Migration has helped rural incomes and, to a certain extent, agriculture. Typically, migrants from rural areas are short-term migrants. Often, adult migrants take their children with them, and this leads to the overall picture being distorted. A 2010 study on the impact of short-term - often as short as a month - migration on the children of migrant labourers finds the process doesn't lead to child labour, but results in school drop-outs. "Child Labour and Short-Term Migration", a research paper by Diane Coffey, a Princeton University scholar, suggests the only way to ensure children in rural areas continue to attend classes is to entice their mothers to stay back - if mothers migrate, children follow. The Mahatma Gandhi National Rural Employment Guarantee Scheme (MGNREGS), which grants 100 days of wage employment in villages, could be a useful tool in this exercise. In villages across the three states the survey covered - Rajasthan, Madhya Pradesh and Gujarat - the scheme provided work for a mere 11 days. The study was based on a survey on migration across the three states in 2010. Of the 700 households across 70 hamlets in five districts, it was found 50 per cent of the population comprised tribals, 71 per cent households had no electricity and most adult women were illiterate. Of those aged up to 13, 30 per cent had migrated along with their families in 2009, against 80 per cent of the adults. In the 10-13 age group, 18 per cent had migrated. According to the survey, these households had three primary sources of income - agriculture, migrant work and MGNREGS. Agriculture is predominantly rain-fed. Therefore, migration is crucial, especially in the summer, when agriculture is unproductive. Coffey said the survey found very little child labour among migrant children.
In cases where it did, the children were from a slightly higher age group. It was found 20.5 per cent of the children were involved in domestic work (for their own households) when they last migrated; 5.7 per cent worked for pay; 3.3 per cent helped adults in their work, but were unpaid; 2.1 per cent went to school; and 79.5 per cent "did nothing". Paid or unpaid work (other than household chores) was done only by children aged at least 10. Coffey also pointed to a finding that could explain why child labour rates were low. Most parents, she said, worked at construction sites, for which they were paid pre-decided daily wages, rather than a piece rate. "Employers may not want to risk hiring an unproductive worker or getting into trouble with the law, as most construction work is done outdoors. It is common to see children at work sites, but uncommon to see them working." But there was a downside. On an average, schooling for migrant children aged up to 13 was two years less than those who didn't migrate. Coffey said this could be addressed by creating conditions that induced women not to migrate, as children migrated with their mothers. "The factor that best explained whether or not a child migrated in the past year was whether or not his/her mother migrated. Thus, it may make sense to have policies that try to address child migration by focusing on giving mothers a reason to stay home. The survey suggested a strong MGNREGS programme might be a way to do that," she said.