Just wanted to flag an Upshot story that ran in the New York Times the other day, looking into the algorithm that the Chicago Police Department uses to predict who is most likely to be involved in a shooting (whether as the shooter or as the victim.)
As I’ve mentioned before – and Nick Diakopoulos has campaigned about – we ought to be doing more to cover the alogorithms that rule increasing parts of our lives, so it’s great when there’s a piece that does exactly that.
But using the publicly available data that (the CPD) have released, we reverse-engineered the impact of each characteristic on the final risk scores with a linear regression model. Because the department didn’t release all the information that the algorithm uses, our estimates of the significance of each characteristic are only approximate. But using what was available to us, we could predict risk score very accurately, suggesting that we are capturing much of the important information that goes into the algorithm.
It’s a nice piece of work that helps shed light on what’s almost certainly an important policy and policing tool in Chicago. It isn’t clear if the algorithm works well or not – gun violence remains a problem – but just being able to show what factors are taken into account is already an important public service.
In particular, victims of assault and battery or shootings were much more likely to be involved in future shootings. Arrests for domestic violence, weapons or drugs were much less predictive. Gang affiliation, which applied to 16.3 percent of people on the list, had barely any impact on the risk score.
The algorithm has been updated several times, and (Illinois Institute of Technology lead researcher on the project Miles) Wernick noted that the variables of gang affiliation and narcotics arrests were dropped from the most recent version.
There’s nothing wrong in theory, of course, with using an algorithm such as this one to help prioritize the use of limited resources – or even to take human bias out of decision-making, so we shouldn’t be approaching these stories with a bias that algorithmic decision-making is a bad thing.
But while we can talk to humans about why they made certain decisions, it can be both much harder and much easier to do so with machines. Easier because – unless it’s a machine-learning algorithm that’s a black box – there’s generally a codified path of logic we can follow to see how a particular decision was arrived at. And harder because that codified path of logic is rarely disclosed.
Hence the need to better understand how they work, if not exactly, then at least with a sense of what kinds of variables it takes into account.
“If the government is going to outsource decision-making to a computer, the public should be able to examine how the decisions are made and whether it’s fair and effective.”
So far, the CPD hasn’t been tremendously forthcoming with those details, although they did release a list with names redacted that the Upshot used to reverse-engineer the algorithm. As the NYT story notes:
To date, the Chicago Police Department has declined to release details of the algorithm, citing proprietary technology. (Last week, The Chicago Sun-Times and three independent journalists filed a Freedom of Information Act suit against Chicago and its police department to release full information on the algorithm.)
There’s clearly a lot more journalists and media organizations should do to keep an eye on how such algorithms work, what factors go into them, how they’re designed, who designs them, and how they’re regulated and overseen. It’s a huge area that we don’t do anywhere as good a job covering as we should, partly because it’s hard to get real information, but also partly because it doesn’t fall into some of the classic themes of journalism, with reasonably well-delineated notions of right and wrong, or standards we expect to be adhered to.
Because, frankly, we’re still at the start of really understanding what we expect from the algorithms that rule us – what does fair treatment mean, especially in an age of mass and granular personalization? Should everyone be charged the same price for the same product, or is it OK when Staples prices products different based on your location? Does it matter what factors a gun violence algorithm takes into account if it helps bring down deaths – not that it seems to? And so on.
There’s much work that we need to dive into, and the sooner we get into high gear on this, the better.
Russia’s Defense Ministry said on Friday it may have killed ISIS’s leader and self-declared caliph Abu Bakr al Baghdadi in an airstrike in May.
Even if the reports turn out to be true, Baghdadi’s death wouldn’t be the end of ISIS, according to counterterrorism expert Richard Barrett. “[I]t’s a setback rather than a defeat, because [ISIS is] an idea as much as it is a movement,” Barrett said. “The idea is still out there. The reasons the idea took root are still out there.”
In 2014, FRONTLINE correspondent Martin Smith spoke to Barrett as part of the film The Rise of ISIS, which traces how the group seized vast swaths of Iraq and Syria. In this video, Barrett explains how Baghdadi rose to lead the terrorist group.
Although Baghdadi’s self-proclaimed caliphate has been losing ground in its core territories in Iraq and Syria, Barrett said even in the event of the ISIS leader’s death, his supporters might be inspired to fight harder to defend the idea of an Islamic state.
And, as FRONTLINE film, The Secret History of ISIS detailed, ISIS survived a similar transition after its founding father Abu Musab al-Zarqawi was killed in 2006. This excerpt from the film explains how Baghdadi, who spent time in an American prison during the U.S. occupation, ultimately helped to rebuild ISIS from the remains of Al Qaeda in Iraq.
If ISIS follows the same path as its predecessor, it will likely outlast Baghdadi, Barrett said, just possibly in a different form. “Just as when Abu Musab al-Zarqawi was killed and it dealt a blow to Al Qaeda in Iraq, but it came back, there will be an underground organization that survives,” he said. “There will be sufficient structure left to ensure that organization at least has some potential to re-germinate and re-grow.”
The Republican health care bill would cause 13 million Americans to lose coverage and steeply increase Obamacare enrollees’ out-of-pocket costs, a new analysis from the Medicare actuary projects.
The report differs significantly from the Congressional Budget Office analysis, which estimates that 23 million Americans would lose coverage under the American Health Care Act.
Both analyses agree that a world in which the AHCA passes is a world in which millions fewer Americans have health insurance coverage. They differ because they make very different assumptions about how states would react to the bill.
The Medicare actuary report, explained
The Medicare actuary estimates that over the next decade, 13 million Americans would lose coverage under the AHCA. Of these people, 8 million would be Medicaid enrollees and 1.4 million would be people who buy coverage on the individual market. (The Medicare actuary is an independent office within the Center for Medicare and Medicaid Services, meaning its analysis does not reflect the views of the Trump administration.)
The report is clear that people who lose Medicaid would not be likely to end up with coverage elsewhere: “For those Medicaid enrollees who would lose coverage under the AHCA, most are assumed to ultimately be uninsured, though a small fraction would choose to purchase individual insurance.”
The Medicare actuary also expects that 3.3 million fewer people will get insurance at work if the AHCA passes, as the law would create incentives for big companies to drop coverage and move people to the individual market.
Similar to the CBO report, the Medicare actuary estimates that the AHCA would hit older Americans especially hard. Individual market enrollment of people between ages 50 and 59 would fall by more than half.
The amount of money the country spends on health care would decline, as fewer people would have health insurance. The amount Americans spend out of pocket, however, would rise significantly.
The Medicare actuary estimates that households would spend an estimated $21 billion more on health care over the next decade, as more people become uninsured and have to pay health care bills on their own. Some of this would also result from the individual market plans becoming less generous, meaning higher deductibles and copays for those enrollees.
This would amount to a 61 percent increase in overall out-of-pocket costs for Americans who buy their own health coverage.
Part of that would be spent on higher insurance premiums. Although base premiums in the individual market would decline by an estimated 13 percent by 2026, what people pay out of pocket would go up 5 percent because the AHCA provides significantly less financial help to Americans who purchase their own coverage.
The Medicare actuary report makes a surprising assumption
Much of the difference between the CBO analysis and the Medicare actuary report comes down to how they think about Medicaid. CBO estimates that 14 million Americans would lose Medicaid under the AHCA, whereas the Medicare actuary pegs the number at 8 million.
One big part of the AHCA, for example, is changing how Medicaid funding works. Right now, the federal government has an open-ended commitment to pay a certain percentage of Medicaid bills, regardless of how high they get. The AHCA would change that. It would give states a lump sum per person and leave local officials to decide how best to spend that money.
This change — completely separate from ending the Medicaid expansion — would significantly cut state Medicaid budgets. But the Medicare actuary thinks this would not affect enrollment in the program. It seems to assume that state Medicaid programs would function the same, even with less funding.
“There is no estimated impact on Medicaid enrollment because of the presence of per capita allotments,” the report concludes.
Most other observers, including CBO, estimate the Medicaid funding cuts would lead to fewer people having Medicaid coverage. The program couldn’t weather the size of the cuts in the Republican bill simply by becoming more efficient.
“Some of that decline would be among people who are currently eligible for Medicaid benefits” and not just the expansion population, the CBO concluded in its most recent report. CBO expects that, facing this new funding formula, states would “restrict eligibility for enrollment” in the program.
Buried in a footnote in the 9th U.S. Circuit Court of Appeals’ unanimous opinion upholding the bulk of the injunction blocking Donald Trump’s travel ban, there is a moment of reckoning in which the panel addresses whether the president’s tweets constitute binding statements of executive intent.
In making a determination that the second version of the executive order exceeds the statutory authority granted to the president, the panel finds that the order “does not provide a rationale explaining why permitting entry of nationals from the six designated countries under current protocols would be detrimental to the interests of the United States.” The panel then drops a footnote to add the following observation about the president’s actual intentions in enacting the order:
Indeed, the President recently confirmed his assessment that it is the “countries” that are inherently dangerous, rather than the 180 million individual nationals of those countries who are barred from entry under the President’s “travel ban.”
See Donald J. Trump (@realDonaldTrump), Twitter (June 5, 2017, 6:20 PM),
https://twitter.com/realDonaldTrump/status/871899511525961728 (“That’s right, we need a TRAVEL BAN for certain DANGEROUS countries, not some politically correct term that won’t help us protect our people!”) (emphasis in original).
Put aside for a second the legal glory that lies in that “emphasis in original,” a parenthetical that does so much work while doing nothing at all. What’s really vital is that the footnote also does away with the claim that such tweets should be ignored or swept aside, noting a CNN piece that reported “the White House Press Secretary’s confirmation that the President’s tweets are ‘considered official statements by the President of the United States.’ ”
The question of how seriously the courts should be taking the president’s informal and spontaneous tweets is a serious one, and—as was discussed on this week’s Amicus podcast—a question that hasn’t thus far been treated with great rigor. In a new article on presidential speech and the courts, Cardozo Law School’s Kate Shaw notes that the judicial branch shouldn’t take casual presidential comments too seriously. She argues, however, that there is a subset of cases in which presidential speech reflects a clear manifestation of intent to enter the legal arena, among them cases touching on foreign relations or national security and those in which government purpose constitutes an element of a legal test. Based in part on Sean Spicer’s assurance that Trump’s tweets are official statements, the per curiam panel of the 9th Circuit has just ruled that the president’s Twitter commentary clearly falls in the category of speech that belongs in the legal arena.
What does it mean that courts may now begin to take the president’s tweets seriously? Beyond the implications for the travel ban, the notion that Trump’s Twitter feed is its own binding constitutional stream of consciousness invites all sorts of other delightful legal interventions. For one thing, the somewhat charming letter sent last week by the Knight First Amendment Institute at Columbia University now looks like a more serious complaint. The letter, which was sent by the Knight Institute’s Executive Director Jameel Jaffer, argues on behalf of a group of Twitter users that the First Amendment precludes him from blocking people on social media. The letter, which at least implies that it may be followed with a lawsuit, describes the Trump Twitter feed as a designated public forum and a curated series of official statements. Thanks to the 9th Circuit, that characterization now has more teeth.
The finding by the 9th Circuit also gives some force to another fanciful enterprise, the daringly named COVFEFE Act, a piece of legislation introduced Monday by Rep. Mike Quigley, a Democrat from Illinois. The Communications Over Various Feeds Electronically for Engagement Act amends the existing Presidential Records Act to include “social media,” a move that could make it illegal for the president to delete his tweets.
I know what you’re thinking here, so let’s just say it aloud: Why are the federal courts wasting valuable time looking at the president’s Twitter feed when they could be assessing his executive authority? And why are legal journalists writing about it? One might answer those questions with yet another question: Why is the president wasting time he could be spending making the country safer by tweeting threats at the federal courts?
Your move, Mr. President. We’d submit that tweeting “see you in court” isn’t helpful given that the courts have now made plain that they see you, too.