Wednesday, February 27, 2008

Immigrants Less Likely to Commit Crimes Than Americans... If You Leave Out Key Data


A new study by Public Policy Institute of California seemingly indicates that immigrants are less likely to commit crimes than U.S. citizens. The study found that U.S.-born men are two and a half more times likely to be incarcerated than foreign-born men.

The report also found that amoung males 18-40, the group most at risk, the U.S.-born men were 10 times more likely to be incarcerated than foreign-born men. The report uses this data to claim that any efforts to increase education level requirements for visas or tougher penalties for criminal immigrants would be a waste of time and money.

These finding all sound reasonable if you stop reading the report at what is contained in the press release, but if you dig a little further you'll find that they admit their data is flawed in several aspects.

For example, their data only looked at criminals that have been incarcerated in prison.

Differences in treatment between the foreign born and the U.S.-born at any juncture in the criminal justice system may lead to differences in institutionalization rates for a given level of criminal activity. Such differences could skew our findings. For example... if the foreign-born are swiftly deported for criminal activity, then their institutionalized numbers will be low relative to their actual criminal activity.

If someone comes to the United States legally, a conviction of anaggravated felony qualifies that personfor deportation, unless he or she is a naturalized citizen.


How much does deportation matter? According to the report they were unable to determine this factor because not enough data is provided by the federal government. Any analyst worth a grain of salt knows that if bad data is input you'll get bad data out. Garbage-In-Garbage-Out, as the saying goes.

They also admit other factors may contribute to the apparent discrepencies:
[I]mmigrants may affect public safety in ways other than direct involvement in criminal activity. For example, immigrants may induce more criminal activity among the U.S.-born by displacing the work opportunities of the U.S.-born; in other words, immigrants may “take away” legal jobs, possibly leading to more crime among natives.

One thing the report does make clear is that this is no simple issue. With the lack of available data it will be difficult for anyone to draw any reliable conclusions as to the real impact of immigrants, legal or otherwise, on crime in the U.S.

1 comment:

Anonymous said...

I am not buying this. Kansascitian, the same article can be found at the Modesto Bee (Ca.)

http://www.modbee.com/local/story/222291.html

Note, from the modbee article is this passage,

The institute's numbers stand in stark contrast to federal prison statistics. Though immigrants made up about 12 percent of the U.S. population in 2005, they made up 27 percent of the population in federal prisons, according to a U.S. Government Accountability Office report.

Now who are you going to believe, A liberal think tank from San Francisco or a GAO report?

Until I know the motivations of the think tank, I will stick with the GAO figures.