Initiatives against slavery face two really difficult measurement questions: How many people are in slavery in the places we are working? And, are we achieving a reduction? In Northern India, we asked the Institute of Development Studies and the Praxis Institute for Participatory Practices in India to try to measure both.

If researchers use standard survey techniques to measure prevalence of slavery, they face a whole series of problems, not least of which is explaining what it is they are asking about – and then being able to elicit an honest response. The other problem with traditional surveys is that the people who most urgently need the data – the community members themselves – are usually the last to hear the results. To try to overcome these problems, IDS and Praxis developed a novel approach that shows promise for our field (see here for the full report).

To ensure descriptions and prompt questions were locally relevant, IDS and Praxis began by asking our local NGO partners to collect open-ended life stories, mainly from individuals affected by bonded labour and trafficking. These life stories were designed to understand the causes of slavery, the indicators of change in people’s lives, as well as which strategies are felt to be the most relevant for tackling slavery (see here for the report). In addition, based on these accounts, specific categories of slavery were refined to be directly applicable to the local context.

The IDS/Praxis team then created a pictorial survey tool that respondents could mark up. For data collection, households within the program communities were randomly selected and participants recorded their own information as well as information for their immediate neighbours, via a facilitated discussion.

Results were then immediately tallied and checked by the group of participants, helped by the NGO. They then discussed the findings, including questions like “Why is this happening here? How have some people got out of the situation? What can we do about this?”

The method gives people a chance to immediately start to use the survey results for locally relevant action. Central collating of the overall results of over 3,000 households is still continuing.

The method brings together the advantages of participatory methods (such as use of narratives, community mapping and joint analysis) with the rigour of standard statistical procedures such as random sampling and ensuring the sample has sufficient statistical power.

There were many interesting challenges: The NGO teams sometimes found crowds approaching the survey participant groups, asking why their household was not included; sometimes randomly selected participants did not wish to sit with members of a different caste; and training NGO field staff to facilitate the process took considerable training and re-training.

But facilitators reported some clear advantages: The participatory process helped to build trust – a prerequisite to obtaining an accurate answer on whether that person was in a situation of slavery. NGOs also reported that the exercise was useful for finding out issues they didn’t know much about, such as the extent of households taking loans for marriage.

This was a baseline assessment and so the IDS/Praxis team plan to follow this up in about two years to see whether there has been a reduction in prevalence in communities where our partners are working, as part of the overall hotspot evaluation. We also plan to use a similar method for measuring community-level prevalence of modern slavery in two other Freedom Fund hotspots. The baseline participatory statistics from northern India will be available shortly but in the meantime, learning from this method provides valuable insights for both measurement and community engagement.

Photo credit: Ginny Baumann © The Freedom Fund
Caption: Samples of tools used for the prevalence study