Poverty & Inequality

Workfare as an effective way to fight poverty: The case of India's MNREGA

  • Blog Post Date 11 December, 2014
  • Articles
  • Print Page
Author Image

Shamika Ravi

Brookings Institute, India Center


The fundamental appeal of a workfare programme, vis-à-vis a welfare programme, is that it helps in targeting the beneficiaries. This column assesses the welfare impact of MNREGA on poor rural households. It finds that the programme had a significant effect on extreme poverty in the first few years of implementation by improving food security, financial inclusion and mental health.

While developed countries are increasingly leaning on workfare programmes as a means to reduce work disincentives provoked by their far reaching social security systems, the concept of cash-for-work has gained importance in less developed countries as well. Looking back to a long history of food-for-work programmes in times of economic distress, developing countries increasingly run public works programmes not only to better target benefits to the poor (vis-à-vis welfare programmes) but also to use the emerging labour force to build up the rural economic infrastructure. India implemented the ambitious Mahatma Gandhi National Rural Employment Guarantee Act (MNREGA) in 2006, and since then, each rural household is guaranteed 100 days of unskilled wage employment per year within proximity of their residence.

India´s Mahatma Gandhi National Rural Employment Guarantee Act (MNREGA) is the largest safety net programme in the world, if measured by the number of participants. Although MNREGA has recently come under serious criticism due to low efficiency and high corruption (Niehaus and Sukhtankar 2013), it started with tremendous promise. There have been studies showcasing the positive impact of MNREGA in terms of agricultural wages (Azam 2011), indirect benefits in the form of higher private earnings (Imbert and Papp 2011), improvements in children´s educational outcomes due to rising women´s participation (Afridi et al. 2012), and higher consumption expenditure, intake of energy and protein, and asset accumulation (Liu and Deininger 2010).

Analysing welfare impact of MNREGA on participating households

So how do the different features of MNREGA – general access in rural areas, work guarantee, wage level, limited participation period of 100 days– combine to impact the welfare situation of the individual, participating household? Does the scheme achieve the intended purpose to alleviate (extreme) poverty? In a forthcoming paper, I along with my co-author Monika Engler, measure the welfare impact of MNREGA on participating households. We use data on 3,485 individuals from 1,064 poor rural households from Medak district1 in Andhra Pradesh over two years (2007-2009) (Ravi and Engler 2015).

The eligibility criteria for the extremely poor households chosen for the study was not having any male working members, owning less than one acre of land, not owning any productive assets, not receiving services from a microfinance institution and scoring less than a threshold number on housing condition. In our sample, 69% of individuals are widows, 27% are divorced and 4% are unmarried women. The male members in these poor households are usually fathers (or fathers-in-law) or children, all of whom are dependents.

In particular, we compare over time, average outcomes of households that applied for and received jobs under MNREGA with the average outcomes of households that applied for and did not receive jobs. The outcomes that we look at are food security (as captured by monthly per capita consumption expenditure, and number of meals foregone by household members), financial inclusion (probability of household holding any savings, and amount of reported savings) and health outcomes (physical and mental health).

The main reason reported by the local administration for denying applicants work was lack of worksites or limited job availability at existing worksites. Incidence of rejection was higher in the early years of MNREGA roll-out and reduced considerably in subsequent years as the programme scaled up in size and scope. In our sample, 63% of households that applied were denied work in 2007; this figure fell to 31% in 2009. Only a small fraction (5%) was denied work in both periods. Moreover, the programme was not implemented simultaneously in all villages within the district. Hence, there is significant difference in the likelihood of participating (or not participating) in the scheme depending on the location of household, or the village where households resided. To gauge the effects of the scheme, we exploit this exogenous variation in participation by households (variation that is not systematically related to household characteristics such as income, health etc.), over time and across the district.

Our baseline survey was conducted from August to October 2007 and the end-line was conducted from August to October of 2009. In our analysis, MNREGA was already in operation when the baseline survey was collected. This means that we are not measuring the effect of the introduction of MNREGA on welfare outcomes; rather, we are tracking growth in welfare outcomes over the course of MNREGA implementation.

Significant welfare gains in first few years of MNREGA implementation

We find that MNREGA had a significant impact on extreme poverty within the first few years of implementation. Participation in the programme improved sharply over the years. We find a significant increase (9.6%) in the monthly per capita expenditure on food among households that were given jobs. This result is similar to previous research by Liu and Deininger (2010), which finds a 10% increase in per capita consumption expenditure due to MNREGA. We also find an increase of 23% in monthly per capita non-food consumption. The programme also improved food security by a significant reduction in the number of meals foregone by participating households per week.

In terms of financial security, the programme raised the probability of a poor household holding savings by 21%, and the per capita amount saved per month increased 15% from Rs. 119, on average. With regard to health outcomes, there was a significant reduction of 12% in the incidence of reported depression. Other self-reported indicators of mental health such as anxiety and tension have also shown significant improvements over time. There are no major improvements in physical health outcomes that we measured such as number of sick days, number of days of work lost due to sickness and other self-reported wellness measures. While one would expect increased food security to have positive impacts on health outcomes of households’ members, the lack of effect on physical health can possibly be explained by the nature of MNREGA work, which is mostly physical labour. This is corroborated by the data on attitude towards MNREGA. We find that approximately 25% of the households which did not apply for MNREGA work, cited “work is too hard” as a reason. It is possible therefore, that the nature of manual labour in MNREGA negates the gains in health due to improved food consumption.

We find significant welfare gains in the first few years after the programme was implemented, but are unable to say much about its long term impact. The data indicates strong seasonality in participation within the first few years of the MNREGA. Participation was overwhelmingly in the summer months when alternate employment was unavailable, and remained low during the rest of the year. This was true at the all-India level as well as for five of the largest states. This seasonality in participation, however, gradually fades over the subsequent years and household participation became more evenly distributed throughout the year. This implies that after the first few years the MNREGA has been affecting the private rural labour market (including agriculture) closely. Rural households are moving in and out of the programme and private employment opportunities throughout the year. To gauge the overall welfare impact of this interaction between the MNREGA and the private rural labour market on poor households would require a general equilibrium approach and dataset richer than we possess.


  1. Medak is one of the poorest districts in India´s semi-arid Deccan plateau. This district was a part of Andhra Pradesh at the time of the survey; it is now a part of Telangana.

Further Reading

No comments yet
Join the conversation
Captcha Captcha Reload

Comments will be held for moderation. Your contact information will not be made public.

Related content

Sign up to our newsletter