Posts Tagged ‘Financial Crisis’

Weapons of Math Destruction

May 10, 2018

The title of this book is identical to the title of a book by Dr. Cathy O’Neil. The subtitle is “How Big Data Increases Inequality and Threatens Democracy.” Dr. O’Neil is a mathematician. She left her academic position to work as a quant (a quantitative expert) for D. E. Shaw, a leading hedge fund. Initially she was excited by working in the global academy. But the economy crashing in the autumn of 2008 caused her to reevaluate what she was doing.

She writes, “The crash made it all too clear that mathematics, once my refuge, was not only deeply entailed in the world’s problems, but also fueling many of them. The housing crisis, the collapse of major financial institutions, the rise of unemployment—all had been aided and abetted by mathematicians wielding magic formulas. What’s more, thanks to the extraordinary powers that I love so much, math was able to combine with technology to multiply the chaos and misfortune, adding efficiency and scale to a system that I now recognized as flawed.”

She writes that the crisis should have caused all to take a step back and try to figure out how math had been misused and how a similar catastrophe in the future could be prevented. She writes, “But instead, in the wake of the crisis, new mathematical techniques were hotter than ever and expanding into still more domains. They churned 24/7 through petabytes of information, much of it scraped from social media, or e-commerce websites. And increasingly they focused not on the movements of global financial markets but on human beings, on us. Mathematicians and statisticians were studying our desires, movements, and spending power. They were predicting our trustworthiness and calculating our potential as students, workers, lovers, criminals.”

These math-powered applications were based on choices made by fallible human beings. Although some choices were made with the best intentions, many of the models encoded human prejudice, misunderstanding, and bias into the software systems that increasingly managed our lives. Dr. O’Neil came up with a name for these harmful kinds of models: Weapons of Math Destruction, or WMDs for short.

She notes that statistical systems require feedback—something to tell them when they’re off track. The example she provides is that if amazon.com, through a faulty correlation, started recommended lawn care books to teenage girls, the click would plummet, and the algorithm would be tweaked until it got it right. However, without feedback, a statistical engine can continue spinning out faulty and damaging analysis while never learning from its mistakes. These models end up defining their own reality and use it to justify its results. She writes that this type of model is self-perpetuating highly destructive—and very common.

This book focuses on the damage inflicted by WMDs and the injustice they perpetuate. It discusses harmful examples that affect people at critical life moments: going to college, borrowing money, getting sentenced to prison, or finding and holding a job.

Trump and Behavioral Economics

June 2, 2016

On the June 6 & 13, 2016 “New Yorker” Financial Page there is an article by James Surowiecki.  He is the regular “New Yorker” correspondent for economics, business, and finance.  He has also written a book that Healthymemory would highly recommend, “The Wisdom of Crowds.”  His article is titled “Losers” and it is about how behavioral economics explains the attitude of Trump supporters.  The field of behavioral economics was founded by Daniel Kahneman and Amos Tversky. There have been many, many healthy memory blog posts on this topic and about these authors.   Prospect Theory is key to behavioral economics and resulted in a Nobel Prize being awarded to Kahneman.  Unfortunately Tversky had already passed away when the award was made.

Surowiecki notes that Trump plays to one of the most powerful emotions in economic life, which is what behavioral economics call loss aversion.  The basic idea is that people feel the pain of loses much more than they feel the pleasure of gains.  Empirical studies estimate that, in general, losing is twice as painful as winning is enjoyable. Consequently, people will go to great lengths to avoid losses, and to recover what they’ve lost.

Suroweicki notes that Trump’s emphasis on losing is unusual  even in bleak times.  But he believes that it has worked for him, because it resonates with what many Republican voters already feel.  A study by the Pew Research Center last fall found that 79% of those who lean Republican believe that their side is losing politically.  A RAND survey in January found that voters who believed that “people like me don’t have any say about what the government does” were 86.5% more likely to prefer Trump.  Trump supporters feel that they, and the country, are losing economically, too.  In the RAND survey, Trump did better  with the people who were the most dissatisfied with their economic situation, and exit polls from the Republican primaries show that almost 70% of those who voted for Trump were “very worried” about the state of the economy as compared to only forty-five % of all voters in Democratic primaries.

Surowiki notes some surprising things about all this.  The first is that, in objective terms, plenty of Trump supporters haven’t lost that much.  We’re familiar with Trump’s appeal among white working class voters, many of whom truly have seen wages stagnate and jobs dry up.  But Nate Silver has recently pointed out that the median Trump voter is actually better educated and richer than the average American.  But an important point of Kahneman and Tversky’s work is that people don’t look at their status objectively, they measure it relative to a reference point, and for many Republicans that reference point is a past time when they had more status and more economic security.  Kahneman argues that even people who simply aren’t doing as well as they expected to be doing feel a loss.  And people don’t adapt their expectations to new circumstances.  A study of loss aversion by Jack Levy concluded that, after losses, an individual will “continue” to use the status quo ex ante as her reference point.”  Suroweicki notes that Trump’s promise is precisely that he’s going to return America to the status quo ex ante.  He tells his supporters that he will will help recoup their losses and safeguard what they have.

Suroweicki goes on to say that the other surprising thing is that you might expect loss-averse voters to be leery of taking a risk on an unpredictable outsider like Trump, since loss aversion often makes people cautious:  offered the choice between five hundred dollars and a 50 % chance at a thousand dollars or nothing, most people take the sure thing.  However, loss aversion promotes caution only when people are considering gains; once people have sustained losses, impulses change dramatically.  Offered the choice between losing five hundred dollars and a 50% chance of losing a thousand dollars or nothing, most people prefer to gamble—opposite of what they did when presented with the chance to win a thousand dollars.  People are willing to run huge risks to avert or recover loses.  In the real world , this is why people hold falling stocks, hoping for a rebound rather than cutting their losses, and it’s why they double down after losing a bet.  For Trump’s voters, the Obama years have felt like a disaster.  Taking a flyer on Trump actually starts to feel sensible.

Suroweicki continues, noting that historical parallels are always tendentious, that loss aversion has been instrumental in the success of authoritarian movements around the world.   The political scientist Kurt Weyland has argued that it played a crucial role in the rise of such regimes in Latin American, where the fear of Communism drove putatively democratic societies toward the radical solution of strongman rule.  Suroweicki notes that Trump may not quite be an American Peron, but, to his his supporters, his unpredictability is a selling point rather than a flaw.

It is important to remember that the basis thesis of behavioral economics, a thesis that has ben consistently supported, is that humans do not behave or think rationally.  Rather they are driven by emotions.

Healthy memory feels compelled to note other facets of human cognition that contribute to flawed political decisions.  One is the success of the big lie and the continued persistence of these lies.  It is extremely difficult to correct these lies.

Another problem is  the fallibility of memory and how selective memory makes it difficult to correct erroneous beliefs.  Consider the Iraq war that the younger Bush took us into.  The weapons of mass destruction, on which the invasion was predicated, were never found.  France and Germany were urging Bush to delay an invasion until the inspection were completed and the existence of these weapons could have been ascertained.

It was also the case that the King of Jordan and Henry Kissinger warned Bush that an invasion would result in a broken country that would serve as a base for radical Islamist groups..  This is exactly what has happened.  So the costs of this war not just monetary, which added to the national debt, but more importantly human, produced a situation that is worse, not better, than what prevailed, before the beginning of the war.

People also seem to have forgotten the financial crisis left by the Bush administration that resulted in the very real possibility of a depression.  In spite of recalcitrant Republicans, Obama managed to prevent the depression and aid in an important economic recovery.  By most objective standards, the U.S. economy is in good shape, and the American economy is one of the best performing economies.

Healtymemory still wonders about Trump.  It is difficult for him to imagine Trump curling up with a copy of Kahneman’s “Thinking Fast and Slow.”  It is also difficult imagining Trump taking consul with an expert informing him how to exploit human information processing shortcomings for political gain.  Using the word “instinct” is inappropriate here, but Trump has a flair for exploiting human information processing shortcomings so that System 2 processing is avoided and System 1 prevails resulting in emotions rather than reasoning governing their voting.

© Douglas Griffith and healthymemory.wordpress.com, 2016. Unauthorized use and/or duplication of this material without express and written permission from this blog’s author and/or owner is strictly prohibited. Excerpts and links may be used, provided that full and clear credit is given to Douglas Griffith and healthymemory.wordpress.com with appropriate and specific direction to the original content.