UK-based criminologists who work with linked administrative datasets often cast an envious eye over to the Scandinavian countries, where information about an individual’s criminal convictions is securely linked together with a wide range of other personal data through ‘register’ based systems. This means that researching changing patterns of conviction, the impact of justice policies, alternative sentencing practices or potential causes of offending in Sweden, Norway, Finland and Denmark can be a fairly straightforward process and has yielded many interesting and policy-relevant results. Whereas, back in the UK, we are currently unable to ask any of the same research questions because data about criminal convictions are not routinely linked to any other administrative datasets.
So when we were invited to Stockholm to attend a workshop on register-based criminological research in Northern Europe we jumped at the chance. The opportunity to discuss criminological research based entirely on administrative sources of data, compare methodological approaches and discuss possibilities for comparative analyses was too much to resist!
The workshop focused on sharing research findings from criminological studies that were conducted exclusively using linked administrative datasets. Teams of researchers from Scandinavia and the Netherlands presented a range of research projects that were of significant policy importance, including: the feasibility (and desirability) of predicting youth crime at the year of birth in Norway; the effect of parental imprisonment on children’s criminal offending in Sweden; detecting bias in policing through data in Sweden; and the changing socio-demographics of convicted offenders in Finland.
One presentation that was of particular relevance and importance to policy and practice in Scotland was presented by Hilde Wermink from the Netherlands. She had used linked data to compare reconviction rates for people given short-term sentences with those given prison sentences, a highly topical subject in Scotland. She explained that just comparing the crude reconviction rates for these two types of sentences was like comparing apples to oranges (because differences in sentences were likely to reflect differences in their characteristics and offending history). To make an ‘apples to apples’ comparison, Hilde had adjusted for differences in the characteristics of the two groups and found that people who were given community sentences had a lower reconviction rate than similar people who received short prison sentences. Being able to replicate such a study in Scotland, and test the impact of different non-custodial disposals, could provide strong support for reducing the prison population!
Another benefit of attending was that we could highlight where research using administrative data in other countries had had an impact on policy developments in Scotland. For example, Britt Østergaard Larsen described a study to test the impact of reducing the age of criminal responsibility from 15 to 14 using a ‘natural experiment’ approach in Denmark. Using data on convictions, they found no evidence of a deterrence effect (that is, the threat of being dealt with in the criminal justice system did not lead to a reduction in the conviction rate). However, being convicted did itself lead to a higher rate of reconviction and the change also resulted in a lowering of educational attainment and greater likelihood of dropping out of school. Britt was surprised to hear that these results had been influential in evidence presented to the Scottish Parliament about increasing the age of criminal responsibility in Scotland.
We presented some of the research that we've been doing as part of the Understanding Inequalities project and the Scottish Centre for Administrative Data Research. Ben described the convictions data held in Scotland in the Scottish Offenders Index (SOI) and provided some interesting illustrations of how it could potentially be used in comparative research with our Scandinavian colleagues. For example, it could help us to determine whether the recent falls in youth conviction in Scotland are due to uniquely Scottish factors – such as policy changes in Scotland – or are common across countries. Meanwhile, Susan presented findings from the Edinburgh Study of Youth Transitions and Crime, an award-winning study using linked administrative and survey data, which highlighted the detrimental effect of early and prolonged justice system contact on young people. Our presentations showcased the great potential for policy-focused research using administrative crime data in Scotland. We also discussed the great strides that Scotland has made in terms of establishing a secure infrastructure and governance framework for linking administrative data for research purposes.
While we thoroughly enjoyed presenting our work, it was disappointing to have to acknowledge that our ability to use the Scottish administrative data infrastructure to study crime and justice topics is severely constrained by the lack of connection between crime-related data, such as the SOI, and any other administrative datasets in Scotland. At the conclusion of the workshop, there was a strong appetite for organising future meetings and developing collaborative work together. However, hearing about the work going on in Scandinavia and the Netherlands made us realise just how far we still have to go in Scotland. For example, criminologists in the Netherlands are currently replicating findings from a study conducted in the early 2000s, whereas we have still to get off the blocks in terms of having convictions data linked to other administrative datasets in Scotland.
As the debates about Brexit reach fever pitch, it is more important than ever that UK-based academics form strong partnerships and working relationships with our counterparts in Europe. However, it is sobering to realise how far we still have to go, in administrative data linkage terms, just to be where our European colleagues were more than a decade ago.
We left Stockholm feeling somewhat envious of the work of our European neighbours. Nevertheless, members of the Understanding Inequalities project and the Scottish Centre for Administrative Data research have been working actively with data controllers in Scotland to identify ways of bringing crime and justice data into Scotland’s data infrastructure, so we are hopeful that Scotland will be able to catch up soon!
Photo c/o Raphael Andres via Unsplash