Policy wonders! Counter-intuitive results of Right to Work Act.

Mahatma Gandhi National Rural Employment Guarantee Act (MGNREGA)

MNREGA, National Rural Employment Guarantee Act, or The Right to Work.

It is one of India's flagship social protection programs. It is a policy which many economists, politicians and intellectuals were sceptical about with a view that wage increases without corresponding gains in productivity would likely reduce private employment and potentially attenuate impacts on poverty.

But the evidence from the positive effects on wages, employment, and incomes, and finding evidence of employer market power, suggest that programs like NREGS can not only reduce poverty, but also be effciency enhancing.

And you know what, I was not wise enough to actually judge a policy by Dr. Manmohan Singh, and designed by Jean Drèze in team, whom I consider my inspirations. So, I came across a wonderful paper by Karthik Muralidharan, Paul Niehaus and Sandeep Sukhtankar titled ‘General equilibrium effects of (improving) public employment programs: experimental evidence from India.’ which through a very robust research, found very astonishing and surprising results, though counter-intuitive!!

Overall, their analysis suggests that in addition to causing large gains in average household income, the reform generated net gains for over 92% of the population, with net losses seen only for large landowners.

By the way, why we need such programmes??

According to the mentioned paper, the Economic rationales for such programs include self-targeting through work requirements, public asset creation, and enforcing wage floors in informal labor markets by making the government an employer of last resort.

The paper at first presents results using their survey data for NREGS-registered households (henceforth, beneficiaries"), who make up 49.5% of all rural households and are the households the NREGS was designed to benefit.

Then turns to examine impacts on the overall economy using the various censuses. Finally, using a combination of survey and census data to characterize the distributional effects of the reform.

They concluded on a positive note stating,

An improved NREGS could boost labor demand in three ways.

First, it could increase the marginal product of labor by augmenting the stock of complementary physical capital.

Second, it could directly make workers more productive through human capital channels such as improved nutrition or skills.

Third, increased income from NREGS could boost demand for goods produced using local labor.

As a solution to the implementation issues in MNREGA, the Department of Rural Development of the Government of AP introduced ‘Smartcards’ for payments involving 2 major changes.

First, the flow of funds shifted in most cases from government run post offices to banks, who worked with local partners (called banking correspondents) to make last-mile cash payments in the village itself. Second, the protocol for authenticating when collecting payments changed from one based on paper documents and ink stamps to one based on biometric authentication using Smartcards.

Their success fastened the payments by 29%, with arrival dates 39% less varied, and took 20% less time to collect.

In economics, we learn that many a times, the impact of spillover effects is more than the impact of the actual activity or policy. The same happened in this case.

The reform raised beneciary households' earnings by 14%, and reduced poverty by 26%. Importantly, 86% of income gains came from non-program earnings, driven by higher private-sector (real) wages and employment.

This has worked out in an interesting way:

First, increased the workers’ reservation wages, i.e., the minimum wages at which workers would choose to work anywhere especially at private entities. It reduced the land returns because land owners would now have to pay higher to the labours and the employment gains were higher in villages with more concentrated landholdings.

Surprise!!! opposite to the conventional economics wisdom, non-agricultural enterprise observed a growing count with employment growth even rapid despite the higher wages.

Let’s again accept how important the local demand is for the structural transformation of any region, these results suggest that public employment programs can effectively reduce poverty in developing countries, and may also improve economic efficiency, a lesson to learn for economists and policy makers of our time.

If you doubt the sample size and scale of the paper, I would love to tell that the researchers worked with the Government of Andhra Pradesh (AP) to randomize the order in which 157 sub-districts (mandals) with an average population of 62,500 each introduced a new system (biometric/Smartcards) for making payments in NREGS during 2010-2012.

First and foremost, improving NREGS implementation substantially increased real incomes of the rural poor. Mean earnings among NREGS-registered households increased by 13.9%, leading to a 25.8% (7.4 percentage points) reduction in an income-based measure of poverty, while consumer goods prices did not change signicantly. Strikingly these income gains came primarily from sources other than the NREGS. Program earnings accounted for only 14% of income gains, whereas the majority (80%) of the total income increase came from private labor-market earnings.

Also, MNREGA lead to 10.1% increase in wages in the treated areas during June. Again counter-intuitive, in spite of higher wages, researchers see an increase, not a decrease, in market employment among NREGS-registered households, which rose by 20% (1.4 days/month), while days self-employed or not working fell by 13% (2.4 days/month).

It also worked wonders for poverty reduction, in a census of rural households, the proportion in the lowest income bracket fell by 3.4% (2.8 percentage points).

Employment in non-agricultural establishments in the Economic Census rose 34% (11 percentage points). The number of these establishments increased by 23% in parallel, with increases concentrated among small owner-operated enterprises.

I think humans appreciate the comparisons very much, so just for putting it in perspective, this employment increase within 3 years of the reform was 11% of the working-age population; whereas the share of the Indian workforce engaged in non-agricultural employment increased by 8% in the entire preceding decade (2000-10).

Calculations suggest that the income gains in the survey of NREGS-registered households fully account for the gains we see in the census of all households. This suggests that poverty reductions were concentrated among NREGS-registered households, as one might expect given they tend to be relatively more dependent on labor as opposed to land income.

As we could think that increased wages may have lead to reduced employment, but opposite to this, it amplified the growth in the region by even increasing the employment too. This happened because of three reasons:

(1) an increase in labor productivity.

(2) an inward shift in labor supply in the context of imperfectly competitive labor markets (e.g. oligopsony).

(3) increases in aggregate demand for locally-produced goods and services.

Clear evidence suggests, labor supply to private sector jobs shifted inwards implying that an improved NREGS increased workers' bargaining power by enhancing outside options. Moreover, changes in market employment covary systematically with proxies for employer market power.

Now, a doubt splashes that okay, I am convinced that income gains happened, but how can we be sure that it had translated to increased demand, it can be saved too, right?

But the survey suggests that most of the income gains were either consumed or used to acquire assets (but not saved), which would boost local demand. Finally, the large increase in the number of non-agricultural firms, and in employment in these firms suggest that the net benets of increased demand exceeded the cost of higher wages.

The next basis we can question the paper is that producers or employers have passed not he increased costs to the consumers, increasing the prices of basic products.

In this case, rural poor would not have benefited, because their increased earnings would have been going for the consumption at higher prices, but ‘this is not the case!!’, Very surprisingly, the increased costs have been borne by the producers themselves and not increasing the prices of any of the commodities of the consumer goods basket. Evidence from the paper:

Given that the reform impacted wages, it is also possible that it affected the local prices of final consumer goods, and thus the overall price level facing consumers. In this case the nominal earnings effects we reported above would overstate the real gains to beneciaries. Using either of our alternative methods we obtain estimates very close to zero and precise enough to reject effects as large as the 14% increase in earnings we observe. One implication of these results is that the costs of higher unskilled wages were absorbed by employers, rather than passed through to consumers.

Now the paper says that it was a win-win game for the entire economy, and had net positive gains for the entire economy, but as we know that MNREGA employment is a very small part of the entire labour market in India, then how can it influence such a large market share?

So, coming to third, results also highlight the influence public options can exert on private markets even when they themselves capture only modest market shares.

A similar premise holds in our data: only 7% of income earned by our control group came from the NREGS. Yet 32% of households actively participated in the NREGS at some point in 2011-12 (in NSS data), and the reform sharply increased their reservation wages. This underscores that the NREGS's impact depends not only on its “market share" but also its credibility as an outside option.

But now I again need to emphasis that these results are produced when the implementation was amazing, so I would like to make the point that instead of focusing on how past policies can be eliminated, we need to think on how their implementation can be improved.

Stating and Explaining some technical results:

• When the paper examines how many households' incomes moved across the expenditure poverty line, it estimates a 7.4 percentage point (25.8%) reduction in poverty among beneciaries.

• Reiterating again, 14% of the earnings gain account for NREGS earnings, and the rest 80% comes from private sector earnings.This pattern of effects aligns with the structure of control group earnings, where NREGS earnings account for just 7% of the total while wage labor earnings account for 35%.

• Consistent with the increase in reservation wages in treated areas, we also see large increases in market wages. The adjusted total effect is a Rs. 13 increase on a base of Rs. 128 per day, or 10.2%. Of this effect, roughly 2/3 is the main effect, and the remaining 1/3 is a (statistically signficant) spillover effect.

• we also see an increase of 1.4 days per month (18%) in private-sector employment.

• Days worked in the NREGS also increased signicantly by 1.3 days per month (29%). These days replaced time spent in self-employment or not working, which fell by 2.4 days per month. These days replace time spent in self-employment or not working, which fell by 2.4 days per month.

“The Socio-Economic Caste Census (SECC) classifies households into low, middle, and high-income categories based on whether their highest earning member reported monthly earnings below Rs. 5,000, between Rs. 5,000 and Rs. 10,000, and greater than Rs. 10,000, respectively.”

Quantitatively, the estimates imply that roughly 3% of households moved out of the bottom bracket, primarily into the middle one. Overall, these results validate using entirely independent census data that the reform had substantial impacts on population-level earnings and poverty.

Based on the state of Indian policy sphere, and a very limited understanding of mine as a 20 years old Economics student, I understand and would like to conclude that thinking for a policy and the policy outcomes should not restrict our scope only to the direct outcomes it does lead to, instead, I suggest and as very evidently proved by this paper, we should also focus on the indirect or spillover effects of a policy. It can provide us astonishing results and give us amazing on ground insights on what a policy is actually doing, and if it is doing wonders even in a small region, think how can we scale it and provide its benifit to the large sections of our society in need of that policy in the other regions of the nation.

If you subscribe, it’s absolutely FREE in terms of money for you, but it is invaluable in terms of the encouragement it gives me seeing my readers’ love and appreciation.