Global Slavery Index

The Global Slavery Index is an annual study of world-wide slavery conditions by country published by the Walk Free Foundation. In 2016, the study estimated a total of 45.8 million people to be in some form of modern slavery in 167 countries.

The report includes three data points for each country: national estimates of the prevalence of modern slavery, vulnerability measures and an assessment of the strength of government responses. The Index pioneered the use of random-sampled, nationally-representative surveys to estimate prevalence. This included commissioning seven such surveys in 2014 and a further 19 surveys through Gallup World Poll in 2015.

The 2014 Index includes country studies with policy recommendations for many countries including: Taiwan, Australia, Pakistan, India, Brazil, the United Kingdom, the United States, Qatar, and others.

The 2014 Global Slavery Index includes data on three key variables: the prevalence of modern slavery in each country, vulnerability, and government responses to modern slavery. The methodology is written up in detail in a methodology paper.

The first of these factors, the prevalence estimates, were derived using a statistical process known as extrapolation. In 2015, Joudo Larsen, Bales and Datta published an article that describes the extrapolation process, and provides a test of it. Writing in Significance, the magazine of the UK Royal Statistics Society, Joudo Larson et al. note that the estimates in the 2014 Global Slavery Index involved several steps:

This resulted in a prevalence estimate for each country, calculated as a proportion of the total population that was enslaved within that country. For all 167 countries this produced a total global estimate in 2014 of 35.8 million enslaved people.

Writing in 2015, Larsen et al. note that newer 2015 survey estimates now allow a comparative test of some of the 2014 extrapolated estimates. This is important for checking the validity of the earlier results and the continued application and use of this methodology. The comparisons suggest that extrapolation, while not perfect, is a useful and valid method. All but one of the extrapolated estimates of the prevalence of slavery fell within one percentage point of the estimates derived from random sample surveys.

...
Wikipedia