Blowing a gaping hole in the Asylum Wall

America's civil asset forfeiture regime has led to evil. Because police there get to keep a good chunk of what they seize, some local governments have allowed their police to fund themselves by stealing from people. I am not exaggerating. 

New Zealand brought in civil asset forfeiture in 2009. It is not good. But it is not as bad as America's.

I wrote in last year's Outside of the Asylum essay:

Incentives in New Zealand are not nearly as perverse as in America. Police here do not directly profit from asset seizures. But they can apply to the pool of funds established by seizures. In 2016, Prime Minister John Key gave millions to anti-meth efforts from seized assets.

If the seized proceeds of crime are not used to compensate victims, those proceeds should be part of general government revenues. If police drug enforcement activities become self-financing because of asset forfeiture, police attention may plausibly shift towards drug crime – at the expense of less profitable lines of policing.

But it's worse than I'd thought then. I did not know that National had set police a KPI for gang asset seizures of $400m by 2021. And Stuart Nash has just increased that target to $500m.

A Cabinet paper seen by Stuff shows Police Minister Stuart Nash and Police Commissioner Mike Bush have set four new "high-level outcome targets", while also retaining most of the previous government's nine performance targets at an operational level.

The targets include $500m in cash and assets seized from gangs and criminals by 2021.

Nash said a small number of key targets would help focus police on priority areas, and since taking on the job, he has been clear about his plan to focus on gang-related crime.

KPIs on seizures can provide similar bad incentives for police, depending on what the rewards and penalties are for hitting those KPIs.

Here's the snapshot of the Cabinet paper from Stuff.

Asset seizures should never ever ever be a KPI target. At best they could be an intermediate target towards some actual KPI of, say, reducing overall criminal activity. But targeting it directly seems an exceptionally bad idea.

In case it is less than obvious why it is an exceptionally bad idea, here are some examples.

• If gangs own less than $500m in total assets, the police have strong incentive to deem more people as being gang-affiliated so that they can steal more stuff from them. The same holds true regardless of how much gang-affiliated people own. What matters is how easy it is to seize assets from different people, and how easy it is to claim that those people are gang-affiliated. 
• It encourages seizing the easiest-to-seize stuff, which may not be the assets that are either most directly tied to actual criminal activity or the assets most critical to continued criminal activity;
• Police always have choices about how to spend their time. Those choices depend on the KPIs. Should police be prioritising work leading to potential seizures over other work? Shouldn't they be instead trying to minimise the overall burden of crime, and figuring out what measures work best toward those ends?

As a general rule, whenever NoRightTurn and I agree that a policy is terrible, we should have a joint veto on it.