What can we learn from India’s Mass Employment Scheme?

I. Introduction
Countries with large rural populations that include high shares of landless laborers typically require formal safety nets to reduce vulnerability and sustain people’s livelihoods. India already has several large, universal safety net programs, including the Public Distribution System (PDS) for food and the Integrated Child Development Services (ICDS) scheme, and in 2006, it launched another. The National Rural Employment Guarantee Scheme (NREGS) emanated from India’s commitment to a legal right to work. Adopted in the wake of persistent public demand, the NREGS is a mass ‘public works program’ (PWP) based on the country’s experience in reducing human distress in recent decades in rural Maharashtra state. It promises to sustain the incomes of rural people while creating physical infrastructure that will benefit the country in the long run. Although the program is relatively new and thus has not been subject to a comprehensive assessment, early indicators point to areas of success and failure, highlighting where improvements can be made and lessons learned in a public employment program of this size.

II. The National Rural Employment Guarantee Act/Scheme (NREGA/S)
The ‘right to work’ is a ‘directive principle’ in Indian Constitution which was formalized through the enactment of NREGA (2005) . It is seen as a mechanism of income transfer, infrastructure development and promoting rural production and consumption markets - a multifarious strategy indeed. NREGA has found priority policy attention in India’s 11th five year plan (2007-12) under a broader objective of ‘Bharat Nirman’ aiming for resurgence in rural areas. Some consider nrega a natural response to non-inclusive growth that occurred during reforms process of last about two decades. The format of nrega and its nationwide implementation was a result of persistence by civil society and activists which is a common mechanism to influence policy in India. NREGS is unique, being large in size , intended to cover long periods, disburse huge funds and be dynamically responsive to climatic and rainfall conditions and above all open to any adult intending to work for wages often lower than local causal wages. Since self-targeting is inherent to scheme, besides chronic poverty manifest for example in food inadequacy, it also intends to mitigate idiosyncratic risks and shocks faced by households due to being differently-abled or death of earning member. NREGS can attract the unemployed or underemployed rural youth; because of immediate cash availability and 100 days of assured work which functions as a short-term relief objective. Indian policy appears confident that nregs can be important normally, even in the absence of price or income shocks and that it can smoothen seasonal fluctuations in labor demand and, therefore, wage rates in rural areas where rainfall patterns and insufficient irrigation preclude year-round crop cultivation (see also Barrett, et. al, 2004). Other objectives include, generation of productive assets, empowering rural women, reducing rural-urban migration, fostering social equity and environmental protection. Box 1 summarized the multiple nrega goals which can be categorized as protective, preventive and promotive in nature.


Multiple Goals of NREGS

Protective : assistance ensuring consumption smoothening and immediate welfare

Preventive : insurance facilitating risk taking such as investment in agriculture

Promotive : economic stimulus, through cash accumulation, local production and enhancing markets
- harmonize labor market, promote rural nonfarm employment (RNFE) and equalize wage rates;
- create durable physical assets impacting local eco-system and climate change;
- facilitate human capital formation through skill development; and
- alleviate poverty, effect equity, reduce distress migration and empower women.

The listed issues, however, are yet to be thoroughly investigated in the context of nrega.



How was the NREGS formulated: The format of nrega/s draws upon positive experiences of reducing human distress by Maharashtra ‘employment guarantee scheme’ (MEGS). The MEGS was first of its kind effort in India, to address high levels of rural distress caused by a severe drought in 1973-74. The MEGS created 3,597 million person days of work on minor irrigation, soil and water conservation, reforestation and local roads since inception until 2004. In its peak year in 1980 about one fifth of capital spending of Maharashtra was spent on EGS. MEGS thus became a model of sorts in which all aspects of governance, political commitment, bureaucratic efficiency, equity objective and concerns to sustain local environment and so on came into play, although at the heart of the issues was clearly the push and pulls of electoral politics. This scheme had central focus on drought proofing activities that led to measurable tangible results. Land holders were direct beneficiaries and the scheme also established a participatory process in the local governance (Moore and Jadhav 2006).

While the MEGS was a prototype and used as a benchmark, before the launch of nrega a plethora of PWPs were tried out all over India. National rural employment program (NREP) was initiated in 1980 followed by rural landless employment guarantee program (RLEGP), and in 1989 they were merged to form Jawahar rozgar youjana (JRY) with a focus on unemployed and creating rural assets. Employment assurance scheme (EAS) introduced in 1993 became universal by 1997-98. Jawahar gram samridhi yojana (JGSY) introduced in 1999 restructured the JRY and made it a central scheme. Sampoorna grameen rozgar yojana (SGRY) was launched in 2001 and in 2002 EAS and JGSY were also merged, aiming at providing rural wage employment as a mechanism to ensure food security, along with the creation of durable community, social and economic assets. The ‘national rural employment guarantee act (NREGA)’ thus is the most resent version of PWP, which for the first time has a legal sanctity inbuilt into the scheme.

Coverage of the program: The national level rural development ministry and associated departments are responsible for implementing nrega across the whole country. They also appropriate financial allocations from national budgetary mechanism and facilitate states to draw respective shares to execute work projects. The nrega coverage during first year of implementation (2006-7) was only to 200 poorest districts, followed by an additional 130 district during 2007-8; and by 2008-9 all the 610 districts across India were brought under nrega. A review of data published biannually through official website suggests that in the year 2007-8 , overall 33.7 million households were provided with 1.43 billion man days of nrega employment and distributed close to Rs. 86 billion. These absolute numbers which account for about 45% of all rural households, suggest a vibrant and highly efficient program implementation and matches with the stated policy and the targets.

Budgetary allocations: Before formalizing the Act, government agencies estimated that full coverage of nregs will cost Rs. 400 billion (about US$ 9-10 billion) which was about 1% of GDP. Some empirical assessments suggest that nrega could help reduce rural poverty to 23 per cent during lean season, at annual cost of 1.7 per cent of GDP (Murgai and Ravallion 2005). Others based on simple average minimum wage aggregates of all states estimated the national annual cost to be 1.3% of GDP; and a case was made that nrega will be sensitive to prevailing minimum wages in respective states (Shariff, 2004).

The allocation and expenditure on nregs during first two years of implementation were low due to staggered implementation; yet even after covering all 610 districts in 2008-9, only Rs. 160 billion (0.37% GDP) were allocated. The allocation for the year 2009-10 is Rs. 780 billion or 0.66 % of GDP, but the actual expenditure is expected to much less. NREGA expenditure for 2007-8 although 0.23 % of GDP compares well with other national programs namely targeted ‘public distribution system (PDS)’ of food products (0.13% GDP) and ICDS (0.10% GDP) respectively. However, a cost-benefit estimate for erstwhile MEGS compared with PDS suggested a ratio of 21.6 % for the former and meager 11.2% for the latter (Parikh et. al. 2007). Although there are substantial improvements lately, the relative efficiency differentials are likely to prevail. The Indian safety net programs appear huge in terms of allocations and coverage but reach a small proportion of poor and only nominal benefits are received by them.

Given both low allocation and under-utilization of funds, further financial expansion of nrega are not expected to impact national level fiscal deficits adversely even if no additional tax or levy is imposed; nor does one expect that nregs proceeds would cause inflation in the local economy. An increased cash flow amongst wage workers will pushup local demand and prices, but a simultaneous accelerating effect on local product market associated with broad based income growth will dampen the inflationary impact. However, nregs appear good in reducing risks and vulnerability, but not sufficient to eliminate poverty in India.


III. Strong Performance in Some States

Given a vast expanse one expects interstate differentials in nrega performance, some due to staggered and lagged universalization others because of administrative bottlenecks unique to each state (CAG, 2007). Corruption, leakage and dominance of private agents are other notable hurdles which are insurmountable where panchayats are nonexistent. Some appreciate self-targeting strategy inherent in nrega, but cautions social welfare losses caused by miss-targeting (Banerjee, 2008). Overall all, Rajasthan, Madhya Pradesh and Chhattisgarh stand out as better performing states. The relative coverage of households is better and average wage accruals are reasonable due to better provisioning of employment days. Given relatively higher poverty it is reasonable to assume that relatively vulnerable households do benefit in these states.

In case of Rajasthan, there are satisfactory assessments as to who benefits from nregs. Wage accruals have helped smoothening of household income (Scandizzo, et al., 2007) and there are large favorable gender impacts (Chandrasekhar and Ghosh, 2009). A silver lining thus has emerged from state of Rajasthan (Shivakumar 2006, Dreze, et al. 2007) which indeed provide much needed confidence to strive for improvement and continuity of nregs. Although vibrant panchayat raj administration is essential, but it is not a sufficient condition for better nrega; for example, Kerala with well developed panchayats is suffering from high leakage of funds. The success story of Rajasthan is due to a fair balance between functioning of panchayats and other enabling factors such as open and transparent monitoring and public audit mechanisms.

IV. Poor Performance in Most States

States which host large number of the poor but doing poorly in nrega are Bihar and Uttar Pradesh. Other states utilizing meager expenditures are Haryana, Gujarat, Karnataka, Kerala, Maharashtra and West Bengal. Another backward state, Jharkhand suffers from absence of panchayat institutions (Bhatia and Dreze 2006). Orissa’s problem is attributed to systemic bureaucratic failure to put in place transparent implementation of nregs. Large scale leakages and corruption are rampant due to absence of documentation system amenable for crosschecks for accuracy of record keeping (Dreze, et al. 2007), and the situation in Bihar is not encouraging either (Pankaj or Sharma 2006).

While nregs suffers from large exclusion errors due to poor coverage, one finds some hope due to greater access that it gives to women, the Scheduled Castes and the Scheduled Tribes (Chandrasekhar and Ghosh, 2009). A recent field study of vulnerable rural households in seven north Indian states (Shariff, 2008) suggests that community participation; information sharing and formulation of an opinion of program stand out as dominating factors that enhance maximization of receipts from nrega. Such attributes are normally prevalent amongst ‘rural middle class’ and therefore poorest of the poor within a micro locality are most likely to face entry barrier to nregs. Mechanisms to overcome such anomalies are available, for example, beneficiary participation and partnerships with local civil society and NGOs are known to have helped in ensuring transparency, equity, timeliness, financial prudence and quality assurance in delivery of public services. Such partnership also strengthens institutional capacity at the grassroots. The future reforms or nregs-course correction must address such anomalies, lest the program is held ransom to middle class values.

 
V. What determines access to NREGA and maximization of employment days?

A perceptions survey of 3200 poor households about government’s safety net programs in sixteen selected most deprived districts in northern parts of India provides some rare data that are amenable for econometric analysis in identifying factors supporting access and use of the nregs. These data evaluates both nregs accessibility/enrollment and number of days of employment received per household form such employment. Qualitative information on participation of households in village level institutions (of local governance), frequency and transparency of panchayat meetings are also available. The data were subjected to the ‘correction for the selectivity bias econometric model’; which assessed in its first stage, determinants of enrolment and factored a computed lambda value in the second stage to find out characteristics supporting maximization of the number of days of employment from amongst those enrolled. The survey was conducted across seven states in north India, the results do present a close to accurate status of the nregs of this region and the main findings are highlighted below .

Accessing NREGS: The econometric analysis suggests that social variables have influence nrega enrollment in expected direction, for example the casual labor and illiterate households have easy enrollment into the scheme. However, the scheme is accessible to households belonging to all caste and religious communities suggesting the fact that the scheme is adequately broad-based although one expects the Scheduled Castes and Scheduled Tribes to show greater access. Important is that fact that governance variables such as participation in panchayat meetings and having an opinion about transparency in nrega meeting have favorable impact on choice of nrega work. View on ‘transparency’ was categorized as yes, no and no opinion. Both divergent views - that nrega meetings were ‘transparent’ and that they were ‘not-transparent’- show a large and significant positive effect on nregs as opposed to those households who did not have a opinion at all or the fact that they were truly not interested in having a view on local panchayat meetings. Given this strong and important relationship additional tests were performed using interactive terms with transparency but did not find significant additional advantage. This analysis suggests that program information and having-a-view on institutions that promote participation are important to determine enrolments. This finding provides a strong signal to program managers that to enhance reach and efficacy of nregs and similar other programs people’s participation is essential irrespective of whether such participants approve or disapprove of a program. There are a few studies which have discovered the importance of participation for example, Krishna (2006, 2001); Weinberger (2000); and Cohen and Norman (1980) in promoting access to program in developing economies and in rural areas.

To understand whether the scheme is accessible to the poor a comprehensive index (using a combination of productive assets such as animals, implements and durables) which captured economic standing was used in the model. It was puzzling to find that relatively better-offs have high accessibility to the scheme; and the evidence challenge the program expectation that nregs is designed to benefit the poorest of the poor through self targeting mechanism. Since nregs-wage in a number of states is higher than local wage rates the beneficiaries are the regular wage workers, and not necessary those abject poor whose reservation wages can be low. Although, by law all those who approach for work must be provided with jobs, there are many constrains that limits creation of all the desired / demanded / needed employment days. Such limits emerge from limited supply of funds, seasonal factors, non-availability of useful projects and local level idiosyncratic factors as well. Given such limitations, many nregs jobs will be appropriated by those who have a better bargaining capacity (can also be the better off within the group), thus inhibiting the poorest of the poor benefiting from the program.


Maximizing Days of Employment: The selection bias correction model used evacuated the mean number of days of employment conditional upon a household getting enrolled. Information on ownership and size of land was used as distinguishing variable which also reflect relative economic condition of the household. Limited numbers of variables show significant effects; for example, even after controlling for land asset variable with no impact, the ownership of pukka (good quality) house shows independent positive impact at less than 5% level. That is labor force living in pukka homes maximize netting nregs employment days. Another dominant effect emerge from a community level factor namely, institutional participation of woman in the village; followed by significant (less than 5%) and unexpected effect from those reporting fair food adequacy. Finally, households having a migrant family member show high degree of incapacity to maximize nregs employment days, although as we found above they were successful in enrolling into the program, however.

It appears that the inner story of who gets to work more number of employment days depends upon one crucial fact that shows little influence on whether one gets to work or not; but once one gets enrolled in to the scheme derives maximum benefit. This crucial fact is women’s formal participation (from within households or own community) in local self governance such as the panchayat, school committees, mahila mandlas and so on which has shown the most dominant impact suggesting the fact that household gets its employment days maximized when a woman from the household or even from own community participates in the panchayati raj linked village level institutions. Strategies to improve program efficiency must be undertaken jointly and concurrently by strengthening local self governance with clear distinctions made between political, administrative and fiscal decentralization (von Braun and Grote, 2002). Best results can accrue if political and administrative decentralization precede the financial but it appears that in case of nregs this step wise transition has not taken a clear shape yet. This finding is highly significant both to the national goal of democratic decentralization on the one hand and favorable implementation of nregs on the other.

An unexpected finding is that even households reporting fair food adequacy have recorded a large positive effect that gives credence to the fact, that even those households who do not feel the pinch of food inadequacy make efforts to hang on to nregs employment for maximum number of days and this effect is prominent even after controlling for other social and economic factors makes it puzzling . Given a vast array of mechanisms through which leakages and discrimination works, it appears that those who manage to show endurance can hang on to more employment through various partnerships that may develop with the managers of muster rolls and payments in this program. In case of the migrant households, what appeared to be a reflection of distress while seeking work turns out to be worse as they are not able to maximize upon wage receipts through higher number of days of employment. In fact such people seem to get penalized and factors that lead to such a situation are not yet clearly documented.


VI. Policy Implications

India’s experience so far suggests lessons for its own implementation of the NREGS as well as for other countries that implement large-scale employment schemes and a few are summarized below:

Lesson 1. Be sure that local institutions exist that are capable of implementing large PWPs. This Indian model depends entirely on the 3rd tier of government identified as the local panchayats which are in principal locally elected bodies reflecting not only people’s participation but also people’s interaction with such institutions on a daily basis. However, panchayat system is essential but not sufficient condition for the success of nregs.

Lesson 2. Make implementation transparent to all stakeholders. Transparency is a trait which is not well factored in program implementation in India. Number of welfare programs therefore suffers from lack of scope for monitoring and mid-course correction if necessary. In built monitoring through a transparent documentation and audit mechanism enables reductions in leakages, better targeting and cost efficient delivery. NREGA has responded well when there is transparency in implementation.

Lesson 3. Monitoring through social audits can help ensure accountability. A social audit is a process in which the people work with the government to monitor and evaluate the planning and implementation of a scheme or indeed of a policy. NREGA social audit can examine local records such as the muster roles, work requisitions and monetary transactions, by nregs-workers and civil society. Such social audits will benefit from use of modern information technology (IT) and banking networks. NREGA has immensely benefitted both by the IT, and the banking and post office networks across India resulting in reduction in corruption, leakage and improvements in timely payments to the enrolled participants.

Lesson 4. Reducing ‘exclusion’ and ‘inclusion’ errors; the former being highly anti-poor and the latter reflecting program inefficiency are essential to register success for such a large scheme. It is ideal, therefore, that political and administrative decentralization precedes the financial one so as to enhance coverage and make nregs inclusive. Sustaining nregs-wages below the minimum wage seems essential so as to eliminate the crowing-out effect adversely impacting self-targeting, thereby causing exclusion.

Lesson 5. A larger policy issue relates to nregs promoting labor market distortions and impacting upon the natural process of migration (see also World Bank 2009). NREGS can inhibit rural to urban and rural to rural migration affecting employment and wages in both place of origin and place of destination across the country. Such a possibility arises when the nregs wage is arbitrarily fixed higher than the reservation wage of the poorest amongst the rural communities. Higher wages will deepen the exclusionary and wrong inclusionary process thus defeating the very purpose of large PWPs. Besides, if nregs create basic public service infrastructure in rural areas it may in fact inhibit migration of the poor in search of such services.


In conclusions it is useful to emphasize, that future of nregs, which is a legal entitlement for the deprived living in rural areas, is securely tied with functioning of the panchayati raj institutions in India. To some this scheme is a consolidation of democratic process and appears revolutionary. But test of the success of such a large scheme is in its ability to carry vulnerable and the poor on board and keep them there for extended period of time. Further mechanism for an exit from nregs is a useful policy to think about. Due to deepening effects on farming and increase in land productivity the household incomes can rise above poverty line or at a level when reservation-wage will be more than nregs-wage, in which case many marginal and small farmers, may not further nregs work. This aspect is not recognized and no policy guide lines exist as ‘exit strategy’. It is useful to document the processes which provide leads to nregs’s relevance for poverty alleviation through improvement in agriculture.

In spite of powerful demands from the monitoring agencies, academics and activists, one finds lack of coordinated government level initiative, innovation and interventions to improve the program. In the end it must be stated that nregs has potential to provide social security to the masses only if its implementation is efficient and synergies are exploited. India should not miss another opportunity to demonstrate that world’s largest democracy also cares for its people especially the deprived and vulnerable, and that it truly is marching ahead to become a welfare state.