The innocence principle
Like freedom of speech, the presumption of innocence before proof of guilt is something that almost everyone agrees is important in principle, but are occasionally reluctant to apply in practice. In recent weeks we have witnessed some examples of this reluctance that, to me, seem chilling. Eric Garner was an obese African-American who was killed by police officers holding him in a chokehold while they arrested him for illegally selling individual cigarettes in New York City. His last words are here.
Virtually everyone who has seen the video agrees that they acted with an extreme amount of force against a man who was not fighting back although he was resisting arrest (passively – that is, in a way that would not harm the officers).
A Grand Jury found that the police officers who killed Eric Garner did not act unlawfully. I defer to the Grand Jury on this, but assuming they are correct this suggests that the scope for lawful killing by police officers is extremely broad. As law professor Glenn Reynolds (and others) has noted, killings by police are treated much more sympathetically by juries than killings by civilians.
Michael Brown was an African-American teenager who was shot and killed by a police officer during an arrest after he (seemingly) robbed a convenience store in Ferguson, Missouri. There is still some disagreement about what happened here. The initial reports suggested that the officer executed Brown as he fled or begged for his life, but the subsequent Grand Jury investigation seems quite conclusive that Brown assaulted the police officer. The Grand Jury’s conclusions prompted looting by people in Ferguson.
If Brown’s shooting was unjust, the Garner lesson applies. But if the narrative found by the Grand Jury is correct then the protests, lootings and slandering of the police officer involved are wrong. In that case, it is the media’s presumption of guilt on the part of the police officer involved (even after the Grand Jury verdict) that has led to significant destruction and violence. People suspended the innocence principle to advance a political point, and the results have been bleak.
Jackie is a student at the University of Virginia by a Rolling Stone article which alleged that she had been gang-raped by a group of fraternity men. Last week Rolling Stone retracted the story after a number of facts given by Jackie in her story proved to be false.
The aftermath of the Rolling Stone story has been extremely disturbing, with very prominent people proudly dispensing with the innocence principle. The Washington Post ran a piece titled “No matter what Jackie said, we should automatically believe rape claims” (this was later changed to “generally” believe them). The Guardian’s Jessica Valenti wrote that “I choose to believe Jackie. I lose nothing by doing so, even if I’m later proven wrong”, and that “the current frenzy to prove Jackie’s story false – whether because the horror of a violent gang rape is too much to face or because disbelief is the misogynist status quo – will do incredible damage to all rape victims.” [my emphasis]
Has Valenti considered that someone else may lose something if we chooses to believe an accusation that is untrue? Or that we may have other reasons than misogyny or incredulity to want to know if a criminal accusation is false?
Sexual assault is very common, but this does not mean that false accusations do not occur. An estimated 1.5% to 7.5% of accusations may be false. Staggeringly, a 2012 study that used DNA testing of old physical evidence and exonerated between 8% and 15% of convicted rapists.
I know why Valenti is eager to believe Jackie: because not believing a genuine story is horrendous for the victim and makes other rape victims less likely to come forward, and hence makes rape an easier crime to commit. But the inverse is also true: believing a false story is horrendous for the wrongly-accused and makes other false accusations more likely. (The Rolling Stone story did not name individuals, but guilt-by-implication can still be enormously harmful.)
In all of these cases, people who would normally say that the presumption of innocence before proof of guilt is a good thing have assumed the opposite. The rule might work in general, they may say, but this case is an exception. Police need to be able to subdue people resisting arrest. The death of an 18-year old must be unjust. Rape is too serious an allegation to question.
Like the principle of free speech, the innocence principle only produces good results if we apply it rigidly and in cases where doing so may feel deeply unsettling.
The innocence principle matters because people who seem guilty may in fact be innocent. This is why mechanisms like jury trials exist – like the ‘thick’ version of free speech that I argued for recently, they are a mechanism for sorting the truth from lies.
Hayek speculated that liberal institutions like these evolved over time, because the societies that lacked them eventually fell behind the ones that upheld them. Politically and culturally, we may be witnessing an erosion of these institutions now. That would be a catastrophe. But it is not too late to change course.
Are immigrants dangerous criminals?
Sometimes supporters of strict migration controls criticise my focus on the economics of immigration. They more or less accept that free movement leads to productivity gains, innovation gains, entrepreneurial gains and fiscal gains for the receiving countries, ‘brain gains’ for the migrants’ home countries, and even that it leads to massive welfare gains for the immigrants themselves, but suggest that the non-economic costs of immigration are very large too, and we basically ignore them.
They have in mind costs like increased crime, reductions in social trust, a decline in democratic support for important institutions like the rule of law and free speech, and the dilution of the native culture. To some extent I believe all of these things are costs of certain kinds of immigration, and may justify certain controls on immigration, but none are reasons to support the immigration controls that we currently have.
In this post I’ll consider how much impact immigrants have on the crime rate. I’ll come back to some of the other points in future posts.
So: crime. Crime only seems to rise in line with certain kinds of immigration, and does so for basically economic reasons. A 2013 study looked at two different waves of immigration to the UK – asylum seekers in the 1990s and early 2000s (mostly from places like Afghanistan, Iraq and Somalia); and immigration from the “A8” countries (Poland, Czech Republic and the six other Eastern European states that joined the EU in 2004).
The paper looked at crime rates according to areas where these waves of immigrants settled. It found that neither of these waves had any statistically significant impact on the overall rate of violent crime. A8 immigration actually had a small negative impact on crime rates, driven by a reduction in property crime rates.
On the other hand, asylum seekers did cause a statistically significant rise in property crime, but still a small one. This equivalent to an increase of 0.7% in property crime rates per 1% percentage point share of the local adult that was asylum seekers. Note that this is a percentage increase to the existing crime rate – if crime rates were already 10% and a new wave of asylum seekers suddenly accounted for 1% of the population, crime rates would rise to 10.07%, not 10.7%.
According to the report’s authors,
“Across all England and Wales [asylum seekers] averaged 0.1% of the local adult population, so the average property crime rate might be 0.07% higher as a result – only around 2% of the average property crime rate of around 2.7%. Of course, some authorities had appreciably more asylum seekers located in the area, though shares larger than 1% of the local population were extremely rare.”
What explains this? There might be because of some inherent difference between, say, Polish and Iraqi people, but violent crime levels were not significantly different (in fact, A8 migrants were slightly worse here than asylum seekers). What may explain the difference is that A8 immigrants were free to work however they wished, whereas asylum seekers are not allowed to seek legal employment. The authors of the 2013 report suggest that this explains why only property crime was higher for asylum seekers.
This data is not broken down by nationality, so it is rather blunt. According to another 2013 paper, arrest rates (which are broken down by nationality) were significantly higher for immigrants than non-immigrants – 2.8 arrests per 1000 for UK nationals and 3.5 arrests per 1000 for non-nationals (excluding immigration-related arrests). However, controlling for age this difference disappears. Having more young people seems to lead to more arrests; in and of itself having more immigrants does not.
Finally, there is the issue of second- and third-generation immigrant groups who cause disproportionately more crime. According to the Metropolitan Police, the majority of men held by police for gun crimes (67%), robberies (59%) and street crimes (54%) in London in 2009-10 were black, even though only 10.6% of London’s population is black. There is no easy explanation for this, and it may be due to a combination of factors that we cannot control – or, indeed, easily change.
The ultimate lesson of all this may be that immigration in general does not have a big impact on crime, but certain immigrant groups might if they do not assimilate culturally. Then again, Eastern Europeans don’t seem to be a problem at all, and they seem to be the ones we’re most concerned about right now.
My friend Ed West has a point when he says that there is no such thing as an ‘immigrant’ – only an American, a Pole, a Somali, and so on. In the course of my posts on the non-economic impacts of immigration, I will suggest that a compromise position between my libertarian preference for very free movement and Ed’s conservative preference for restrictions. We may yet be able to square the circle of immigration policy.