There is an Echo Chamber, but It is Neither Liberal Nor Conservative

The talk of alleged “liberal” echo chamber that stifles speech on campuses has been going for some time, with New York Times’ Nicholas Kristof recently joining the fray.  Aaron Hanlon, an English professor at Colby has just chimed in on The New Republic in attempt to counter these accusations, but I think he just wound up sounding whiney and petulant rather than successfully counter the accusations.

I think the problem is twofold.  First, I think there has to be an honest recognition that there IS an echo chamber in academia, or indeed, any semi-insulated group of people.  BUT, equally, it needs to be noted that this echo chamber effect is not necessarily “political” in the usual sense.

While we use the term derogatively and flippantly, echo chambers are hardly abnormal or unusual.  We try to understand the universe by processing the information we get in terms that (we think) make sense to us.  As George Lakoff might say, we understand things by drawing metaphors, or even similes.  Thus, if we understand what a wall is, then an elephant is like a wall.

This, of course, is not exactly wrong, from a certain perspective, at least, but hardly “right,” or, at least, complete, either.  An elephant IS, in some context, like a wall indeed, but there’s so much information lost through this analogy. The problem is how to explain what an elephant is to a group of people whose understanding is limited to that of the wall.  (Note the wording:. “exolaining” is harder than merely “describing,” although simply describing would be hard enough.)

If trying to explain an elephant with just one set of audience with a certain perspective is difficult, try doing it simultaneously with two different audiences:. one who understand only the wall and the other who understand only the fan.  In a limited sense, an elephant IS simultaneously like a wall and a fan, although the neither analogy would make sense to both audiences at the same time–and saying that an elephant is like both wall and fan is still woefully and nonsensically incomplete.  If, moreover, there are bad feelings between the wall people and the fan people, things can get ugly, without having enlightened anyone.

The problem is not that academics are necessarily trapped in a bubble–although many are, no doubt.  It is far more that the audiences, especially the students are trapped in their own bubbles and feel that it is their right to keep them–and are offended when their bubbles are attacked, or so they think.

From personal experiences, I know well the whiny procluvities of politically active college students, who believe that the elephant should be like the wall, or the fan, or whatever.   They are not so much wrong, as much as lacking in perspective.  BUT faculty challenging their worldviews is strictly discouraged, especially if you are low on the totem pole–the admins have had their share of issues with unhappy “customers.” So an elephant is like a wall, and that’s that, if there are enough students who want to hear that version of the story.

This is a far more serious problem in social sciences, because, here, facts are squishy.  In physics, if Newton’s physics implies light bends so much and it does not, then Newton is wrong, at least as far as how gravity affects light is concerned. Most social science questions don’t have such obvious factual answers.  Indeed, way too many questions are normative ones that have NO factual answers.  In case of Newton and gravity, there is no disagreeing with the experiment.  If you say people should vote and I say they shouldn’t, then there is no reason for an agreement:. all that we do by disagreeing is indicating that we hold different values, and if the values we hold are sufficiently impirtant, disagreements are grounds for distrust: after all, how can you take seriously someone who thinks people should not vote?

It is easy for the normative mindset to slip into even the questions of facts. Whether one thinks people should or should not vote, it is a point of fact that mant people do not, for all sorts of reasons.  It is all too common, once this point is made, that students react by saying that is terrible and that they should be made to vote somehow.  When the point of the lesson is how turnout fluctuates from election to election, for different demographics, with political consequences, students’ reaction is completely missing the point, but attempting to explain why the nonvoters are justified in their abstentions requires defending the morality of their action, thereby offending students’ moral sense of how the world should be.

This is where we come up hard against that bubble again: I don’t disagree with how you think the world SHOULD be, but the world does not work that way because many people don’t share that view for reasons that are important to them.  This is a matter of fact that should not cause problems, but this invariably does.

An excellent example of this is creationism.  Creationism as “science,” may be wrong, but belief in it is not–belief is an opinion, and as such, it is neither right nor wrong.  The existence of many people who believe in some form of creationism is a matter of indisputable fact, on the other hand.  Yet, there are many perfectly reasonable people who insistently deny this fact because it offends their bubbly morality, and they pride themselves on how grounded they are on “reality”–apparently the reality they believe in, not the one they got.

It is not just one bubble that ails academia, or indeed, society at large.  Too many people are not only trapped in their bubbles, but are actively using the power and influence they have to actively maintain them, and that you have “facts” on your side does not mean you are not trapped in an echo chamber–see the creationism example above.  In the academic universe, as well as in journalism, policymaking, and any number of other fields, not only are we trapped in our own bubbles, we must take care not to pop the bubbles that people who are more influential than we want to see preserved.  What is more, unlike Einstein and Newton, we will not be vindicated by a single experiment, for social facts are noisy and often inconclusive (to be fair, when Harold Urey suggested in 1970s a comet might have done in dinosaurs, he was met by a deafening silence, except perhaps some snickers about a chemist not knowing about paleontology (bubble).  If he weren’t a famous Nobel prize winner, his career could have been ruined for such “nonsense.”. Of the Alvareses a decade later, it was Luis, the Nobel-physicist, not his paleontologist son Walter, who thought up comet first, or so I heard.  But facts in paleontology are also murky.)

All the sacred cows that cannot be touched create profit centers for those who peddle comfortable myths that cannot be easily challenged, and we wind up with a cult, of a particularly cargoish variety.  Sometimes, the elephant just isn’t all that much like a wall, and all the nonsense we concocted on the premise that the elephant is “just like” the wall need to be chucked aside if the “facts” don’t seem right–even if the “facts” come in form of creationists.

PS.  The point I am raising is that the bubble is mutual–all sides are wrapped in their own bubbles, and they do not want their bubbles shattered.  Sometimes, worse still, those in a bubble do not realize that they are–even as they are quick to realize the bubble the others are in.  Thus my point about the evolutionists vs. creationists:  the former cannot conceive reasonable reasons to believe in creationism, rational reasons to vote for Trump, or any other seemingly “unreasonable” things.  All analogies that they can conceive of suggest that it is they who are the reasonable ones while the rest of the universe is not.  Thus, those who do not act and think as they do cannot be rational–a variation of the “true Scotsman fallacy” if you will:  no reasonable person cannot possibly disagree with them, so those who disagree must be unreasonable people.  Of course, the alternative is to reckon that your notion of what is reasonable might be too limited, that others may have different sense of what is “reasonable” and are acting accordingly that you need to understand.  If you will, you need to learn other analogies, except, unlike the usual analogies, you need to learn what the others know in their usual life so that you can understand how they use their analogies.  To return to the elephant example, you do not know what a fan is.  The other guy insists that an elephant is like a fan, even though you “know” that an elephant is like a wall.  Therefore, you need to understand first what a fan is and then why the other guy thinks an elephant is like a fan.  Only at that point can you start wondering how that analogy fits with the analogy that you were familiar with to begin with:  that elephant is like a wall.  This is not an easy step, especially since, to the wall people, the analogy between the elephant and the wall is both natural and obvious, while that between the elephant and the fan, whatever this fan is, is not.  But that is exactly the point:  you do not “learn” what is obvious and natural.  If you think creationism, Bernie Bros, or Trump voters are strange, but they actually exist in spite of your believe that they shouldn’t, that’s all the more reason that you should look at them closely and seek to learn.

 

 

 

 

Advertisements

Deaton on 2016

Finantial Times has an excellent interview with Angus Deaton that provokes a lot of thoughts.  While full commentary will have to wait, a few quick reactions.

  1.  Deaton is fundamentally right that there is a lot of damage done by globalization, but that on the whole globalization has done on enormous amount of good as well.  This is, indeed, the fundamental dilemma–how to minimize the number of babies being thrown away with bathwater, while taking, if possible, an appropriate amount of baths.
  2. Further, point #1 requires a subtler thinking about politics and economics.  There is something fundamentally wrong with the analogy of “centrism” for the politico-economic status quo.  Spatial models, at least of Euclidean variety, implicitly assume zero sum games:  if you move to one side, you move away from the other direction. In theory, centrism should minimize the losses to all, but this is obviously not what is taking place, since #1 implies that there is a massive inequality in gains and losses among the politico-economic actors, which, by definition, cannot be “centrist.”  You could keep spatial metaphors and deal with this in theoretical terms, if you introduce more sophisticated mathematics…but Hilbert spaces drive even former math majors batty, let alone people with little or no math background pretending to be math savvy.  We need something creative, simple, and decidedly not spatial.
  3. I love the way Deaton responds to some of the questions as “simple minded questions.”  I suppose that’s more diplomatic than my more usual way of addressing them:  all answers are right, but some, even many, questions are wrong.  His answer, that the problem is plutocracy (as he puts it, “inequality that comes through rent-seeking”) cuts to the heart of the matter.
  4. This is a huge quote worth keeping in mind always:  “And I’ve been there where your life is really, really shadowed by not knowing where the money is going to come from . . . It’s a misery.”

Was Obama Bad for the Democrats?

The title of this post is also the title of the op-ed piece on New York Times by the Greenbergs today.   It is not clear what exactly their answer to this question is, ironically, enough.  Their tentativeness is captured by the weak concluding sentence of the piece:

We think voters were sending a clear message: They want more than a recovery. They want an economy and government that works for them, and that task is unfinished.

Personally, I think the problem with the Democrats is that they were bad for themselves. The political skills of Obama and his operatives and the resulting personal popularity that he enjoyed obscured how badly out of touch the Democrats have been.

With the triumph of Donald Trump, it became fashionable for the Thomas Franks of the world to claim “I told you so.”  Yet, it is not in Kansas, literally or proverbially, that the electoral tide swung against the Democrats.  Trump’s margin of victory in Kansas, and indeed, much of hte so-called Red States, was smaller than Romney’s (Romney captured nearly 60% of the votes in Kansas, while Trump only captured 56%, obscured by a large number of votes that went third party–more than 7%).  The pattern is repeated in states like Georgia, Texas, or Utah.  Trump gained, or at least, Democrats lost in the North.  While the Democrats’ loss of votes in the Midwest received much attention, their loss of votes in the Northeastern states like Connecticut, New Jersey, and New York, EVEN IF they won these states by large margins, deserves closer scrutiny.

I don’t think this is a new pattern, but a continuation of what has already been seen.  Andrew Gelman has already documented that, in the North, especially in the states dominated by the Democrats, income is not a good predictor of a voter’s partisanship.  The wealthy tend to be, relatively speaking, more Democratic, while the poorer tend to be more Republican, again relatively speaking.  While, in no state, the wealthy are more Democratic than the poor, the idea of “wealthy, out of touch liberal” applies to the blue states, not Kansas, proverbial or literal.  In what sense are the wealthy in the liberal states “out of touch” with the poorer (the choice of the word is deliberate–it is not the very poor, but those who are in the lower rungs of the middle class, those who could use some help but not poor enough to qualify for a lot of assistance that are most discontented)?  My hunch is that the extent to which they are “out of touch” cannot be easily quantified or specified in precise policy terms.  Rather, it is “cultural” in the sense that the wealthy elites, as per my previous post, ooh and ahh over the latest sensations that are utterly irrelevant to the poorer.  (Hamilton might be another example.  Here, I’m torn, since, even though I never did actually watch Hamilton–I don’t live in New York, after all–I actually love the idea.  But I don’t run the country and I actually am convinced that people like me should not–thus my hostility to wonkism.)  In a sense, it’s a class war:  the poor are against the rich who have political power in their hands, and in the blue states, the rich with all the political power tend to be Democrats.

The problem is that the Democrats in the blue states do not actively care to seek support from the poorer.  “Cultural liberalism,” broadly defined, is enough to win over the support from the many wealthier liberals found in these states anyways and their support is usually good enough to dominate the politics.  The rich do find it easier to rule, like it or not, as wealth provides many tools useful for an effective political machinery–much the same reason Republicans will be difficult to eject from positions of power in the proverbial Kansas.  It may not necessarily the case that the Democrats are actively against the poorer per se, but are engaged in a form of malign neglect.  What the Democrats offer do not fit the needs of the poorer, even if they might extol themselves on how good they are for the poor(er)–again, the choice of the words here might be relevant.  The Democrats may indeed be good for the poor, but are they good for the “poorer”?  Marie Antoinette might have been sincere in offering her cake to the poor, but the poor don’t want her cake.  They need help getting by the concerns that matter to them and they are not getting it from the cake eaters.

I think this is where Trump is significant.  For all his money, Trump is, at least, electorally speaking, a representative of the Northeastern Republicans–the lower middle class who might be found in, say, Staten Island.  All other Republican candidates represented the more typical Republicans of the proverbial Kansas.  He was, whether deservingly or not, credible to the poorer Northeastern Republicans in a manner than no other Republican could have been–thus the geography of his electoral support.  The Northeastern Republican is not necessarily a bigot, a small government fanatic, or a religious fundamentalist–even though he may have shades of all three, I doubt any of them is a make or break it issue like for more conventional Republicans.  He is primarily interested in a government that works for them, and, I suspect, not even especially a “Republican.” This is the group for whom “get government out of my Medicare” actually makes an intuitive sense–they like government programs that work, but they are not trusting of the people who run government to do the right thing when they mess with them, and to the degree that they have dealt with their state governments dominated by the Democrats–since most programs are, after all, administered by states–Democrats are usually the target of their distrust.

This is also a group that received relatively little attention in politics so far, because of their lack of impact on national or even state politics.  They are the out group in states that consistently elect Democrats.  Most of the Republicans seen on the national politics come from the states where Republicans dominate–the proverbial Kansas–and they don’t exactly command trust of these voters, for they are as out of touch with them as the liberal Democrats.  Lack of familiarity means that we do not know where to place them on our scale, which, after all, is built on, literally, the records by the politicians we have, not the hypothetical politicians we don’t have.  The point raised by the Greenbergs at the end of their op ed is the obvious one and, also, the right one.  But it is also far harder to accomplish than not.  Trump won by heresthetics, by recognizing the contradictions of the semi-fictitious liberal-conservative dichotomy in U.S. national politics.  No one will defeat him by going more “conservative” or “liberal” (except possibly, the Republicans in Congress, by forcing him to give up his heresthetics for conventional conservatism.)  The Democrats, if they want to win, will have to start thinking different thoughts, other than liberalism and conservatism.

Why Are We Surprised When Nationalists Turn Out to be Nationalists?

This article in the New Republic provides another example of how deluded Westerners who think “democracy” equals Western values are.

Aung San Suu Kyi is, whether one likes it or not, a Myanmarese/Burmese nationalist above and beyond all else.  She doubtlessly shares the proclivities, both positive and negative, of her people, and this includes, as shown repeatedly throughout history, a certain streak of ultranationalism and xenophobia (i.e. how many Indians are left in post-independence Myanmar/Burma where there were more than a million of them?  Answer is not many, because they were driven out by the locals who considered them, not without reason, as associates of British Imperialism.)  That she is showing that, she is, after all a Myanmarese/Burmese should surprise the Westerners is the real shocker, not that she is acting like one.

What is more is that Aung San’s own changing political status reinforces her Myanmarese/Burmese-ness.  In the past, she was under house arrest, without any formal power or influence other than goodwill of the Westerners.  Naturally, she was willing to say the things that flattered the Westerners and appeased their sensibilities.  That is no longer the case now that she is part of the government, indeed, an integral part of the government.  Her opponents have, not entirely without justification, that she is not a real Myanmarese/Burmese, but a pawn of the Western imperialist powers–which, in a sense, is not entirely untrue.  She has every reason to dispel such perceptions in order to secure and expand her influence in present day politics of Myanmar/Burma.  Parrotting Western do-gooders is the last thing she’d want to do.

This is, like it or not, the new face of democracy, the Fourth Wave, one might call it.  The leaders of this wave increasingly represent majorities/pluralities among the locals, the actual denizens of the countries that they seek to govern, and these majorities/pluralities are increasingly insular and inward looking in their confrontation with an increasingly unfamiliar and foreign landscape of the internationalized world.  These local faces belong to far right nationalists in Hungary and Lithuania, Euro-skeptics in UK, the Ikhwan in Egypt and Syria, the ultra-nationalists on both sides in Ukraine, FN in France, and Donald Trump in United States–in other words, the West is not exactly immune to this wave of “democracy.” These people are not friends of the cosmopolitan elites, not in Egypt, not in Ukraine, not in UK or US of A, or, indeed, in Burma either.  As I had written about illiberal democrats and liberal authoritarians before, in fact, several times.

I don’t think this means that we need to choose between “liberalism” (in the old fashioned sense) and “democracy.”  I do think that it does mean that we need to choose between liberalism ™ and “democracy.”  A lot of what passes as liberalism today is indeed foreign, alien, and most importantly, utterly irrelevant to the lives of the average person.  As I saw on Twitter earlier today (which I unfortunately cannot find again), for many “liberalism” has come to mean oohing and ahing over “bespoke insects prepared by a multicultural Nigerian-Swedish chef  touted by New York Times.”  What does this have to do with the regular people, whether in Kentucky or Yangon?  Aung San and Trump are both clever people, with an eye for spotting what it is that their respective people–the real people who actually make up the majorities or large pluralities of their countries’ populations–and are acting accordingly.  If anything, given that most of the cosmopolitans reside in New York or London and not Yangon, Cairo, or even Kiev or Donetsk, one might expect the leaders of the Fourth Wave in the developing world should be far more illiberal than their Western counterparts.

Perhaps one should add a third act to Marx’s 18th of Brumaire:  what was once a tragedy turns to farce, then becomes an even bigger tragedy.  Bad things are unfolding across the world.

 

Follies of Asking “What Do You Want” and Measuring Them by Wrong Yardsticks.

I think this article by Guy Molyneux published in The American Prospect is one of more thoughtful and thought-provoking pieces on the topic.

The first instinct that many people have in evaluating almost any (seemingly) new political movement is to place it in context of the divisions in the conventional politics. This is erroneous, as the article notes, because the very reason that these insurgent electorates emerge is because they simply don’t buy into conventional politics in the first place.  The division that Molyneux suggests is instructive:  perhaps over half the white working class has now become fairly reliable partisan Republicans, while a very small fraction remains reliable Democrats, but more than a third–a huge number–lie outside the conventional political divide where the usual markers of liberalism and conservatism simply don’t apply to them.

I don’t like the term “moderate” white working class (MWWC), as this implies a spatial ideological frame, that they hold views that straddle the geometric middle between liberalism and conservatism.  Perhaps a better term might be apartisan or non-ideological (although the latter still suffers from the insinuation that they lack an ideology, not that their ideology does not correspond neatly to the conventional divisions.)  A useful illustration of this is their attitude towards government vs. government functions.  This “moderate” white working class is deeply distrustful of the former, but not of the latter.  If anything, they are in favor of much more activist government that does more things on behalf of “the people” than it does now, but they are not trusting that the actual government, composed of the politicians who presently run it, would actually set up the details to ensure that government services actually serve “the people,” rather than narrow subsets of politically useful allies of politicians.

Hints of this have been found since the early days of the so-called Tea Party movement, but these have been dismissed as absurd and insignificant, e.g. “get government out of my Medicare.”  These MWWC voters love Medicare and government services generally, and, especially important, these are services that are set in stone, beyond the “politics.” Whether Republicans like it or not, many aspects of Obamacare have entered the realm of Medicare–government services that people like and expect to continue as rights set in stone.  What they do not like and trust is the “politics” involved in the implementation and the politicians involved, who have done nothing to earn their trust, of whose motives they have no understanding, and whose actions are shrouded in mystery.

It is tempting to insist that greater transparency would somehow earn greater trust.  It is unlikely or necessary:  people do not understand Social Security or Medicare–they are complicated programs run by enormous bureaucracies.  Nobody really trusts them on the basis of understanding how exactly they work, not that it is realistic to expect millions of voters to understand anything more than the rudimentary basics of them anyways. The proof of these services, however, is that they work, which they do quite efficiently.  In the beginning, moreover, they were sold by politicians who have spent decades working their districts in one form or another and earned the trust at the personal level of the voters who elected them.  Their salesmanship depended on not politicians trying to explain the nitty gritty of the programs, but emphasizing that they are trustworthy people who sincerely have the best interests of the voters at heart, proven by decades of service and neighborliness.

Matthew Stoller’s essay in the Atlantic  is applicable here, although, I suspect, more at the level of tactics and appearance than the substance–of “populism,” that is, not the article’s content.  “Populism” does not rest on the specifics of the policy programs, but on the credibility of the policymakers and the salesmen–the politicians.  Supporters of populism do not do so because of the policy specifics, but because they find the populist more trustworthy than the alternatives.  Wright Patman could do what he did because he earned the trust of his voters, not because of particular expertise.  But this comes with a huge caveat:  wonks are not really all that different from the populists, in the sense that they sell trust, faith, and confidence (and I mean this in the best possible fashion).  I’ve often commented on how economism, wonkism, and scientism resemble a cargo cult more than actual science, but the flip side is that they generate faith in a manner that an actual science cannot and must not.  Many people today believe in facts and figures and the experts who generate them trust them to work, not unlike the faith of many people in Social Security and the populist politicians who brought it about in the New Deal Era.  If populism of yesterday was a genuine revivalist religious movement acting as a political movement, so to speak, wonkism today is more of a cargo cult in the sense Feynman used it–a religious movement that wraps itself in the trappings of reason and “science.” Economism and scientism fit this description well, regardless of one’s moral views on them:  if religion were the opiate of the old masses and the populism a political variant thereof, scientism is the antidepressant of the new elites and wonkism the political outgrowth.

Both rest fundamentally on “faith” in the institutions and actors associated with them, rather than actual facts and reason, and as such, their clash necessarily carries with it a sort of religious undertone.  This is not a helpful situation:  one can argue endlessly about whether Hillary Clinton has had a set of populist political agenda or whether Donald Trump is a real populist.  The truth, however, is that, since populism is a faith-based thing, it remains that Trump did inspire actual faith of many who believed (not too many perhaps, but enough to win him the election) while Clinton did not.  That the two are not completely mutually exclusive should be obvious given the experience of Barack Obama, who simultaneously won the trust of both the wonks and populists, at least enough of them to secure the huge majority of 2008 and the respectable majority of 2012.  But this is the aspect that has only received attention lately, especially in the aftermath of the Trump victory.  Every wonk, it seems, saw in Obama one of their own, and assumed that he had no populist in him, even if he ran on a vacuous slogan that effectively asked only “faith” out of the voters–“hope and change.”  2016, consequently, saw no attempt at blending the wonkism and populism:  during the GOP primaries, populist Trump mopped the floor with the Republican wonks, while the populist Sanders nearly upset wonkish Clinton, with the same religious war spilling over into the general election.  Without commenting on the morality, I think this is a mistake.  Religious wars can never be “won,” except by exterminating unbelievers, after a long, hard, and ugly struggle (if you believe otherwise, tell me where the nearest Temple to Mithras is.)  What made FDR, Obama, and perhaps even Nixon effective leaders was that they could simultaneously gain trust of both populists and wonks.  This we need more of, not the political equivalent of fire and brimstone preachers on both sides calling holy wars on the infidels.

Econ 101 and Economism: Advocacy for the Devil.

James Kwak at the Baselinescenario has been writing about economism often lately.   I suppose he does have a new book coming out so that he has a pecuniary incentive to draw attention (To invoke economic theory!  Ha).

I have been posting a lot about evils of economism, wonkism, and scientism.  I am hardly a supporter of the tendency towards economism, but I find the currently fashionable attacks on economism troubling.  A bit of background:  although I wound up in political science academia via history and economics, I had begun in physics and math (I wound up in social sciences because I wanted to study history like a mathematical physicist–which, in retrospect, seems to have been a very bad idea), and I had experience teaching physics to high school students at a fairly advanced level, of which I am quite proud.

My modus operandus when teaching physics was always geared to show the limits and pitfalls of theories that were “abstractly” right:  e.g. the laws of motion.  The typical demonstration would involve calculating some properties of moving objects on different surfaces using the textbook formulas followed by experiments that show how they pan out.  For example, do objects with equal mass that are subject to the same force accelerate at the same rate?  The short answer, of course, will be no:  smoother surfaces lead to greater acceleration, but, at no time, does the observed acceleration match that predicted by the formulas.  The consequence of this was twofold:  1) the formulas are of limited value when in real life because there are other forces at work not accounted for in the “theories”; 2) these other forces vary depending on the smoothness of the surfaces, among other things.  Of course, these “other forces at work” reside in the residuals from the prediction from the theory:  in other words, you use the theory to establish a baseline, and you theorize about these other forces based on the patterns of deviations from the theory’s prediction.  In other words, the theory is valuable because it is quite wrong, even if “fundamentally right.”

The take-home point from these exercises is not so much that the laws of motion are wrong because they cannot describe the actual observations perfectly.  Indeed, you SHOULD accept the laws of motion as if they are true, as the baseline from which you begin your inquiry.  The real point is that the actual motions are more complicated and there are other potential discoveries to be made because the existing theories are incomplete, as reflected in the mismatch between the theory and the data.  The consequence of this is that, by going through these demonstrations, the students learned something about friction.

Feynman, of course, made this general point often, as exemplar of what he considered good science.  Consider the following passage from his famous address about “Cargo Cult science.”

For example, there have been many experiments running rats through all kinds of mazes, and so on—with little clear result.  But in 1937 a man named Young did a very interesting one.  He had a long corridor with doors all along one side where the rats came in, and doors along the other side where the food was.  He wanted to see if he could train the rats to go in at the third door down from wherever he started them off.  No.  The rats went immediately to the door where the food had been the time before.

The question was, how did the rats know, because the corridor was so beautifully built and so uniform, that this was the same door as before?  Obviously there was something about the door that was different from the other doors.  So he painted the doors very carefully, arranging the textures on the faces of the doors exactly the same.  Still the rats could tell.  Then he thought maybe the rats were smelling the food, so he used chemicals to change the smell after each run.  Still the rats could tell.  Then he realized the rats might be able to tell by seeing the lights and the arrangement in the laboratory like any commonsense person.  So he covered the corridor, and, still the rats could tell.

He finally found that they could tell by the way the floor sounded when they ran over it.  And he could only fix that by putting his corridor in sand.  So he covered one after another of all possible clues and finally was able to fool the rats so that they had to learn to go in the third door.  If he relaxed any of his conditions, the rats could tell.

Now, from a scientific standpoint, that is an A‑Number‑l experiment. That is the experiment that makes rat‑running experiments sensible, because it uncovers the clues that the rat is really using—not what you think it’s using.  And that is the experiment that tells exactly what conditions you have to use in order to be careful and control everything in an experiment with rat‑running.

I looked into the subsequent history of this research.  The subsequent experiment, and the one after that, never referred to Mr. Young.  They never used any of his criteria of putting the corridor on sand, or being very careful.  They just went right on running rats in the same old way, and paid no attention to the great discoveries of Mr. Young, and his papers are not referred to, because he didn’t discover anything about the rats.  In fact, he discovered all the things you have to do to discover something about rats.  But not paying attention to experiments like that is a characteristic of Cargo Cult Science.

The part about how Mr. Young did not discover anything about the rats always stuck at me.  Of course, Mr. Young discovered things about rats, just that he discovered nothing “fundamental” about rats.  Something that I hadn’t realized, until years later, was that I was guilty of the same sin as Mr. Young, so to speak, in my friction and motion demonstrations:  of course, they showed rougher surfaces lead to lower acceleration, like funny noises leading to rats finding their way around the maze.  They do say something about actual motion and rats–the data says so.  But I did not say anything fundamental about motion:  a = f/m is still the basic law of motion.  I simply made a slight modification:  a = f/m – f(smoothness), i.e. just an “error” term that happens to be biased (i.e with a mean not equal to zero.)  Mr. Young did not say something about how rats find their way in the maze, but, in a sense, he discovered something even more fundamental about rats than their maze-exploration skills–rats have a very good sense of hearing and making use of it in their lives.

The trouble with the counterarguments against economism is that they are as laced with political motives as economism, and, as such, as as guilty of cargo cult tendencies.  So if the rising minimum wage does not lead to higher unemployment empirically, there is no dispute that this raises significant questions about the theoretical proposition, but this is not necessarily a “proof” that unconvincingly shows that minimum wage has nothing to do with employment.  A good experimental scientist will want to know whether some conditions can be found where an increase in minimum wage does affect employment, both in theoretical and empirical terms:  the colors, lights smells, etc. of the rat maze, if you will.  This is rarely the argument posed by the critics of economism, and as such, they engage in ascientific cargo cultism of their own, precisely because they want to raise the minimum wage as a matter of policy.

The nice thing about natural sciences is that they offer little incentive for people other than abstract, theoretical curiosity.  I have nothing invested in how rats find their way–unless I have some grand theory about rats’ maze finding skills which would be undermined if rats hear their way through mazes.  So I think it’s great to learn that rats have great sense of hearing and they use it in their lives, including, incidentally, how they find their way around mazes.  If I want to lower taxes, decrease unemployment, reduce the number of abortions, or raise students’ test scores, I am no longer the detached observer motivated only by curiosity.  I become an advocate and I become dumb, blinded by what I want.  Advocates make good cargo cultists, but not good scientists because they believe too much.  Social sciences offer too much to believe in and, as such, they become too normative too quickly, on one side or another.  There is no such thing as a normative science.

DWNominate is NOT a Measure of Ideology, At Least Not Really.

There is a concept in the philosophy of language called “Empty Name.”  Basically, the idea is that that something has a name does not mean it actually exists.  One might say that that something can be measured (indirectly, especially) does not mean that whatever is being measured actually exists.

DWNominate and its abuses (as far as I’m concerned) is something that always drives up the wall, and Kevin Drum on Mother Jones blog, has put on another example.

The argument is that Sanders would have been beaten in 2016 because he is, supposedly, very very liberal.  That Sanders may well have been beaten, one cannot argue with too much since he did not run in the general election and all we have are counterfactuals.  But whether he stood to lose because he was allegedly too liberal requires that, first, there is such a thing as “ideology” exists that can be clearly identified and measured and, second, that we have a measurement thereof that is reasonably reliable.  Naturally, like everyone else who does not know (or, even many who do) how exactly DWNominate is calculated, he appeals to Sanders’ DWNominate scores to supposedly show that Sanders is very very liberal.

The problem, of course, is that DWNominate measures ideology only to the degree that legislators’ votes are products mostly of their ideology and all other considerations (e.g. “politics”) are just random noise.  This is problematic for all manner of reasons, and gets worse when all manner of indirect evidence are mobilized to put political actors outside Congress (e.g. governors) on the same scale.  Basically, someone is identified as a liberal in DWNominate scores if they vote with the Democrats and conservative if they vote with the Republicans.  Someone is identified as very liberal if they vote against both the average Democrat and the average Republican often, but vote with the Democrats more than they do with Republican, and vice versa for the very conservative.  Basically, Sanders’ DWNominate score shows that he’s an outsider who does not often vote with either Democratic or Republican establishments who nevertheless tends to vote with the Democrats, if he does side with one versus the other.  In other words, his scores say nothing that we did not already know, except for the superficial and misleading appearance of precision due to its numerical nature.

This is not quite the same thing as saying that DWNominate has nothing to do with “ideology,” but the linkage is a lot more tenuous than what a simple number seems to indicate.  The overall patterns are not nonsensical, but the specific numbers should not be given the due credence because they are nowhere near as precise as they seem to be.  Yes, Sanders is probably more “liberal” than the average Democrat–but it is not clear what exactly that means, and his voting record, as captured in his DWNominate scores, does nothing to add any more clarity.  Like “empty names” in the philosophy of language, DWNominate (and other quantifications of vague and poorly defined concepts, like ideology) do not refer to something that is terribly well defined.  This is a danger in the drive to quantify everything:  so we have numbers, but what do they really mean?  If you don’t know the answer (and, speaking for myself, I could not explain “liberal” and “conservative” when I was in graduate school because I did not know what they really are, and I know even less what they really mean nowadays.), that could mean that the concepts you are trying to explain may not actually “really” exist, at least not in a form as obvious as one might think at first.

Strengths and Limits of Statistics

Andrew Gelman has a thoughtful explainer that purports to be about “Bayesian statistics,” but really should be a set of points that anyone trying to use data scientifically should be cognizant of.

This, in particular, is critical:

Is there any warnings? As a famous cartoon character once said, With great power comes great responsibility. Bayesian inference is powerful in the sense that it allows the sophisticated combination of information from multiple sources via partial pooling (that is, local inferences are constructed in part from local information and in part from models fit to non-local data), but the flip side is that when assumptions are very wrong, conclusions can be far off too. That’s why Bayesian methods need to be continually evaluated with calibration checks, comparisons of observed data to simulated replications under the model, and other exercises that give the model an opportunity to fail. Statistical model building, but maybe especially in its Bayesian form, is an ongoing process of feedback and quality control.

A statistical procedure is a sort of machine that can run for awhile on its own, but eventually needs maintenance and adaptation to new conditions. That’s what we’ve seen in the recent replication crisis in psychology and other social sciences: methods of null hypothesis significance testing and p-values, which had been developed for analysis of certain designed experiments in the 1930s, were no longer working a modern settings of noisy data and uncontrolled studies. Savvy observers had realized this for awhile—psychologist Paul Meehl was writing acerbically about statistically-driven pseudoscience as early as the 1960s—but it took awhile for researchers in many professions to catch on. I’m hoping that Bayesian modelers will be sooner to recognize their dead ends, and in my own research I’ve put a lot of effort into developing methods for checking model fit and evaluating predictions.

Statistics, I think, rests on three things.  First, we use statistics precisely because we don’t know what the real world looks like.  I don’t remember too much else of what Gary Lorden taught us when I was undergrad (I learned most of my statistical techniques in physics, in statistical mechanics, which is probably why I still tend to do strange things with data), but this is one thing that I remember and has formed the basis of all my thinking.  Second, statistics is a process of inference from a combination of some theoretical assumptions and limited data.  Sometimes, theory obscures facts.  Other times, theory helps make connections not seen in the facts.  An atheoretical model, like Google Translate, sometimes works better than a theoretical model, but that, in turn, assumes a theory–that the existing theories are too narrow for their own good and should be ignored.  Perhaps reasonable sometimes, but not at other times.  As the famous computer science story has it:

“What are you doing?”, asked Minsky.
“I am training a randomly wired neural net to play Tic-tac-toe“, Sussman replied.
“Why is the net wired randomly?”, asked Minsky.
“I do not want it to have any preconceptions of how to play”, Sussman said.
Minsky then shut his eyes.
“Why do you close your eyes?” Sussman asked his teacher.
“So that the room will be empty.”
At that moment, Sussman was enlightened.

The third leg of statistics rests on both of the previous legs.  We don’t know the truth.  Our theory and data are probably wrong, or, at least, incomplete.  As the new data rolls in, we need to keep rethinking what this tells us about how we should think about the world.

The tricky part, though, is that both the theory and the data can be wrong or incomplete.  One could trust that the theory is right and the data is wrong, and this can be dealt with by adjusting the data collection procedures and reweighing the data already on hand.  Or, one could accept that the data is true and modify the theory to better fit the data, or, more likely, do both.  The problem with this, of course, is that there is no “right” answer.  Both the theory and the data are potentially wrong.  You may cross-validate the theory with the data (as is the preference of the data science types) but it is really just a robustness check, to winnow away inferences that depend on a handful of outliers in the existing data.  It is just a mechanical step that addresses neither the theory or the data being wrong or incomplete.  The consequence of this, of course, is that this allows for multiple “statistical truths” to coexist:  different versions that are compatible with different parts of theory or data, and they are all “true” in the statistical sense because we do not know the full scope of the “truth” to contrast them against.  (This echoes the observation by Kuhn about scientific revolutions, and, more narrowly, the sunspots in macroeconomics.  We have different theories of the truth, but we don’t know what is true because of the lack of data.  In a sense, this is analogous to the identification problem, but, perhaps a bit more fundamental.  In the theoretical identification problem, we at least have full data–it just supports multiple theories.  In the scenario I laid out, we KNOW that the data we have is probably wrong or incomplete, even if we don’t know how exactly.  This is far more common in social problems.)

Gelman had written about potential problems where “statistical hypotheses” and “scientific hypotheses” not coinciding neatly, many times, in fact.  Paul Meehl’s original paper is worth a close read as well.  The problem, too, is that, as a commenter to this post points out:

I recall a student asking him, why don’t people acknowledge any of this stuff you’re talking about? He said in his delightful Minnesota dialect (heavily affected at times), “Because if they did it would mean they’d all be selling shoes!”

The problem is not so much that this cannot be done, but this runs into a lot of problems if the scholars try to follow this path.  For the junior scholars who need publications to stay afloat, this is not a worthwhile endeavor (I speak from personal experience, as I am now in the proverbial business of “selling shoes.”  I found that trying to publish in top journals, trying to address these more carefully, is easier than trying to publish in more pedestrian journals.  But most people, myself included, don’t have enough materials for top journals and they need to pad their CV’s with weak publications.) and apparently a problem in the business, too, if the points of note need to be condensed to 500 word memos that need to make the claims emphatically.  This brings back the multiarmed economist problem:  the value added of academic expertise is that it offers nuanced, conditional, and detailed guidance.  But most of the nuances, conditionalities, and details get in the way.  There are very few Hyman Rickovers who see salvation in the details, and to be fair, even Rickover considered abstract and theoretical details to be distracting–although he, in turn, thought that academics were too eager to assume away the details of practical implementation–different lines of work, different incentives, I suppose.

Gelman’s answer is, in a sense, more of a perspective than the solution.  We don’t know the truth.  Our theories are data are both wrong, at least some of the time.  We need the failures, and, indeed, we need to set up the models to fail so that we can see what breaks under controlled conditions so that we can fix them before when we need to bet big–and, even then, maybe we should hedge our bets and not say too much with too much confidence.  This actually looks like good science in general to me, and is something that we should constantly keep in our mind as we work with data.

A New Reformation?

This article by Pascal-Emmanuel Gobry is very insightful, with regards the way informational environment has been changing in face of technology.   The analogy to the Reformation, in particular, is spot on.

If you have been following my posts, you’d have discovered that I am absolutely fascinated by the history and philosophy of science, and the whole process of epistemology in general (I call this a process since, it is essentially a theory of probabilistic–not statistical–learning and unlearning in the abstract.) and have been very critical of the Whiggish interpretation of the history of knowledge, where people consistently get more and more enlightened.  Almost invariably, new ideas that become fashionable in an allegedly enlightened era are grotesquely twisted, overly simplified versions of the original that claim to be capable of doing great and practical things: Mesmer, not Lavoisier, was the ultimate science hero of Enlightenment Europe, if only because the masses, even those who fancy themselves educated and knowledgeable, cannot tell between spurious nonsense and real science–especially since real science is almost always circumstantial, conditional, and uncertain–No scientist actually should “believe” in science.

Science establishment, in other words, is necessarily “conservative.”  It is necessarily skeptical of new ideas and odd phenomenon, although, in principle, not necessarily dismissive.  Cardinal Bellarmine, one of the churchmen presiding over the trial of Galileo but also a learned scientist in his own right and a personal friend of Galileo, in many ways, exemplified the ways of science establishment:  he was willing to accept that Galileo’s argument could make good sense, but that the latter didn’t have much evidence to support his argument to be sufficiently convincing.  This, of course, is echoed by the Catholic Church’s view of witchcraft prior to the Reformation:  weird things happen, but it’s not a big deal.

The trouble, of course, is that the Reformation was not just an isolated event.  It coincided with a great deal of other challenges to the existing body of knowledge, literally from all quarters.  The great era of discovery where Europeans saw distant worlds where the rules of universe that they thought they knew did not apply was unfolding at the same time.  The idea of “progress” was born out of this millieu, along with the realization that the establishment did not know all the answers.  That, for the sake of science, was a good thing.

In this context, it is the establishment intellectuals, both the left and the right, who are acting as the agents of the Counter-Reformation.  Modern day wonks, in a sense, are reprising the role of the Jesuits and the hunt for “fake news” takes on a role not unlike the Inquisition–which, given the anti-Catholic millieu that the modern day Western intellectualism grew out of, strikes me as deliciously ironic.  They are defenders of the intellectual status quo, of the theories that are true because they should be, and we don’t have good explanations for that which we cannot understand–even if we can see them.  To be fair, I’m an admirer of the Jesuit Order and not especially respectful of the Mesmer-like lunacies that grew out of Reformation and other intellectual movements (I’d written about Thomas Muentzer in an earlier post–an actual example from the Reformation Era). But it is also true that Reformation did end the central role played by the Catholic Church as the locus of Western intellectual development, as the Catholic Church no longer offered the foundations of the “theories” that the West would understand the universe.  (One could point to LeMaitre or Mendel as the counterexamples, but they were churchmen by job, and their “science” had nothing to do with the Church.  Indeed, LeMaitre was very insistent that Church has nothing to do with the science–God is not a theory, he said, while all of science is a theory, he might have added.)

We have discovered, proverbially, a whole new continent whose ways we are finding unfathomable:  Trump and other weirdness, like LePen, Brexit, and others.  The old conflicts, like the crusades at the time of the Reformation, may still be continuing–the conflict between the establishment liberals and conservatives, may still continue.  After all, even as the Reformation was unfolding, the Catholic Europe was still fighting against the Muslims, now in the form of the Ottoman Turks across a geographical scope much broader than before:  the Balkans, the Mediterranean, and even the Indian Ocean.  Yet, these were side shows to the greater sweep of European history in retrospect, except in some parts of Europe.  The old political conflicts may not exactly be irrelevant just yet, but making sense of the new discoveries, so to speak, will supersede them in importance doubtless.

Will this be more liberating, or will this be more disruptive, I wonder.  The effects of the Reformation were mixed:  Luther and Henry VIII, among others, were themselves deeply conservative and did not care to disrupt the existing order of things too much.  Leibnitz and Newton, despite being from Protestant parts of Europe, did not cut themselves off from the Catholic parts of Europe.  Catholic theology was reformed as much as that of the Protestants’ precisely because the Catholic churchmen did not refuse to address the theological and intellectual challenges that the Protestant thinkers were raising.  If the wonks today do not insist too much on burning Protestants instead of listening to their arguments, this might lead to interesting developments.  But the Thirty Years’ War, for all the progress in the aftermath thereof, was  a destructive, deadly conflict.  This is likely where we are headed in not too distant future.  May the conflict be only proverbial, fought only in words, and not actual.

 

Risk, Profits, Outcomes, and Theories.

ESPN blog has an interesting set of insights about how use of data is transforming strategy in baseball.  The short version:  people are substituting risk for “data,” so to speak.

Now, that is a bit misleading:  people are using data to evaluate risk, and taking risk intelligently, and that has always been the case, with or without the big data.  Mickey Mantle and Sandy Koufax were so good that you didn’t need data to tell you that they were great ballplayers.  You use data so that you can systematically identify and evaluate the risk of the ballplayers who are not that obviously talented.  So, in a sense, is the traditional scouting:  ballplayer X “looks like” Koufax based on qualities that may not be so easily quantifiable, so I am willing to risk that he is a good ballplayer like Koufax.  Of course, the chances are considerable that X is not Koufax, and the disadvantage of non-quantification is that you can’t retrospectively evaluate where you went wrong with precision, whether about X or about Koufax.

The part about being able to go back and retrospectively evaluate with some precision where things went wrong is the critical component of what makes something “science.”  It is not necessarily impossible without quantification, but the limits in “precision” are reached quickly without the numbers.  Yes, X did not look like Koufax so much after all, upon closer look…but how exactly NOT like Koufax was X?  Numbers help round out these differences with precision, whereas non-numerical differences, even if real, are harder to describe.

The problem with Sabermetrics and Big Data mentality, in general, though, is that it dispenses with the science to a large degree.  So is Danny Espinosa more valuable than Johnny Giovatella by 2-3 games because of fWAR?  That is, will Angels win 2-3 more games by replacing Giovatella with Espinosa, while holding all things equal?  It is highly unlikely. fWAR might be a useful aggregate statistic, but how exactly it is reached is not clearly known (bWAR is more easily calculable, but there’s no way to tell if it means actual “wins above replacement” since it is just a patchwork of existing baseball stats.) and, even when it can be calculated, there is no way to tell if it really means what it thinks it is without a sense of imprecision of the statistic in general and with regards its application to Espinosa, Angels, and Giovatella specifically.  Of course, this imprecision is the source of both risk and opportunity:  while the fWAR says something, we still don’t know exactly what will happen and there is no way to evaluate it beforehand.  So Eppler, the Angels’ GM, can take chance that Espinosa will overperform fWAR and Rizzo, the Nationals’ GM, that he would not.  That subjective evaluation|Angels > subjective evaluation|Nationals implies that there is a profit to be made from the trade on the present information although, when the full truth is revealed at the end of 2017 season, one will have lost relative to the other, with much likelihood.  (i.e. are prospects Angels sent to Nationals, who looked terrible last year, any good in next year’s reality or what?)  Angels are willing to pay more than what the Nationals consider Espinosa to be.  So both sides win, until the truth is revealed, that is, by the end of the 2017 season.

In other words, evaluating risk and gambling on them is, in a sense, fundamentally the opposite of what people usually think they are gaining when they do “science.” If the risks and gains can be made more precise, with systematic evaluation of data, the former can be minimized and the latter maximized.  You can identify players who are like Koufax in every manner quantifiable and, if no one else knows the magic formula, lock him up before anyone else gets in the act.  You pay no penalty for the risk and reap all the gains.  But once everyone knows the formula, the competition will wipe out any gain that you stand to make.  You have to recognize the differences between X and Koufax and ask yourself how big a gap these imply for the outcomes expected from X and Koufax.  You may quantify these differences, you may subject them to complex models, or whatever.  In the end, the answer is not precisely known–if it were, there’d be no risk.  So the choice becomes subjective.

One great irony is that, once the bounds of certainty are reached, there is no real gain to be had from quantification, at least in the short term.  You may crunch numbers and churn out models, but what you don’t know, you still don’t know.  You go through the number crunching motions, at least in my part, to satisfy your audiences, to give them the impression that you know what you are doing–even if you don’t, at least not precisely.  You are no better than a traditional scout or, indeed, a fortune teller or witch doctor.  At best, you are making an educated guess.  At worst, you are guessing randomly.  Thus is explained why there are neither true believer nor true atheists in the trenches:  you know what you know and you know what you don’t.  You minmax risk and rewards when you can, and you are going off on a leap of faith when you don’t.

What you gain by having a theory–again, quantified or not–is that you’d know what to do when you are wrong.  The theory, if properly concocted, would have given you an understanding of what to expect from different moving parts.  The results that you expected did not obtain because at least some of the moving parts did not move as you expected.  So what parts of your theory did not work?  Again, quantification helps make things precise, but it is neither necessary nor sufficient.

The trouble starts when the numbers are no longer part of “science,” but mere formulas whose value is simply that “they work,” just because.  So what if they fail?  What if a pitcher with terribly FIP keeps pitching like an ace day in and day out?  Some sabermetrics fans insist on insisting that the pitcher is wrong because he is really terrible–FIP says so.  But, in the end, baseball games do have winners and losers and that’s final, no matter what FIP says. Of course, I can’t help but insert some political commentary here:  in 2012 and again in 2016, the losers and winners of presidential elections alike went off blaming and crediting “big data” for the outcomes, with the losers degenerating to outright denialism and conspiracy theories.  This is a sign of data being abused, in service of cargo-cultish pseudoscience, exactly the opposite of science.  As described above with regards sabermetrics, a science that works too well can breed this kind of attitude:  up to some point, better application of existing science CAN simultaneously raise the mean and lower the variance.  When the boundary of the science is reached, the only way forward is to gamble, because you don’t know.  In a sense, this is far harder a challenge for social sciences than the natural, because the boundaries of uncertainty are far too close.  Deviations in the extent to which sunlight is bent by planet Mercury is too esoteric a thing.  On the other hand, while the vast majority of the people may have party id and vote party id, the exceptions are numerous enough and are important enough in elections for the “theory” (that people are partisan and vote party) to be any useful except as an abstract starting point.  Whether Jered Weaver or David Cone, in their heydays, were great pitchers or not, notwithstanding their terrible FIP, I suppose, is somewhere in between.  Like Enrico Fermi said, when results confirm hypotheses, you have made a measurement, when results contradict, you have made a discovery.  Sometimes, discoveries are good things.