CSI: Crime Statistics Investigation – An interview with ROA research fellow Olivier Marie

How does crime participation relate to income and job opportunities? Whatever assumptions you might have, the topic hadn’t yet been researched in a thorough, scientific way.

That is, until Olivier Marie, PhD, received a Veni grant allowing him to apply advanced micro-econometric techniques to vast datasets from Statistics Netherlands (CBS). “We can already see that two or three years before the crime is committed, there’s a drop in income.”

“Pinning down causality empirically, measuring the relative importance of various factors that affect the complex relationship between labour market opportunities and criminal activity, and assessing whether effect sizes vary across different personality types and circumstances is a challenging task”, Marie wrote in the summary of his Veni project.

And indeed it is. Marie speaks English with a nice French accent, energetically jumping from one aspect of his research to another.

Amazing data

It all started with the unique data gathered by the CBS. “One of the reasons I moved to the Netherlands after doing my PhD in the UK was that the data is amazing here.”

In large surveys, starting at the age of 12, Dutch citizens are questioned about their education, character traits, income, friends and more. If you combine this with the police information on everyone arrested between 1997 and 2012, the result is a dataset with over a million individuals followed over ten years.

“We can link the two datasets using the personal identification numbers. And then we can see what happens with things like a person’s legal income a few years before he or she commits a crime. Not everyone will turn to crime after an income drop, so we also try to identify certain groups of people who are more likely to commit a crime. Is it young people, ethnic minorities? Immigrants in the Netherlands are disproportionately represented in crime statistics, but if you control for education level and parents’ social background, the effect disappears. And then we can even try to pinpoint certain personality traits that enhance the likelihood of committing a crime.” Individual anonymity and data security is, of course, always guaranteed.”

Policy effects

Using sophisticated econometric methods, Marie will also try to find out what causes the income shocks that seem to play a major role in turning to crime.

“It may be that certain Dutch policies have had a big effect on the income of some parts of the population. For example, research in the United States has shown that it’s better to pay welfare benefits to poorer people weekly rather than monthly. In the latter case, three days before the end of the month you’d see a big increase in crime. People with a low income are more likely to commit a crime, so you need to keep this in mind if you really want to influence crime participation.”

The final stage of the research project thus looks at policy implications. The optimal way to reduce crime is through prevention, as this is much cheaper than repression.

“Locking someone up in jail costs a good salary a year. Estimates show that ten percent of people who commit a crime are responsible for fifty percent of all crimes. So if we can target those individuals with policies that prevent an income drop, we might be able to prevent a lot of crimes.”

Solid science

But, as Marie stresses throughout the interview, the most important thing is that the research is carried out in a solid, scientific way.

“Up to the 1990s, lots of crime research was done semi-badly. The distinction between correlation and causation was not always made.”

Take the question whether having more police reduces crime. “You would assume that it does. But a lot of research in the 80s and 90s concluded that in areas with more crime, there are more police, therefore police increase crime. That’s reverse causality. One of my PhD papers showed that having more police has a strong effect on reducing crime; it’s one of the only European papers on this topic. I did my PhD ‘Essays on the economics of crime’ between 2005 and 2009, and I’m from one of the first generations of economics PhDs that was taught to do crime research very seriously.”

Beat Scandinavia

His prediction that the Netherlands will ‘beat’ Scandinavia, the present leader in data quality for crime research, appealed to the Netherlands Organisation for Scientific Research (NWO), which awarded him the Veni grant after a tough selection procedure.

At least, that’s what he thinks. “I’m worried about how difficult it’s becoming to get a Veni scholarship. You can only apply a few years after your PhD and from what I see from the candidates you need to have published at least two A-rated papers. I was lucky to have three, but economic papers take on average more than three years from start to publication.”

Apart from his Veni research, he is attempting to obtain the data necessary to investigate the impact on crime of the different coffee shop bans in Maastricht in recent years.

“You might not find a better example of drug tourism than Maastricht,” he concludes with enthusiasm. “Crime research goes in many directions.”

Olivier Marie is Post-Doctoral research fellow at the Research Centre for Education and the Labour Market (ROA) since February 2009. He is a also research associate at the Centre for Economic Performance at LSE and a member of the Home Office Academic Advisory Panel. He received a PhD from Royal Holloway in 2009 and has an MSc from LSE. His main research interests are in economics of crime, economics of education, applied econometrics, and public economics.

Source: Maastricht University webmagazine, 5 December 2012

Post Your Thoughts