There was an interesting paper published by the World Bank (March 1, 2012) – Does India’s employment guarantee scheme guarantee employment? – which offers some insights into how the Indian employment guarantee works. I thought it was an odd title because by definition the NREGA scheme is an employment guarantee. The relevant issue is a guarantee to whom. The World Bank research confirms the outcomes of my own work on the Indian scheme that it’s conditionality reduces its effectiveness. Those who gain jobs benefit but there is a shortage of jobs on offer relative to the demand for them. Modern Monetary Theory (MMT) shows that an unconditional, demand-driven employment guarantee, run as an automatic stabiliser, is the most superior buffer stock approach to price stability. Conditional (supply-driven) approaches not only undermine the job creating potential but also reduce the capacity of the scheme to act as a nominal anchor.
I discuss the concept of a Job Guarantee in considerable detail in this blog – When is a job guarantee a Job Guarantee? . Essentially, as developed within Modern Monetary Theory (MMT), the Job Guarantee comprises an unconditional and universal job offer at the minimum wage (calibrated to be an inclusive living wage) to anyone who wants a job. It would mostly eliminate unemployment (except frictional) and probably eliminate underemployment.
The reality is the Job Guarantee approach is the only guaranteed way that the national government can ensure there are enough jobs available at all times without activating an inflationary spiral. It is a very modest approach given those aims – choosing to work via the automatic stabilisers.
The NREGA was considered by the Economist article (November 5, 2009)– Faring well – where they noted that “India’s grand experiment with public works enjoys a moment in the sun”.
The experiment they were referring to is the National Rural Employment Guarantee Act (NREGA), which was re-named the Mahatma Gandhi National Rural Employment Guarantee Scheme (MGNREGS) in 2009.
The Scheme guarantees 100 days of minimum-wage employment on public works to every rural household that asks for it. The adults must be willing to undertake unskilled manual labour a the legal minimum wage. Once an adult applies for work, the scheme must employ them within 15 days or pay unemployment benefits.
Local communities have input into the selection and design of the jobs and local government plans and implements the work activities.
The Indian Government maintains a wonderful NREGA Home Page with a wealth of information available.
The Economist article noted some administrative issues associated with the NREGA scheme but concluded that:
Despite such flaws, the NREGA is winning praise from unexpected quarters. One reason India weathered the financial crisis of the past year was the strength of rural demand, many economists argue, and one reason for that strength was the expansion of the act to every rural district in April 2008. Once dismissed as a reckless fiscal sop, the scheme is now lauded as a timely fiscal stimulus. Because it must accommodate anyone who demands work, it can expand naturally as the need arises.
The World Bank article says that in introducing this MGNREGS “India embarked on an ambitious attempt to fight rural poverty.” They consider the ways in which the scheme fights poverty:
1. Direct employment and income guarantee to poor rural areas.
2. Poor people will leave the scheme only when something better arises and non-poor will not seek to access it.
3. The minimum wage becomes binding for all casual work whether in the scheme or not. It thus “can radically alter the bargaining power of poor men and women in the labor market”.
4. It “can help underpin otherwise risky investments” because “(e)ven those who do not normally need such work can benefit from knowing it is available”.
In my 2008 book with Joan Muysken Full employment abandoned we wrote (p.255) that the NREGA was designed “to bridge the vast rural-urban income disparities inequality that have emerged as India’s information technology service sector has boomed”.
Was it an anti-poverty scheme as the World Bank Report suggests? Answer: Yes but it was designed to accomplish other aims.
The problem faced by the Indian government was that there was too much migration from the poor rural areas into the burgeoning urban areas. They decided, sensibly, to reduce the incentive for the migration by creating work and lifting standards of living in the rural areas.
The scheme, despite its shortcomings (see below) has been very successful and millions of jobs have been created and a noticeable dent in poverty has occurred.
Wages paid sometimes exceed the going private sector wage and this has led to complaints from employers who want to pay below what effectively becomes the minimum wage.
In my work I did for the ILO in 2008-09, evaluating the South African Expanded Public Works Programme and designing a minimum wage framework for the scheme (and hence the economy as a whole), the same issue was continually raised by those opposed to the Scheme.
I noted in my Report:
First, the aim of the South African government is to significantly reduce unemployment and to eliminate absolute poverty and also provide for an improved personal capacity to manage risk via savings (that is, reducing relative poverty). The proposal to increase the minimum EPWP wage is consistent with that overriding objective. However, maintaining sectors in the private labour market that pay “poverty wages” is not consistent with that policy. It in the interests of the South African economy that higher productivity employment is fostered rather than relying on low-wage, working poor jobs to absorb the unskilled labour force.
Second, the EPWP can serve an industry policy to promote a quickening of this move to a high-wage, high productivity economy by placing pressure on market economy employers through the wage floor it establishes
At the time, I argued that if the EPWP wage became the national minimum then both demand and supply effects would be present. Employers currently paying below the wage would be confronted with the decision of operating that new legal minimum or closing down. What happened to the workers who lost their jobs depends on how many EPWP jobs were created and the impact of the higher wages on spending and overall job creation. There would also be a dynamic present to restructure existing employment.
If the EPWP was scaled up into an unconditional wage offer to anyone who wants a jobs (which we recommend) then the supply effects are likely to be significant.
In this latter context, employers paying below the proposed minimum would start to find it difficult to attract labour as the EPWP jobs (being always available, local and better paid) would become far better alternatives to the available labour. The employers would then be forced to invest in productive capital to increase the productivity of labour and pay at least proposed minimum per month to retain labour.
There may be some cases where a worker would agree to working below that if the job provided them with other non-pecuniary rewards that compensated. It is unlikely that all the workers who are currently earning below proposed minumum per month would be attracted to the EPWP.
If the EPWP minimum wage became the statutory minimum in South Africa, it is clear that this is the most desirable way in which to introduce and sustain a national employment guarantee system, then the private sector employers would face an immediate need to restructure their workplaces (invest in higher quality capital) to meet the new legal minimum wage levels but more importantly to stop the migration of their labour forces to more attractive EPWP employment.
Some employers would close their operations because they would not be able to operate at the higher costs. Economic development always involves a movement from lower productivity-higher cost production to higher productivity-lower cost production.
The ability of the EPWP to absorb this displaced labour would depend, in turn, on its scale. If there was a true EPWP safety net operating then these closures would shift workers into higher income areas and represent an improvement. That is the rationale of using the EPWP as a quasi-industry policy which can stimulate the South African economy towards the desirable high-wage, high productivity growth path.
The World Bank article examines the concerns expressed by opponents of the Scheme that the “the wage rate on MGNREGS is being set too high, relative to actual casual labor market wages” are unfounded.
In fact, “for India as a whole the two wages are quite close”.
Jobs versus cash handouts
The Economist article then makes an interesting point:
Policy wonks argue that cash handouts to the poor would be easier to administer, and would leave the recipients free to work the fields or roll beedis for private employers. But the poor themselves seem surprisingly sceptical of such an idea. “If money comes for free, it will never stay with us,” one elderly farmer says. “The men will drink it.” To wring anything out of India’s calcified bureaucracy takes a fight. If people feel they have earned their money from the government, they become more determined to claim it, even if that means waiting all day outside the village bank.
Mainstream economists always think cash handouts are better way of solving income insecurity because it allows the recipient to choose and it is always assumed that the individual knows best. The problems with this conception are manifest. The most obvious one noted in the quotation above is that individuals do not hunt alone. They tend to have families and have to assume wider responsibilities.
So often more enlightened policy advocates will argue that “in-kind” transfers are better – because at least the children will be fed! The NREGA is a twist on that theme and raises another very important point.
The “in-kind” component is the job. And it works to get “food on the table” instead of “grog down the throat” because people intrinsically value their involvement in productive work. There is a sense of achievement in earning one’s living.
Mainstream economics textbooks have labour-leisure choice models to determine labour supply. Labour is a bad and hence undesirable, leisure is a good. The only way you will engage in a bad is if you are paid. But this extremely blinkered view of the world that mainstream economists have reflects their ignorance of work in other social sciences. Economists have one of the worst records as a discipline in cross-citations of other disciplines. They just think they know everything – and end up knowing nothing much at all.
Other disciplines (sociology and psychology, for example) reveal that work is seen as desirable (rather than a bad) and the value of it extends far beyond the wage earned. That is one the reasons I always advocate creating jobs rather than paying basic income guarantees or other forms of income support.
The NREGA is a cut-down partial Job Guarantee, in the sense, that the latter is unconditional and demand-driven (that is, the government employs at a fixed price up to the last person who seeks work) whereas the NREGA is conditional and supply-driven (that is, the government rations the scheme according to some rules – number of jobs, hours of work or some other rationing device).
The following Table is taken from Full employment abandoned (Table 9.1, p.257) and compares the ideal Job Guarantee with the partial schemes introduced in South Africa (EPWP), Argentina (Jefes) and the NREGA in India. You can readily appreciate that the underlying modern monetary theory (MMT) which drives the JG conception is missing from the other schemes.
For example, in the NREGA sub-national governments, which are revenue-constrained, contribute to the investment outlays of the scheme. Further, the scheme is not universal nor unconditional. Nor does it provide a permanent job offer which means there is no “buffer stocks” capacity available in India.
Also, given the wage arrangements (some of the wage can be paid in food) the NREGA does not provide a comprehensive nominal price anchor. By paying a minimum wage to all workers, the JG creates what we call “loose full employment” – in the sense it places no pressures on the price level in its own right. It can also be used to discipline the inflation process by redistributing workers to the fixed price sector. There is no such capacity in NREGA.
Finally, NREGA provides no training capacity. The ideal JG would integrate skills development into the unconditional job offer and thus build dynamic efficiencies into the economy and provide the least-advantaged workers income security but a ladder to move to higher productivity jobs in the future.
The World Bank article considers this issue in some detail. The authors say:
The idea of an ―employment guarantee‖ is clearly important to realizing the full benefits of such a scheme. The gains depend heavily on the scheme‘s ability to accommodate the supply of work to the demand. That is not going to be easy, given that it requires an open-ended public spending commitment; similarly to an insurance company, the government must pay up when shocks hit. This kind of uncertainty about disbursements in risky environments would be a challenge for any government at any level of economic development.
If the maximum level of spending on the scheme by the center is exogenously fixed for budget planning purposes then rationing may well be unavoidable at any socially acceptable wage rate. Or, to put the point slightly differently, the implied wage rate—given the supply of labor to the scheme and the budget—may be too low to be socially acceptable, with rationing deemed (implicitly) to be the preferred outcome.
They consider a simple model aimed at assessing the performance of MGNREGS “in meeting the demand for work across states”. They note that the adminstrative data available from the Scheme (available from the link I provided above) shows that “52.865 million households in India demanded work in 2009/10, and 99.4% (52.53 million) were provided work”.
They say the such a claim “demand”” is unlikely to reflect true demand for work because there were administrative disincentives for states to provide a full reckoning of “excess demand” and other processes were likely to reduce the demand for work (ignorance, officials discouraging certain cohorts).
These are well-known problems with the Scheme but do not detract from the fact that in one financial year 52.8 million people were provided work.
To get a better estimate of “demand”, the World Bank authors used household survey data (“66th Round of the NSS for 2009/10 which included questions on participation and demand for work in MGNREGS”).
The following Table (taken from their Table 1) shows summary statistics for 2009-10 by Indian State. The Participation rate (%) is compared to the Demand rate (%) derived from the NSS data) to derive a so-called “Rationing Rate (%)” which is the extent to which unmet demand within the MGNREGS exists.
You can see that there are significant spatial disparities in the unmet demand (and the rationing rates). For the nation as a whole, PR = 0.249 while DR = 0447, given a RR of 0.444.
This means that 45 per cent of rural households in India desired a job under the scheme but only 56 per cent of them obtained work (100*PR/DR) which gives the rationing rate of 44 per cent (100-56).
The other point that the World Bank makes is that:
Participation rates in MGNREGS are only weakly correlated with the incidence of poverty across the states of India.
You can see that by assessing the first two columns of data in the Table. Relatedly, the “correlation between MGNREGS spending per capita and the poverty rate is -0.02 using spending in 2009/10 and 0.04 for 2010/11.”
The question then is “Why is MGNREGS not more active in poorer states?”. Note, first that the scheme wasn’t all about poverty targetting. It wanted to stop migration flows and the only way it could do that was to provide work in the rural areas.
But the question remains and we dealt with it in our 2008 book mentioned above. The World Bank article rehearses similar arguments which bear on the way the scheme is designed.
First, by forcing the states to share in the “budget” of the Scheme, the implementation is biased against poor states who are not as well endowed with available resources – given they are not currency issuers.
Second, the poorer states have less developed administrative capacity
Third, the poorest workers are “less empowered in poorer states”.
The overall conclusion is that the demand for the work in the Scheme is highest in poorer states, which is what we would expect.
But the World Bank research finds that:
… actual participation rates in the scheme are not (as a rule) any higher in poorer states where it is needed the most. The reason for this paradox lies in the differences in the extent to which the employment guarantee is honored. The answer to the question posed in our title is clearly ―no.
The problem with the MGNREGS is that is not an unconditional employment guarantee. Unless the government stands ready to offer work at the stated wage rate to all-comers then by definition it is not offering an employment guarantee.
By forcing non-currency issuing units within the nation (state and local governments) to share in the outlays required to render the Scheme operational, the Indian government is ensuring that rationing will occur. Under these circumstances, job creation becomes constrained by budget outlays.
Interestingly, the World Bank says that “despite the pervasive rationing we find, it is plain that the scheme is still reaching poor people and also reaching the scheduled tribes and backward castes.”
But once the “allocation of work through the local-level rationing” becomes a feature of the system it is obvious that there will be “many poor people who are not getting help because the employment guarantee is not in operation almost anywhere”.
The World Bank authors thus conclude that:
The first-order problem for MGNREGS is the level of un-met demand.
An open-ended guarantee is only viable if it is funded by the currency-issuer. It should be administered and operationalised locally but funded nationally.
Local communities, in liaison with their governments are best able to assess unmet community need which then defines productive avenues of work in the employment guarantee. But sub-national governmental units are always budget-constrained because, like households, they use the currency issued by the national government.
An effective employment guarantee has to be open-ended and unconditional. Governments think that large deficits are bad spend on a quantity rule – that is, allocate $x billiion – which they think is politically acceptable. It may not bear any relation to what is required to address the existing spending gap.
MMT shows you how it is far better to implement an employment guarantee by spending on a price rule. That is, the government just has to fix the price and “buy” whatever is available at that price (the employment guarantee wage) to ensure price stability.
That is enough for today!
(c) Copyright 2012 Bill Mitchell. All Rights Reserved.