Predicting White-Collar Crime

Affluent people commit white-collar crimes? Tell me more!

SuccessThere’s a familiar pattern to law enforcement techniques — first they start in street crime, then they move to white-collar cases.

Wiretaps used to be the kind of thing one used for mob cases. Now they’re used when the feds want to bring charges against people in the board room. Preet Bharara said as much in his Harvard commencement address. Cops, like most of us, love toys, and the white-collar cops aren’t going to be inclined to let the street crime people have all the fun.

There’s a new chapter in the story of bringing blue-collar methods to white-collar cases.

For years, models like Hunchlab.com have used data about street crimes and where they occur to try to predict where crime is more likely to happen. As Hunchlab.com explains it:

HunchLab is a web-based proactive patrol management system. Advanced statistical models forecast when and where crimes are likely to emerge. But it’s not just about anticipating crime, it’s about figuring out the best way to respond.
Policing tactics should not only be effective, but also reflect the community’s priorities. HunchLab provides features that: (1) align patrol activities with the priorities of the community, (2) intelligently allocate resources to prevent over-policing, and (3) determine which tactics work and which don’t.

These are laudable goals. Of course the government — or any entity — should use techniques that work and should intelligently allocate resources.

More troubling, some of these predictive models have been explored as a way for judges to sentence people who have been convicted, based on the likelihood that they will reoffend.
Propublica
has a piece describing this trend.

In some jurisdictions, such as Napa County, California, the probation department uses risk assessments to suggest to the judge an appropriate probation or treatment plan for individuals being sentenced. Napa County Superior Court Judge Mark Boessenecker said he finds the recommendations helpful. “We have a dearth of good treatment programs, so filling a slot in a program with someone who doesn’t need it is foolish,” he said.

Sponsored

There are, though, some odd results from this work:

“A guy who has molested a small child every day for a year could still come out as a low risk because he probably has a job,” Boessenecker said. “Meanwhile, a drunk guy will look high risk because he’s homeless.”

A recent white paper and web-based app applies this same data driven predictive model to white-collar crimes. Here’s the method:

We collected data provided by the Financial Regulatory Authority (FINRA) 14 to compile incidents of financial malfeasance dating back to 1964. Using these data, we correlated financial crimes to the location of the perpetrating individual or organization. Financial crimes were geographically clustered according to geohashes computed from these locations.

As a result of this work, the resulting app identifies a number of things that, it finds, are highly predictive of white-collar crime. The study notes that certain geographic features highly correlate with white-collar crime:

Sponsored

Screen Shot 2017-04-27 at 7.30.47 AM

Or, as the web announcement for the app and white paper explains:

Unlike typical predictive policing apps which criminalize poverty, White Collar Crime Risk Zones criminalizes wealth.

One would think, as a result, that Trump’s cabinet would be a prime area for criminal investigation.

This report is, obviously, a joke. And, to the extent that it isn’t, any experienced practitioner in this area would observe that using FINRA data is going to skew the results, because FINRA’s jurisdiction is generally limited to a subset of white-collar conduct that takes place where there are people interacting with the financial industry, which means it’s going to tend to be where there are wealthy people and skyscrapers. If it were accessible, using data on criminal prosecutions for what are commonly thought of as white-collar crimes would lead to likely different results.

Though, as a marketing exercise, where firms are looking for both clients who are likely to engage in behavior that invites scrutiny as well as can fund a high-powered defense, this app may prove most useful.

That said, if you didn’t already know that living in a wealthy neighborhood covaries with being able to afford an expensive defense, I’m not sure this app is going to save you.


Matt Kaiser is a white-collar defense attorney at KaiserDillon. He’s represented stockbrokers, tax preparers, doctors, drug dealers, and political appointees in federal investigations and indicted cases. His twitter handle is @mattkaiser. His email is mkaiser@kaiserdillon.com He’d love to hear from you if you’re inclined to say something nice.