Governance

Can MNREGA buffer negative shocks in early childhood?

  • Blog Post Date 29 August, 2014
  • Print Page

Exposure to negative shocks such as drought during early childhood is known to have lasting, detrimental effects on human development outcomes. This column examines whether a household’s access to MNREGA, later in the life of the child, can offset the impact of early childhood shocks. It finds that programme access, although incapable of correcting for past deficiencies, does mitigate the impact of recent shocks.

There is a substantial body of evidence highlighting the importance of early childhood environment and the role of shocks in affecting long-term human development outcomes. Recent studies (Maccini et al. 2009, Currie et al. 2011) indicate that exposure to negative shocks (example, deficient rainfall) in early childhood can have persistent negative effects on human development outcomes such as height-for-age, schooling years, test scores, cognitive skills etc. This is more often the case in developing countries such as India where there is very limited access to formal credit markets to smooth consumption1. For instance in the state of Andhra Pradesh in India, where over 80% of the population depends on subsistence agriculture, even slight deviation from the expected rainfall can substantially affect household income, which can have persistent negative impacts on child nutrition2.

In such crisis situations, poor households often have to resort to suboptimal coping mechanisms such as pulling children out of school, deferring healthcare expenditure etc. (Subbarao et al. 2012). This poses a serious concern as we know that investments in the first few years of life are the most critical and tend to have a persistent impact in the long run. Although it is well established that negative shocks affect human capital outcomes significantly, there has been very limited focus in the literature on examining the extent to which these deficits can be made up later on in life when there is access to a social safety net.

Assessing role of MNREGA in offsetting negative shocks in early childhood

In this context, my research analyses whether households are able to offset the negative effects of early childhood shocks3 when they have access to a social safety net in the form of the Mahatma Gandhi National Rural Employment Guarantee Act (MNREGA)4, at a later stage in the child’s life (up to eight years of age) (Dasgupta 2013). Although the main objective of MNREGA is not tied to improving child nutrition, it can be thought of as having an important indirect effect in safeguarding child outcomes5. The idea is that the employment guarantee scheme should ideally enable vulnerable households to achieve food security and thus, protect the long-term health status of the children. Surveys from medical literature (Martorell 1999) have found evidence of catch-up growth, especially for younger children, when living condition is improved - highlighting the potential role of early nutritional intervention in accelerating growth. The underlying hypothesis is that when there is an external income opportunity, in the form of wages earned from public work, it can feed into child investments and help overcome past deficiencies. This external source of income can be all the more beneficial in a situation of crisis6 , or for dealing with negative impacts of past crises.

The state of Andhra Pradesh provides a nice setting to examine this question as it has been among the better implementers of MNREGA by institutionalising social audits (Subarrao et al. 2013) and digitalising transactions of wage payments.

MNREGA has been implemented in a phase-wise manner making use of a ‘backwardness index’ developed by the Planning Commission7; the more backward districts of the country got access to the programme in the first phase (starting 2006), and the next two phases were rolled out subsequently in 2007 and 2008 respectively. My analysis exploits this variation in MNREGA over time and across regions, together with variation in drought exposure across households.

Data and results

My analysis uses data from the Young Lives Survey in Andhra Pradesh, India conducted over three waves (2002, 2007 and 2009-2010) along with administrative records of the programme and rainfall at the sub-district level. The household sample consists of 2,000 children who were around one-year old in 2002, from 20 sub-districts of the state. The main outcome variable in my analysis is height-for-age z-score which is a standardised measure of health status and is considered to be a long term indicator of well-being. I also look at the average percentage of stunted children in a sub-district as another variable of interest to get an aggregate picture of programme impact on health outcomes at the community level8.

For identifying the variation in access and intensity of MNREGA9, I primarily use the average number of work days provided under MNREGA per household for a particular sub-district, obtained from administrative records. The sample children were around five years of age when households in these districts got access to the programme. The annual record of rainfall for the monsoon season (June-September) at the sub-district level was used and deviation from the long-term average rainfall indicated drought.

I find that programme access, although incapable of correcting for past deficiencies, helps to mitigate recent shocks (in the past year) in health outcomes, especially in the case of lesser educated households and Scheduled Castes (SCs). Interestingly, there is no significant difference in programme impact by the gender of the child. The analysis shows that while exposure to drought in the past year significantly reduces height-for-age (by 0.4 standard deviations), access to the programme was able to mitigate about half of this negative impact. I find that an increase in 22 MNREGA days10 per household leads to substantial increase in height-for-age -for-age (by around 0.26 standard deviations) for those recently exposed to drought. Such an increase can bridge about half the rural-urban gap in height-for-age in my sample11. However, I do not find any significant effect of programme access on the negative impacts from cumulative past shocks (drought exposure from birth year till time of survey).

Policy insights

I find evidence that MNREGA helps mitigate recent exposure of children to negative shocks such as droughts, especially in the case of lesser educated households and SCs that are presumably more vulnerable. In this context, there is a lot of scope for policy to play a role by ensuring food security for vulnerable households.

Also noteworthy is the fact that the increase in height-for-age due to MNREGA is quite significant, given that the children were all around five years of age when the programme came in place. Perhaps, the mitigation effects could have been larger if it had come earlier in their lives.

Now, given that Andhra Pradesh has been one of the better performers in the implementation of MNREGA, one has to be careful in generalising these results for other states. The availability of longitudinal data (same set of children were followed from 2002 to 2010) was critical to the measurement of programme impact over time; it would be difficult to undertake the same exercise for other states due to lack of such data.

Finally, it is important to note is that any cost-effectiveness analysis of a programme which is based solely on labour force participation and income is likely to underestimate the total gains if it does not take into consideration the important spillover effects that accrue to the next generation in terms of improved long-term health status of children and potential associated human capital gains.

Notes:

  1. For example, if there is a negative shock to agricultural income due to deficient rainfall, farmers that have access to credit markets can obtain loans to fund their consumption during the lean period. These loans can be returned during periods of good rainfall (and hence, good income) in the future. In this way, their consumption levels remain more or less the same during periods of negative shocks.
  2. According to National Family Health Survey (NFHS-3; 2006), prevalence of malnutrition among children (0-59 months) in Andhra Pradesh is very high (32.5% underweight; 42.7% stunted).
  3. I separately examine cumulative drought exposure (fraction of years a child is exposed to drought from birth till the time of survey), and recent drought shocks (exposure to drought in the past year). Health outcomes for the same child are measured at three points in time: at birth year, at five years of age, and at eight years of age.
  4. MNREGA provides a legal guarantee for at least 100 days of employment in every financial year to adult members of any rural household willing to do unskilled manual work at the notified wage. It was rolled out across India in 2005.
  5. It would be crucial to look at this, especially in the light of proposed integration of various public programmes based on their synergies.
  6. There can be several other mechanisms at work by which access to MNREGA positively impacts child outcomes - through increased participation by females in the workforce and hence, higher earnings and more bargaining power in the household, and/ or reduction in migration. My analysis does not look into the exact mechanism through which the effect is taking place.
  7. The Backwardness Index comprises three parameters with equal weights assigned to each: (i) value of output per agricultural worker; (ii) agriculture wage rate; and (iii) percentage of Scheduled Caste/ Scheduled Tribe (SC/ ST) population of the districts.
  8. Height-for-age z-score shows the height of the child relative to an international reference group of healthy children. A child is considered to be stunted if the height-for-age is less than minus two standard deviations of the reference group. Stunting, or low height-for-age, is a measure of chronic malnutrition and is generally considered a long-term indicator for health status. It might become permanent when nutritional deficits begin early and are prolonged.
  9. 11 sub-districts, out of the total 20 sub-districts in the sample, were covered by the programme in the first phase of implementation in 2005-06.
  10. 22 working days is the average number of days available per household under MNREGA in a sub-district in Phase 1.
  11. The average difference in height-for-age between rural and urban is around 0.5 standard deviations in my sample

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