Mumbai: In conflict-affected areas, rural social insurance programmes such as the Mahatma Gandhi National Rural Employment Guarantee Scheme (MGNREGS) can reduce violent events by half by shielding workers from income volatility caused by shock events, concludes a new study published by the University of Warwick, UK. MGNREGS now guarantees 150 days’ work per year in 14 drought-affected states, and a further 200 days in Odisha, as IndiaSpend reported in 2016.

The Warwick university study, based on events between 2005 and 2014, focused on the links between MGNREGS and conflict in 10 states that have a history of Maoist insurgency. This conflict resulted in 5,235 casualties between 2007 and 2013, including around 3,000 civilians, 1,000 security personnel and 900 Naxals, IndiaSpend reported in 2013.

“The districts where agricultural labour markets and output depend the most on monsoon rainfall benefited the most from MGNREGS providing insurance, seeing drops in conflict of up to 50%,” said Thiemo Fetzer, the author of the study, and associate professor in the department of economics at the University of Warwick.

Shock events such as droughts that affect income stability in rural areas are being increasingly pegged to climate change, and the impacts of climate change on rural areas are being increasingly viewed as global security risks. India is located in the world’s most disaster-prone region. Almost half its land area faced drought last year with rainfall reduced to a third of the annual average.

But in 2019, MGNREGS, the world’s largest rural workers programme, had its budget cut for the first time in five years--from Rs 61,084 crore in 2018-19 to Rs 60,000 crore in 2019-20. The scheme has been criticised for failing to provide the work promised, not paying wages on time, and not being viably implementable in India’s weaker states.

None of these shortcomings take away from the programme’s vital social value, Fetzer said.

Fetzer is an Associate Professor of Economics at the University of Warwick, a visiting fellow at the London School of Economics, and is also affiliated with the Pears on Institute at the University of Chicago and the Centre for Economic Policy Research in London.

The rise in demand for MGNREGS during poor monsoons coincided with a reduction in conflict levels in around 60% of villages located in Naxalism-prone areas, or the ‘red corridor’ as it is called, your study has found. Can you explain the link?

Yes, what the paper does is study the patterns between conflict in the red corridor and monsoon season rainfall. I observe that prior to the introduction of MGNREGS, droughts were associated with significantly more conflict and insurgent activities in the year after a bad harvest. This relationship seems to have become much weaker since the introduction of MGNREGS. This raises the question--is this due to MGNREGS or not?

To answer this question, I study in detail how MGNREGS participation indeed follows this seasonal pattern--with significant increases in participation in the programme following a bad monsoon. This observation--that MGNREGS participation is strong in areas that are experiencing periods of drought--is particularly pronounced in districts where agricultural labour markets and production rely heavily on the monsoon. It is those districts, where the value of insurance is the biggest, that see the biggest drops in conflict.

Despite the introduction of the MGNREGS programme, the total number of conflict events has increased over time across Maoist areas between 2005 and 2014. How significant was MGNREGS to the Maoist conflict overall?

From the aggregate trends, we do not really learn much about what is happening on the ground in places and how MGNREGS affected places. The aggregate trend could be driven by a whole lot of factors and changes, for example, improved reporting on the conflict--which is definitely a factor.

My study focuses on detailed micro data and suggests that had MGNREGS not been introduced, the overall increase in conflict would likely have been significantly higher. The parts of the country that benefited most from the scheme--specifically districts where agricultural labour markets and output depend strongly on monsoon rainfall--saw drops in conflict of up to 50%. This is sizeable and, of course, while aggregating back up from the micro evidence is difficult, would nevertheless suggest sizeable impacts at the aggregate level.

In 2017, it was estimated for 89 countries that natural disasters would put an additional 26 million people into extreme poverty (living on less than $1.90 a day) in the next year. Studies (such as this) have shown that climate-related natural disasters such as storms, floods and droughts have also affected the risk of civil war. Could programmes such as MGNREGS offer preventive and post-disaster reconstruction solutions?

This is a very important question. Countries around the world struggle with the climate emergency. It is a dramatic challenge especially for developing countries as they are set to bear most of the cost and have the least state capacity to mitigate the climate crisis. I think MGNREGS provides an interesting example of a potential policy that can work in reducing the experienced volatility of incomes among people in rural areas.

A basic form of social insurance can encourage productive investments, increase growth and development more broadly. It is a myth to say that a welfare state is undermining development--a basic social security net is absolutely crucial.

And this is where the problem is: How to design such an insurance programme in a context with weak state capacity? You want to ensure that insurance is well-targeted, i.e. only people in need actually benefit. You also want to make it easy for central governments and civil society to monitor programme implementation to reduce corruption.

MGNREGS does a fairly decent job at delivering on these two fronts. First, the fact that MGNREGS pays wages at the state minimum wage-level means that it is only attractive when wages in labour markets are low (for example, following a bad monsoon). Also, in order to get an income from the scheme people need to actually work--this ensures that people with good outside options or high incomes are not skimming the system as is for example the case with the PDS (public distribution system). In the PDS, it is well known that lots of households benefiting from the system are not needy by any reasonable definition.

And lastly, the requirement that MGNREGS produces physical assets and infrastructure makes monitoring and enforcement easier as civil society, the press and higher levels of government can always check on whether infrastructure actually gets built.

Of course, it would be great if one would not have to rely on these indirect mechanisms to ensure that social insurance becomes targeted--I am a bit doubtful about the quality and the social value of some of the infrastructure that ends up being built. But if a state is not able to provide directed targeted transfers to people affected by a shock that is none of their own making--such as a bad monsoon--MGNREGS does represent a valuable and important policy tool.

The programme, in that sense, serves as a significant inter-Indian transfer that helps stabilise rural economies following a shock and thereby, actually reduces the impact of the monsoon on the economic cycle.

At the end of April 2018, 57% of MGNREGS wages remained unpaid. In 2014-15, only 6% of households were able to hit 100 days of work, with the average being 40 days. Can it still play an effective role in providing income security and stemming conflict?

It is clear that there are many problems--after all the programme is just gigantic in scale. Regarding the two figures you mention, there is an evident erosion in the quality of MGNREGS in recent years and that really needs to be tackled. After all, its value is that it does provide for a stable wage floor and assured alternative for rural households in distress. This in and of itself has an effect on rural labour markets, helping stabilise wages following a bad shock. So the programme also indirectly has some benefits even for people not participating in it. But this indirect benefit is only maintained if those who do opt for MGNREGS actually end up getting paid.

Regarding the 100 days--it is true that there are cases of rationing. This can have many reasons. For example, panchayats may be unwilling to provide employment if only a small number of people request employment because there is an actual cost to panchayats to come up with public works projects. There are also likely conflicts of interest if village officials are also agricultural land owners who have an incentive in not providing MGNREGS work during the agricultural season. These all contribute to a problem whereby the scheme is not really effective in providing insurance against household idiosyncratic shocks that affect individual households, but not whole villages. This highlights that MGNREGS is not a genuine social insurance because it predominantly provides insurance against shocks that affect whole districts of villages at the same time.

This is, in fact, what I show in my paper. While there is rationing of MGNREGS work, participation for the programme is high following district-level shocks, for example following a bad monsoon.

In its first term, the ruling National Democratic Alliance put emphasis on promoting entrepreneurship through the Skill India Initiative and Make in India programme. Budget allocation to the Ministry of Skill Development and Entrepreneurship (MSDE) has increased 237% over the last four years, according to government data. But, for the first time since 2013, MGNREGS funding will drop by 1.8% in the 2019-20 budget. Are other programmes likely to work better than MGNREGS? In its second term, what should the NDA government’s approach be?

I am quite concerned about MGNREGS being hollowed out. It is a policy that--(though) far from being perfect--is delivering something that has been missing in India’s social policy: An effective form of social insurance particularly benefiting the rural parts of India. Rather than hollowing it out by depriving it of the funds it needs, focus should be on investing in the programme to ensure even less leakage to allow it to function.

Upskilling is definitely important but I have yet to see evidence that these programmes actually work. It is very important that MSDE engages with academics to evaluate the initiatives against their effectiveness.

The Make in India initiative is interesting but again, it does raise important questions as to what policy instruments are being used. Manufacturing of tradable goods is likely to be an important source of private sector employment that can absorb workforce from rural areas but I worry that the policy tools used focus around the government picking select industries or even companies, rather than focus on building the right institutional support to facilitate private initiative. It is important for any such policy for the government to transparently engage with all stakeholders--the focus should be on developing an enabling and supportive institutional environment to foster private initiative.

India ranked amongst the highest religion-related social hostilities index, at 9.6 compared to the world average of 1.8 in 2016, according to a Pew Research Center study. The motivations for this kind of conflict are identity-based, whereas the Maoist conflict is to a significant extent motivated by land rights and other development factors. Do you believe rural income guarantee programmes such as MGNREGS could mitigate conflict motivated by religion/ethnicity?

This is an interesting question. The focus of my study was indeed on the Maoist conflict. But I do think that the scheme may have some positive features as well that may be of relevance to other conflicts. The fact that people participating in MGNREGS work side by side is something that may actually foster and improve social cohesion. Of course, if at the local level implementation and access to work becomes co-opted then just as any other government programme it can become a tool fostering exclusion and producing grievances. So again, it depends on the specific design and the realities on the ground.

By 2030, the share of the global poor living in fragile and conflict-affected areas is projected to reach nearly 50%. How do we prepare for this?

As indicated, I sincerely think that developing countries more broadly should consider programmes like MGNREGS--not because they are perfect, but because they can actually work vis-à-vis other initiatives that may end up providing un-targeted transfers to households that do not actually need them. It is absolutely important that the organisation of these programmes is sufficiently decentralised to ensure that they do not become a tool to be co-opted for political gain. MGNREGS is a good example because since its introduction we have learned quite a few things already as to how we can make it work better. So other countries can really learn from the Indian experience here.

It is also important to not confound MGNREGS with other types of cash-for-work programmes--some of which have been studied, for example in Tunisia. The key value of MGNREGS is the fact that it is there to fall back on. Most cash-for-work programmes are temporary by nature and hence, do fail to provide social insurance simply because they are not guaranteed to be available in the future when demand may be there.

So MGNREGS has a social value. Even if nobody is doing any work, the certainty that there is a fall-back has a significant positive effect by itself.

(Habershon, a graduate from the University of Manchester, is an intern with IndiaSpend.)

We welcome feedback. Please write to respond@indiaspend.org. We reserve the right to edit responses for language and grammar.