower their incomes in order to qualify. How would you construct the scheme to minimise the damage from this obvious problem? 

I agree that the proposed version of NYAY is likely to create highly perverse incentives.  

Research on anti-poverty programmes around the world highlights that the “phasing out” of targeted benefits for the poor as they earn more, creates considerable disincentives for work (since the phase-out imposes a high marginal tax rate on earnings). 

This is a key reason for why I recommend the IGD approach.  

The deficiencies in the existing poverty alleviation schemes stems from weak State capacity.  Will NYAY also not be hampered by the same? 

Agreed. Again, a key reason for my recommending an IGD approach rather than the one in NYAY is that I think that an IGD is much more implementable at current levels of State capacity than NYAY as currently envisaged. 

This post is part of I4I’s e-symposium on NYAY: https://www.ideasforindia.in/topics/poverty-inequality/decoding-congress-nyay.html