Thursday 25 July 2013

Predictive Policing and Race

On facing pages in The Economist's print edition this week: a piece on predictive policing and a piece on Trayvon Martin. The juxtaposition raises questions that are worth exploring. Will predictive policing help keep future Trayvons, Oscar Grants, and Amadou Diallos alive? Or will it make matters worse?

Predictive policing purports to identify high-risk locations for crime within neighbourhoods, based, in part, on recent crimes there and nearby. It also promises to help identify likely perpetrators (see also this unfortunately gated New Yorker piece on predicting domestic violence in Massachusetts). I want to start with the direct effects on police interactions with the public, before turning to vigilantes like Zimmerman and how predictive policing might change regulation.

The first Economist piece outlines some problems with prediction:

Misuse and overuse of data can amplify biases. It matters, for example, whether software crunches reports of crimes or arrests; if the latter, police activity risks creating a vicious circle. And report-based systems may favour rich neighbourhoods which turn to the police more readily rather than poor ones where crime is rife. Crimes such as burglary and car theft are more consistently reported than drug dealing or gang-related violence.

However, the Economist is still overall optimistic about the effects of predictive policing on race:

But mathematical models might make policing more equitable by curbing prejudice. A suspicious individual’s presence in a “high-crime area” is among the criteria American police may use to determine whether a search is acceptable: a more rigorous definition of those locations will stop that justification being abused. Detailed analysis of a convict’s personal history may be a fairer reason to refuse parole than similarity to a stereotype.

In sum, prediction will allow the cops to do profiling better, and more fairly. I think this is overly optimistic. When a model identifies a high-risk location, the police show up and look around. But what are they looking for? Many appear to be looking for "suspicious-looking" black male teenagers with "furtive" movements. A mathematically-derived predictive model won't stop them doing so, no more than a hunch or prejudice about a violent neighbourhood.

It might, indeed, make matters worse. We're not just bad at predicting. We are bad at understanding prediction. We don't think enough about margins for error, we distort extreme probabilities badly, and we have very weird patterns of risk acceptance and risk aversion, especially at extremes of probability. What's more, the ex ante likelihood of crime on any given night in any given location is pretty small. So even if Block A is more likely than Block B to experience a robbery, we can't say with much confidence that a robbery will occur at Block A. You could make money betting that robbery is more likely to occur on Block A than Block B, but not by betting, at even odds, that a robbery was going to happen on Block A on a given night. But I don't know that we understand all that well: we may become inclined to believe that because Block A is "risky" in the sense of "more risky than Block B," it is then risky in the sense of "having an absolutely, not relatively, high risk of crime." That would be a non-sequitur, but it's one I think people will commit. Under the illusion of computers and their cachet, police may well become more inclined to shoot first. And what goes for a police goes for a jury.

Ultimately, if a cop thinks a location is high-risk and sees a black kid looking "furtive," whatever the hell that is supposed to mean, I don't think it matters much whether a computer told them the location was high-risk or whether their hunch told them. Whether we have a dead kid or not comes down to the cop's ethics, training, professionalism, and prejudice. How a jury responds depends on their prejudices and those of the legal system. There are massive problems with each.

As for vigilantes: it is probably not encouraging to contemplate Zimmerman having access to predictive crime models. Even absent that scenario, I think I'd ideally want the predictive models that are being used to be able to predict when a vigilante is about to be a Zimmerman. I have no idea how feasible this is. It may not even matter: Zimmerman was told by a police dispatcher that he shouldn't chase Travyon, and he did anyway. A cop on the scene may have helped, or, going by all of the above, may have made everything worse. But it is, in any case, well to remember the Economist's warning about what kinds of crime get plugged in to a predictive model. Is it the break-ins alone--i.e. what supposedly prompted Zimmerman to be driving around that night? Or is it the act of a Zimmerman taking the law into his own hands? And if that isn't defined as a crime, what happens?

Thursday 4 July 2013

This was a coup. But then, so was 2011.

What just happened in Egypt was a coup, by prevailing scholarly definitions. But the thing is, by the definitions offered, so was the endgame in 2011. Here is Zoltan Barany in Journal of Democracy:

The generals concluded that Mubarak’s mix of concessions (agreeing not to seek reelection or have his son succeed him) and repression (the February 2 attacks) had failed, and that rising violence and disorder would only hurt the military’s legitimacy and influence. Thus, on February 10, the Supreme Council of the Armed Forces (SCAF) assumed control of the country and, the next day, persuaded a reluctant Mubarak to resign and head for internal exile.
This sounds rather similar: the armed forces replacing a leader unconstitutionally (say what you will about the constitution in question) after several days of hundreds of thousands of people protesting. I want to throw out a new word for the conjunction of uprising and coup: the "couprising." Let me explain.

Even at the time, some scholars and commentators treated Mubarak's ouster primarily as a revolution but also as a coup. Posts at the Monkey Cage referred to comparisons to 1989, the revolutionaryauthoritarian-durabilitycollective action, and media frames; Erik Voetens also referred to a paper by Marinov and Goemans on the consequences of coups on democracy. He did so again today. The armed forces are, as was recognized at the time and as corresponds well with what we know about nonviolent uprisings, pivotal to revolutionary success: there isn't necessarily too much contradiction.

But looking back now, it's easy to forget that it was a coup. I think it's the case that when people now, in the present, talk about what happened in 2011, they'll remember that the military got involved and was ultimately the agent to oust Mubarak. But they won't mention that the intervention constituted a coup at the time. Here, for instance, is the New York Times editorial today (emphasis added):

Despite his failings, and there were plenty, President Mohamed Morsi was Egypt’s first democratically elected leader, and his overthrow by the military on Wednesday was unquestionably a coup. It would be tragic if Egyptians allowed the 2011 revolution that overthrew the dictator Hosni Mubarak to end with this rejection of democracy....
The military was the force behind Mr. Mubarak, then became the protesters’ protector when Mr. Mubarak was overthrown. Subsequently, the generals became the interim government, only to incur the peoples’ wrath when they proved inept at governing yet clung to power even after Mr. Morsi was elected.  
So, in the first paragraph, the current coup is juxtaposed against the 2011 revolution. Then, in the second, the passive voice allows the editorialist to avoid the question of who it was who, in the end, overthrew Mubarak. The military "subsequently...became the interim government", but this is a little misleading: subsequently suggests a time lag between Mubarak's ouster and the military becoming the government--a lag that did not take place.

There are several possible reasons why the two events are viewed rather differently from our vantage point--why 2013 is a coup, but we don't remember as easily that 2011 was too. I want to consider two related ones.

The first is that "revolution" was an available frame in 2011, but not so much in 2013. And this is, to a large extent, fine: just a sensible way of dealing with different events. Treating the events of 2011 as a revolution captured the major reason why they mattered: the ousting of a dictator of thirty years and the end of a seemingly invulnerable regime in favour of a major shakeup in political rules. To zero in on the coup instead would be, to some extent, to miss the forest for the trees. In contrast, on July 2nd, 2013, Egypt was probably a hybrid, "competitive-authoritarian" regime trapped in that difficult "grey zone"; after July 3rd, 2013, it is probably just the same, with someone different at the helm.

So, it is entirely possible that the "revolutionary" frame has pushed the "coup" frame out in the memory of many thinking back to 2011. If--as seems rather unlikely--the coup this year manages to install a stable democracy, we might come to treat it similarly.

A second reason is correspondence to normative positions. Morsi was elected, Mubarak was not, and so Morsi's ouster makes it feel more appropriate to use a term with more pejorative connotations, like coup, over one with more positive connotations like revolution. But this corresponds well, also, to the resistance among many Egyptian liberals to calling what happened a coup. Coup seems to lend Morsi an air of legitimacy (to use Morsi's word) that he should not have had, given his "Calvinball" approach to government--making up rules as he went along to further his own control of the situation.

On this analysis, whether you call what happened a coup has a lot to do with whether you support it or not. And there are, to be sure, good reasons why liberals in Egypt and elsewhere would be against it: like it or not, Morsi was elected, and the massive wave of popular protest in the last few weeks could potentially have been channeled into votes, even in an election without a level playing field. Calling it a coup draws attention to the strong possibility that the Egyptian "deep state" military establishment is reasserting its control.

But the word coup, if that's how the events of the last few weeks are reduced, also draws attention away from the massive protests that Egypt has had, again provoking the resentment of some Egyptian opposition members. A friend of mine, a liberal Egyptian activist, posted to Facebook one of the ubiquitous overhead photos of Tahrir Square, with the caption "Egypt's 'Coup.'" This is with plenty justification. This terrific piece by Trinity College Dublin PhD student Catriona Dowd, based on ACLED data, suggests graphically that protests have, if anything, lasted longer and happened more broadly this year than two years ago.

Dowd asks whether Egypt is experiencing a second revolution. Karl Sharro, getting right at the ambiguity of the officers' motives and intentions, poses a similar possibility in a more cynical vein:

I think it is very important to note that what happened in Egypt, in 2011 as in 2013, involved both a popular uprising and a coup. One also involved a clear change in the rules of the political game, the other... we don't know yet, but there's a good chance not. We have a tendency to apply one label to complex processes. So 2011 is easy to read as a revolution--covering the public uprisings, the change in rule, and, if we think about it but often not, if we don't think about it, the military intervention. 2013: not a revolution. And now, to fight over one term for 2013--coup or uprising?--would be to miss that it was both.

Perhaps we need a single word for the conjunction of an uprising and a coup, capturing their ambiguity. Off the top of my head, how about "couprising"? Thus Egypt had a couprising in 2011, and in 2013; the first was also part of a revolution, the second, we don't know. Or it may be that settling on a single word is itself a problem.