Author: Matt Liu, University of Wyoming director of debate Thanks to the online UW practice debates, I’ve judged more debates on the Predictive Policing topic than I ever thought I would. Thanks to that, we at WDR have a few more thoughts we’d like to share on the topic. Below the fold find our 6 new tips for debating Predictive Policing. 1. Getting the Most out of the Discrimination / Bias Contention (aff)
Arguments that predictive policing reproduce discrimination and bias are the heart of any good aff on this topic. Beyond the obvious, core of the literature, there’s two strategic nuances I want to talk about: responding to negative uniqueness arguments and utilizing the bias debate as a solvency takeout. I will talk about this more in the neg section, but the negative has one overwhelmingly compelling response to the discrimination contention: it’s all happening in the status quo. The argument that discrimination in policing is non-unique is a major potential trouble area for the aff for three reasons: it’s a capital T True argument, it’s supported by all the best aff evidence, and it amplifies the potential for offense (for a link turn- because a link turn’s best friend is a non-unique). When I say it’s supported by the best aff evidence, I mean that every aff card that explains why the data will be biased explains that that is the case because the police are producing biased statistics through status quo, non-predictive policing methods. The key to beating the non-unique argument that discrimination exists now is having a quantifiable explanation for how predictive policing makes it worse. “How does predictive policing make it (discrimination, the world) worse” – that’s an NR line that’s devastating if you don’t have an answer. I suggest you look into research on predictive policing and feedback loops as your starting point. The second important part of the discrimination contention is the ability to weaponize it to take out the solvency of the negative contentions. This is impact calculus 101: identifying the ways the impacts in the round interact with each other. One of the coolest phrases I heard during the online debates is that predictive policing is “garbage in, garbage out.” The downside of this phrase is it highlights the uniqueness issues with the discrimination contention, but the upside is it highlights the most powerful argument in the aff’s arsenal: that if the data from predictive policing is flawed because of discrimination and bias, than it can’t be trusted to effectively reduce crime. This is a powerful argument if prepared well, because it takes out all the most common negative contentions. 2. Taking it Global (aff) Smart debaters have pointed out this resolution does not specify the United States. What does predictive policing look like in the rest of the world? Directing attention this way overwhelmingly favors the aff. In fascist and authoritarian states, increasing the efficacy of predictive policing poses a direct danger to the rights, lives, and causes of the people. The most popular example by far has been China. The use of predictive policing to oppress Uighurs in the Xinjiang province is one of the earliest returns you’re going to get for basic google searches of predictive policing. There are so many good articles on predictive policing in China. There’s no doubt that China uses predictive policing to oppress its Muslim population. I’ve seen some debaters get more in depth on this than others. I’m going to unpack this debate to raise the bar for what you need to say to win this argument, on both sides. The most common negative retort is that China’s use of predictive policing doesn’t prove that predictive policing is flawed, it proves that China is flawed: “don’t blame the tool, blame the one who uses the tool” or “anything can be violent in the wrong hands.” I don’t like the phrasing of this argument. What’s my problem with the “tool” phrasing? It obviates every negative argument. Why is “China will do a bad thing with predictive policing” any different than “the US will do a bad thing – increase discrimination – with predictive policing”? Obviously both could choose to not do the bad thing, but if they will do it, than it would certainly not be just to allow that. This is a weaker phrasing of a great argument: that the oppression of the Uighurs or others by fascist states is inevitable. China certainly detained Muslims before predictive policing. Why is the ability to target discrimination uniquely bad when the alternative is untargeted discrimination, if people are still being rounded up? Heck, Stalin didn’t need predictive policing for the Gulags and Germany didn’t need it for Auschwitz. This argument from the negative is pretty good. Let’s talk about setting up the aff China contention to have next-level internal links that predict this negative argument and respond to it before it’s even made. I think there are three pretty good arguments for why the violence caused by predictive policing in China is unique to predictive policing: First: empirics. A quantitative look at detention shows that arrests in Xinjiang sky-rocketed after predictive policing was instituted. Perhaps this is demonstrative that police were not able to just round up anyone they wanted, but that even in China a pre-text was crucial to policing actions. Second: the chilling effect. Some of the best evidence on China’s predictive policing isn’t on its efficacy as a police tactic, but on its effect on the population. Predictive policing creates a true surveillance state, it makes you afraid to do anything that might be perceived as wrong (and often that means it makes you afraid to do anything). For Muslims in Xinjiang, that might mean that predictive policing creates a chilling effect on gathering as a community to worship. Maybe it makes you afraid to visit your family in Egypt or Lebanon. These interruptions to the social lives of religious and ethnic minorities are a unique expansion of the oppressive state due to predictive policing. Third: the black box. Hat tip to one school for already finding this argument: that because predictive policing is so complex, communities have less tools to challenge it. There are less avenues for resistance because the technology is only understood by the state, the algorithms not known to the people they are being used to lock up. It’s worth mentioning that the global predictive policing debate doesn’t begin and end with China. Think about what predictive policing might mean in Saudi Arabia. What might predictive policing mean for burgeoning democratic movements in semi-authoritarian states? What might it mean for resistance groups in truly authoritarian ones? These are good lines to think through the aff contention because it puts a spotlight on the efficacy of predictive policing. The initial, intuitive response to the China contention (“it’s inevitable”) makes sense because it calls into question whether predictive policing is necessary for a state to do evil things, or just policing. But when we’re talking about hunting down and arresting democratic activists in Iran, women’s rights activists in Saudi Arabia, labor movements in China, environmental activists in Brazil, free speech advocates in Egypt… the picture looks different. Because the efficacy of predictive policing truly matters for whether or not the state can restrict those movements. 3. Answering the Discrimination / Bias Contention (neg) I’ve already talked about how uniqueness issues are baked into the aff discrimination contention. Every aff card that says the data is biased explains that’s because the police are already biased now, without predictive policing. By itself, this is a phenomenal uniqueness argument. If predictive policing can do something, anything, to make the world better, and it doesn’t make the world worse because the police are going to be biased either way, than surely it’s more just to do something to make the world better. Here’s two tips for turning this argument from a good uniqueness argument into a slayer: First, identify neg uniqueness arguments in the aff evidence. If the aff says “garbage in, garbage out”, point out how they are literally calling the police garbage in the status quo. If the aff says “policing is infected with racial bias” (a common, and powerful, line in aff evidence), explain how that is a description of the status quo. Second, couple your non-unique with link turns to make them offense. A non-unique says the world is bad now. A link turn says “but we make it better.” If you win the non-unique that policing is biased now, then you have an intuitive argument on your side: there’s nowhere to go but up. The link turn that seems most persuasive to me is about creating a focal point for change, both technological and political. Predictive policing focuses bias into one singular node: the algorithm. There are 38,422 officers in the NYPD. If the aff is right that all or nearly all of them suffer from implicit, subconscious, or explicit racial biases, it is going to be a lot easier to reform 1 algorithm than the implicit biases of 38,422 officers. That’s the technological angle. The political angle is that there is going to be focused blowback on the algorithm if it’s biased. The New York Times is going to run an op-ed once a month explaining how the algorithm is biased. That’s going to create political incentives for police departments to fix the biases, out of their own political self-interest. 4. Complexity is the Key to Neg Contentions (neg) I don’t mean your contentions should be hard to understand, I mean that you should defend niche areas where there are better warrants for why predictive policing is critical to policing’s efficacy. In our last take on predictive policing, we told you we were skeptical about neg ground, and that meant you needed to put more time and thought into being neg. One way to do that is to identify areas of crime that are too complex for regular policing to regularly succeed. Human trafficking that is run through the Dark Web. White collar crime like financial schemes. Cybersecurity. Anything related to terrorism, be it conventional, bioterror, cyberterrorism, etc. All of these areas of law enforcement have unique predictive policing key warrants because of the complexity of the crime. Moving your contentions in these directions also weakens the aff arg that bias undercuts solvency, because these are areas of law enforcement where biases don’t play out the same way. 5. COVID-19 and Predictive Policing (neg) Let’s talk about one particularly complex area of law enforcement: COVID-19. In the United States, a patchwork system of shelter-in-place laws (mostly executive orders by governors) is being adopted. Israel is considering the use of cell phone tracking to evaluate who infected persons could have interacted with. More than anywhere else, China contained COVID-19 through strict lockdowns and law enforcement tactics. Is predictive policing helpful to contain COVID-19? Given the scale and immediacy of this impact, it’s worth looking into. I’ve heard two responses to the negative COVID-19 contention: first, the no link, that predictive policing is not being used to contain the coronavirus. This is a good argument- the neg will need very specific examples about predictive policing in the context of COVID-19 to win. The second argument I’ve heard is that the best negative example of predictive policing being used to contain COVID-19, China, is wrong because China doesn’t really have COVID-19 under control. I want to unpack this one because I firmly believe it’s wrong. China does plenty of messed up things: detaining Muslims (though most of the people who would point this out won’t say word one about the scores of innocent Muslims the US is executing as collateral damage with drones in the Middle East), restricting free speech, undermining labor movements. But I don’t think they’re fudging the numbers on COVID-19. Here’s why: A. Their interventions exactly match what the best science says a state should do. They sent 40,000 doctors to Wuhan. They instituted strict lockdowns that make the shelter-in-place patchwork in the US look like a timeout corner in Kindergarten. They instituted mandatory temperature checks for anyone leaving their apartment. There’s a bunch of graph that compares flattening the curve through three different methods: no action, social distancing, and “Wuhan-style lockdowns.” The Wuhan-style lockdown is basically a flat line. We know it works; we know they did it. We also know that a proper intervention can work: no one doubts South Korea’s numbers. B. They’re decreasing interventions as a result of those success. You don’t need to trust their numbers to trust their actions. Why would they reduce internal quarantines if they didn’t have it under control? At the height of the pandemic in China they restricted the exporting of N95 masks. They’ve undone those restrictions and are exporting masks again. Why would they sell them abroad if they needed them domestically? All these actions are real things we know are happening. It’s not possible to fake relaxing a quarantine in Shenzhen or Shijiazhuang to those of us that have family and friends there. It’s a next-level conspiracy if you want me to believe NSDA China is lying to us. Now, they are increasing restrictions on foreign travel to China, but that actually proves my argument: most new cases of COVID-19 in China are coming from outsiders, not domestic sources. C The ev on “China = lying” is very questionable. It’s usually coming from far-right pundits trying to find someone to blame but Trump. One example of a source for this argument is the Epoch Times. Now the Epoch Times always has a bone to pick with China, and that’s not inherently bad, but their track record at fudging the facts to prove their agenda is. Not to mention they’ve morphed into a pro-Trump, conspiracy-toting paper. Ironically, they also push a strong anti-vaxx agenda. D. It’s not just China. The neg should not forgot to give other examples of countries where predictive policing in law enforcement is being used to contain COVID-19. 6. Args to Avoid My plea is to stay away from the argument that something doesn’t matter because it’s not related to “justice”, or the argument that something isn’t “my job to prove as per the resolution.” Often, I find these arguments are copouts (no pun intended) that would have the effect of obviating any argument the other side makes. The road to hell is paved with good intentions, folks. It’s not just to ignore the consequences of predictive policing if it means people will suffer or die. It’s not just to live in a dream world that ignores questions of how will people actually use this, vs how could they use it under ideal circumstances. Justice includes evaluating all the relevant questions to determine if the world is a better or worse place.
0 Comments
Your comment will be posted after it is approved.
Leave a Reply. |
MissionWyoming Debate Roundup is dedicated to providing quality debate content to Wyoming and Rocky Mountain area high school debaters. We’re a resource for Wyoming debaters by Wyoming debate coaches. Categories
All
Archives
February 2024
|