Modern Slavery Research


This is a summary of my recent and current research relevant to modern slavery. The dates refer to publications listed (with links) on my publications page. For example 2020d indicates publication (d) in 2020.


Approaches to estimating prevalence (2018a, 2020d, 2021b, 2022b)

Modern slavery is an example of a hidden crime, where there is typically a large `dark figure’ of victims that do not easily come to attention. Many different approaches have been taken to try to get a reasonable estimate of the extent of this crime, in various contexts. One of the standard methods is multiple systems analysis, which collates lists of known victims and uses a mathematical model to quantify the number that are not observed. This was the approach used in constructing the estimates that supported the Modern Slavery Act 2015. Subsequently, I have written a number of papers discussing and comparing various methods that have been proposed, and have also applied multiple systems analysis elsewhere, for example in New Orleans.

Multiple Systems Analysis methodology (2020e, 2021a, work in progress, R packages)

Multiple systems analysis has been used for a wide range of different problems. The application to modern slavery has motivated much broader methodological development, for example studying the role that Bayesian and bootstrap methods play. Particular attention has been paid to methods for constructing confidence intervals that take account of the inevitable model selection intrinsic to the approach, as well as the appropriate way to handle data where many of the possible overlaps between the various lists of cases do not actually occur in the data. Software packages in R have been written to implement the methodology developed.

Covid-19 and child criminal exploitation (2022a, 2021c, 2021d, 2021e, 2020a)

One form of modern slavery is the exploitation of children in the context of County Lines drug supply. Detailed qualitative research showed that the Covid pandemic and its consequences induced shifts to the drug supply model that had a significant impact on some young people’s vulnerability to exploitation, on the way in which police and frontline practitioners respond to County Lines and child criminal exploitation, and on the way illegal drugs are being moved and sold.

The modern slavery statement registry (2022c)

The Modern Slavery Act 2015 requires larger businesses and other organisations to publish an annual statement setting out the steps that they are taking to address modern slavery risk in their operations and supply chains. There is now a government modern slavery statement registry that aims to be the repository of these statements. A study of this registry indicates that the coverage of the repository is possibly not as comprehensive as might appear on first sight.

IP addresses and child sexual abuse (ongoing work, with Rowland Seymour)

Reports of online child sexual exploitation to government agencies typically include source IP addresses. The digital marketplace is dynamic, with IP addresses dropping in and out of use, for example because of being detected, blocked, or leaving the market of their own accord. This dynamic nature of the system needs to be taken into account when estimating the number of relevant devices. An approach combining Markov modelling and multiple systems estimation is being developed and investigated. This is of specific relevance to fighting online abuse with victims based in the Philippines.

Globalization and modern slavery (2019a)

The relationship between globalization and modern slavery is contested. Some have argued that a race to the bottom, and the structure of economic incentives associated with globalization have contributed to the problem of modern slavery. Others argue that increased openness and the diffusion of values, the spread of democratic forms of rule, and the advance of human rights that come with globalization limit modern slavery. An empirical analysis of these arguments using data on slavery prevalence across more than 60 countries and various measures of economic and political globalization shows that economic measures of globalization and higher levels of democracy are significantly related to lower levels of slavery prevalence, even after controlling for armed conflict and regional differentiation.

Mapping research on modern slavery (2018b)

Evidence-based response is key to ending slavery, and research plays an important role in informing the policies and practices that underpin this anti-slavery effort. Our report and accompanying website therefore survey the UK’s modern slavery research landscape. We seek to understand the evidence base, and identify strengths and gaps in our research understanding.