2003

In 2001, CARE India and the Population Council carried out an experimental intervention for girls aged 14-19 that provided reproductive health information, vocational counselling and training, and assistance with opening savings accounts in slum areas of Allahabad in Uttar Pradesh, India. Community participation and face-to-face capacity-building activities - in particular, peer group formation - were at the core of the effort.

The Centre for Operations Research and Training (CORT) collected data collection for the baseline survey between April and June 2001, prior to the start of the intervention. All adolescents aged 14-19 years who were living in the study areas for at least one year
were specified for inclusion in the baseline survey - both sexes, married and unmarried,
in school and out of school. This survey indicated the cultural appropriateness of the
intervention:

• Approximately half of the girls indicated that they had not travelled outside of Allahabad during the past six months, compared with only about one-fourth of the
  boys.
• Girls were much more likely than boys to report that they needed to seek permission to visit outside of their homes; both boys and girls reported that there
  were no places in the community where unmarried girls could safely congregate for any purpose.
• Girls reported spending more than 4 times as many hours as boys on chores (4.3 and 1.0 hours, respectively).
• The proportion of boys who reported that they had ever worked for pay was 5 times greater than that of girls (34.3% of boys vs. 6.0% of girls).
• Despite the fact that girls were much less likely to work for pay, they were more inclined to save; 54% of the girls and 26% of the boys had some savings.

The evaluation uses a quasi-experimental pre- and post-test design that compares
adolescent girls who participated in the intervention with those residing in a control area. Specifically, Allahabad's 143 slum areas had been divided by CARE staff into 7 wards for programme and research purposes. For the project, 2 comparable wards were selected; one was randomly assigned as the experimental site and the other as the control site. Five slum areas within the experimental ward were randomly selected as project sites, and 9 smaller areas were chosen in the control ward.

Evaluators point out that approximately 22 months elapsed between the baseline and endline interviews. Moreover, the intervention did not continue for the entire interval between surveys. The livelihoods training ended in June 2002, and follow-up support was available until December 2002, 3 months before the onset of data collection for the endline survey.

A total of 27 groups were formed by the peer educators. Of the 525 girls who took
part in the reproductive health information sessions, all but 38 completed at least one
vocational course.

Compared to the matched control respondents, girls exposed to the intervention were more likely to have knowledge of safe spaces for girls: At baseline, 16.9% of the intervention group had such knowledge; this percentage increased to 83.2% at the endline (22 months later) among this group. In contrast, 4.6% of the matched control group, compared to 33.7% of that same group at endline, had such knowledge. Again, compared to the matched control respondents, girls exposed to the intervention were significantly more likely to be informed about reproductive health. The mean knowledge among the intervention group increased from 4.0% to 6.7%; among matched respondents, that percentage change was less (from 3.9% to 5.7%).

Compared to the matched control respondents, girls exposed to the intervention were significantly more likely to:

• be a member of a group - increase was from 1.7% to 15.6% in the intervention group, but only from 9.2% to 11.0% in the matched control group (p<0.01).
• score higher on the social skills index - increase was from 9.6% to 12.0% in the intervention group, but only from 2.6% to 5.1% in the matched control group (p<0.05).
• spend time on leisure activities - increase was from 3.8% to 4.4% in the intervention group, but decreased from 3.9% to 3.7% in the matched control group (p<0.05).

Although the livelihoods programme was found to be acceptable to parents and feasible to implement, the project had only a minimal impact on the behaviour and attitudes of adolescent girls in the experimental slums: "No effect was found on gender-role attitudes, mobility, self-esteem, work expectations, or on number of hours visiting friends, performing domestic chores, or engaging in labor-market work." After explaining several factors that may have contributed to this lack of impact, the evaluators conclude that, "In order to reduce deeply entrenched gender disparities and enhance girls' ability to have a greater voice in decisionmaking about their own lives, however, future interventions should involve many more contact hours than did the experimental project described here. They should also devote greater effort to developing group cohesion and to improving communication, negotiation, and decisionmaking skills. Finally, substantially greater resources must be provided for data collection so that the program can be properly evaluated."

"The Effect of a Livelihoods Intervention in an Urban Slum in India" [PDF], by Barbara S. Mensch, Monica J. Grant, Mary P. Sebastian, Paul C. Hewett, and Dale Huntington, Population Council, 2004; and email from Debra Warn to The Communication Initiative on August 22 2006.