Failed Prophecy and Sunk Costs

October 3, 2014
Posted by Jay Livingston

Four years ago, twenty-three economists (mostly conservative) signed a letter to Ben Bernanke warning that the Fed’s quantitative easing policy – adding billions of dollars to the economy – would be disastrous. It would “debase the currency,” create high inflation, distort financial markets, and do nothing to reduce unemployment.

Four years later, it’s clear that they were wrong (as Paul Krugman never tires of reminding us). Have they changed their beliefs?

Of course not.

Bloomberg (here) asked the letter-signers what they now thought about their prophecy.  Here’s the headline:

Fed Critics Say ’10 Letter Warning Inflation Still Right

This despite the actual low inflation.

(Click for a larger view. The original graph is here.)

I don’t know why I assume that high-level economists would be more likely than some ordinary people to change their ideas to adjust for new facts. Fifty years ago, in The Structure of Scientific Revolutions, Thomas Kuhn showed that even in areas like chemistry and physics, scientists cling to their paradigms even in the face of accumulated anomalous facts. Why should big-shot economists be any different? It also occurs to me that it’s the most eminent in a profession who will be more resistant to change.  After all, it’s the people at the top who have the greatest amount invested in their ideas – publications, reputations, consultantships, and of course ego. Economists call these “sunk costs.”

So how do they maintain their beliefs? 

Most of the 23 declined to comment; a few could not be reached (including Ronald McKinnon, who died the previous day).  Of those who responded, only one, Peter Wallison at the American Enterprise Institute, came close to saying, “My prediction was wrong.”

“All of us, I think, who signed the letter have never seen anything like what’s happened here.”

Most of the others preferred denial:

“The letter was correct as stated.”  (David Malpass. He worked in Treasury under Reagan and Bush I)

“The letter mentioned several things . . and all have happened.” (John Taylor, Stanford)

“I think there’s plenty of inflation -- not at the checkout counter, necessarily, but on Wall Street.” (Jim Grant of “Grant’s Interest Rate Observer.” Kinda makes you wonder how closely he’s been observing interest rates.)

Then there was equivocation. After Thursday night’s debacle – Giants 8, Pirates 0, knocking Pittsburgh out of the playoffs– someone reminded me, “Hey, didn’t you tell me that the Pirates would win the World Series?”
“Yes, but I didn’t say when.”

Some of the letter-signers used this same tactic, and just about as convincingly.

“Note that word ‘risk.’ And note the absence of a date.” (Niall Ferguson, Harvard)

“Inflation could come . . .” (Amity Shlaes, Calvin Coolidge Memorial Foundation)

The 1954 sociology classic When Prophecy Fails describes a group built around a prediction that the world would soon be destroyed and that they, the believers, would be saved by flying saucers from outer space.  When it didn’t happen, they too faced the problem of cognitive dissonance – dissonance between belief and fact.* But because they had been very specific about what would happen and when it would happen, they could not very well use the denial and equivocation favored by the economists. Instead, they first claimed that what had averted the disaster was their own faith. By meeting and planning and believing so strongly in their extraterrestrial rescuers, they had literally saved the world. The economists, by contrast, could not claim that their warnings saved us from inflation, for their warning – their predictions and prescriptions – had been ignored by Fed. So instead they argue that there actually is, or will be, serious inflation.

The other tactic that the millennarian group seized on was to start proselytizing – trying to convert others and to bring new members into the fold.  For the conservative economists, this tactic is practically a given, but it is not necessarily a change. They had already been spreading their faith, as professors and as advisors (to policy makers, political candidates, wealthy investors, et al.). They haven’t necessarily redoubled their efforts, but the evidence has not given them pause. They continue to publish and sell their unreconstructed views to as wide an audience as possible.

That’s the curious thing about cognitive dissonance. The goal is to reduce the dissonance, and it really doesn’t matter how. Of course, you could change your ideas, but letting go of long and deeply held ideas when the facts no longer co-operate is difficult. Apparently it’s easier to change the facts (by denial, equivocation, etc.). Or, equally effective in reducing the dissonance, you can convince others that you are right. That validation is just as effective as a friendly set of facts, especially if it comes from powerful and important people and comes with rewards both social and financial.

------------------------
* This blog has had previous posts on cognitive dissonance:

• You loudly preach the dangers of vaccine. Then it turns out that the one scientific paper supporting your views used faked data. Do you change your beliefs? (here)
• You rail against “gun-free zones” as evil and dangerous. Then your son smuggles a .357 into his officially gun-free university dorm, making his room considerably less gun-free. He uses the gun to commit suicide. Do you change your beliefs? (here)
• You predict that the Republicans will win the presidential elections because they’ve secretly rigged the electronic voting machines in key states. When the Democrat wins those states, do you change your beliefs? (here)

Social Capital and Social Values

October 1, 2014
Posted by Jay Livingston

In the previous post, I offered data showing that among professional footballers, wide receivers and cornerbacks, compared with other positions, were more likely to have been arrested. By coincidence, shortly after posting that I listened to a recent interview on WNYC’s “Death, Sex, Money”  (here) with former cornerback Dominique Foxworth. Googling Foxworth led me to an April 2014 video from the Harvard Business School. Given the representation – two cornerbacks, one wide receiver, and a running back – I figured that the famous HBS case method was now including cases involving sociopaths. 

Here is a brief excerpt. The speaker is cornerback and media-certified thug* Richard Sherman.

So it turns out the athletes are there not as examples of social pathology but for their particular expertise on social capital, though the discussion winds through many other topics. The other panelists along with cornerbacks Sherman and Foxworth are wide receiver Larry Fitzgerald, and running back Arian Foster. They are bright, informative, and frequently funny. If you have an hour, listen to the whole thing.

If you watch from the beginning, you will also hear the moderator’s introduction of the panel members, which is worth noting for what it says about HBS values. In addition to listing some statistics about their athletic records, Prof. Elberse is careful to specify how much money each of them is making.  In the football stadium, what matters most is the final score. At the Harvard Business School it’s income. What’s surprising is not the implicit value itself; it’s that Prof. Elberse makes it so blatant.**

------------------------

* In the wake of his comments after the most recent Superbowl, Sherman was widely labeled a “thug.” The word was used about him at least 600 times on television. For more, go here, or just Google “Richard Sherman thug” and check out any of the 112,000 pages.

** For more on money as the ultimate value, and how this value emerges in ordinary conversation, see this earlier post.

Football Violence - Position or Disposition

September 27, 2014
Posted by Jay Livingston

When sports stories wind up in the headlines and network news, something’s usually very wrong. The news biz, whether print or TV, usually keeps athletes confined in the sports section.  So now we have the network anchors talking about Adrian Peterson leaving welts on the flesh of his son, age four, or showing us the video of Ray Rice coldcocking his fiancee in the elevator. Other NFL domestic violence stories, previously ignored (no superstar players, no video), are now mentioned since they fit the news theme.

These incidents all suggest that maybe football players are just violent people – men with a streak of violence in their dispositions. This personality trait that allows them to flourish on the field, but too often it gets them in trouble after they leave the stadium. 

This is the kind of psychological “kinds of people” explanation that I ask students to avoid or at least question, and to question it with data. Conveniently, we have some data. USA Today has the entire NFL rap sheet (here), and it looks like a long one – more than 700 arrests since 2000.  Nearly 100 arrests for assault, another 85 or so for domestic violence. And those are just the arrests. No doubt many battered wives or girlfriends and many bruised bodies in bars didn’t make it into these statistics. Are football players simply violent people – violent off the field as well as on?

Well, no. The largest category of arrests is drunk driving  – potentially very harmful, but not what most people would call violent.  And besides, NFL players are arrested at a lower rate than are their uncleated counterparts – men in their late twenties.*

(Click to enlarge. The graph comes from Benjamin Morris at FiveThirtyEight. )

This suggests that the violence we see in the stadiums on Sunday is situational (perhaps like the piety and moral rectitude we encounter elsewhere on Sunday).  The violence resides not in the players but in the game.  On every down, players must be willing to use violence against another person. Few off-the-field situations call for violence, so we shouldn’t be surprised that these same men have a relatively low rate of arrest (low relative to other young men).

But let's not discard the personal angle completely. If we look at arrests within the NFL, we see two things that suggest there might be something to this idea that violence, or at least a lack of restraint, might have an individual component as well.  First, although NFL arrests are lower for all crimes, they are much, much lower for non-violent offenses like theft. But for domestic violence, the rate is closer that of non-footballers.  The NFL rate for domestic violence is still substantially lower than the national average – 55 NFL arrests for every 100 among non-NFL men. But for theft, the ratio is one-tenth of that – 5.5 NFL arrests per 100 non-NFL. Also on the higher side are other offenses against a person (murder, sex offenses) and offenses that might indicate a careless attitude toward danger – DUI, guns. 

Second, some positions have a disproportionate number of offenders. The graphs below show the percent of all arrests accounted for by each position and also the percent the position represents of the total NFL roster.  For example, cornerbacks make up about 10% of all players, but they accounted for about 14% of all arrests. (The difference is not huge, but it’s something; there would be a very slight overlap in the error bars if my version of Excel made it easy to include them.)

The positions disproportionately likely to be arrested are wide receivers and defensive tackles. Those most under-represented in arrests are the offensive linemen. 

This fits with my own image of these positions. The wide-outs seem to have more than their share of free-spirits – players who care little for convention or rules. Some are just oddball amusing, like Chad Ochocinco formerly of the Bengals.** Others are trouble and get traded from team to team despite their abilities, like Terrell Owens of the 49ers, Eagles, Cowboys, Bills, and Bengals.

As for the linemen, the arrest differential down in the trenches also might be expected.  Back in the 1970s, a psychiatrist hired by the San Diego Chargers noted this difference on his first visit to the locker room. It wasn’t the players - the offensive and defensive lineman themselves looked about the same (huge, strong guys) – it was their lockers. They were a metaphor for on-the-field play.  Defensive linemen charge, push, pull, slap – whatever they can do to knock over opponents, especially the one holding the ball. Their lockers were messy, clothes and equipment thrown about carelessly. Offensive lineman, by contrast, are more restricted. Even on a run play, their movements are carefully co-ordinated, almost choreographed. Watch a slo-mo of the offensive line on a sweep, and you’ll see legs moving in chorus-line unison.  Correspondingly, their lockers were models of organization and restraint.

Maybe these same personal qualities prevail off the field as well. Those offensive lineman get arrested at a rate only half of what we would expect from their numbers in the NFL population. Arrests of defensive linemen and wide receivers are 50% more likely than their proportion on the rosters. Position can’t be the entire explanation of course. Running counter to this “kinds of people” approach are the other hard-hitting defensive players – defensive ends and linebackers. According to the principle of violent people in violent positions, they should be over-represented in arrest figures just like the defensive tackles and cornerbacks. But they are not.

If this were a real article, a journal article, this final paragraph would be where the author calls for more data. But the trend in NFL arrests has been downward, and if fewer arrests means less data but also less domestic violence, that’s fine with me.


-----------------------
* Some critics have questioned this comparison since it does not take income into account. Of course, arrest is a fairly rare event, and it would be kind of hard to find a large enough sample to allow for estimates among men 25-29 with incomes over $500,000.  

** Changing his name to match his uniform number is one example. For another, once during an official video review of his catch of a pass to determine if he was in bounds, Ochocinco borrowed a dollar bill from an assistant on the sideline, went up to one of the refs on the field, and offered the dollar as a “bribe” to rule in his favor.  Everyone who saw the gag found it funny – everyone except the NFL brass, who fined him $20,000.

That Isn’t Funny

September 24, 2014
Posted by Jay Livingston

Most of the time, when people talk about humor – TV  sitcoms, movie romcoms, comedians, jokes, etc. – they’ll say things like, “That show is so funny,” or conversely, “That show is definitely not funny.” 

They assume that the funniness resides completely in the joke or show or comedian and that they themselves are objective observers.  But as any comedian knows, the funniness depends on the interaction between the joke and the audience. If everyone in the room is laughing, it doesn’t make much sense for you to say that the joke wasn’t funny. 

The funniness depends not just on the joke but on the ideas, assumptions, values, and knowledge that we bring to it. Some of that background knowledge is knowledge of other jokes.  Here’s a cartoon from the current New Yorker.

It stands on its own, I guess, but it’s funnier if you know the joke it’s referring to, which goes something like this.*

A grandmother (she doesn’t have to be Jewish, but she probably is, and she certainly was in the version that I first heard, and besides, it’s Rosh Hashanah, so we’ll say she is). A Jewish grandmother is standing at the edge of the ocean pointing out at the crashing waves and screaming for help. “My grandson, my grandson.” A lifeguard hears her, runs into the surf, swims out through the rough water, dives under, comes up with the boy, carries him back to shore, performs every kind of artificial respiration until finally the kid coughs and sputters and comes back to consciousness. The lifeguard, exhausted looks up at the grandmother. She looks down at him and says accusingly, “He had a hat.”

OK, maybe it’s not so funny on the page. If you heard me tell it in person . . . or maybe not even then.  Some guys know how to tell ’em (and that’s a punch line to another joke).  Anyway, the New Yorker cartoon is a meta-joke, a joke about a joke. But the other cartoons too, I realized as I paged through the magazine, require background knowledge. If someone from a distant culture, or a member of our own society who has not acquired that cultural knowledge (i.e., a child), looked at any of those cartoons, we would have to fill in that missing background. Without it, the joke would not be funny. Of course, then we’d be explaining the joke, and it wouldn’t be funny anyway. You can’t win.

But the larger point is that despite our sense that the funniness is in the joke or that the “wonderfulnesss” of a poem** lies completely within the joke or poem, we would be more accurate if we said, “That joke is not funny to me and to people like me.”


----------------------
* Like many other teachers, I’m often disappointed and frustrated by students’ lack of cultural and historical knowledge. On the other hand, when I say, “It’s like the old joke . . .” I realize that most of them don’t know that joke. And if I tell it right, I get a laugh.

** Andrew Gelman and his commenters had a discussion about this recently (his blog post is here).