Evil and the Agent-Structure Problem of American Racism

I was listening to the "Do You Like Prince Movies" podcast while biking today, and it got me thinking.  Alex Pappademas and Wesley Morris noted that Hollywood tends to depict people who have messed up lives by having them eat out of Chinese takeout cartons.  This is anti-takeout-ist they noted.  Eating out of takeout cartons is not wrong!

Anyhow, their conversation eventually turned to the appearance of President Obama on the Marc Maron podcast.  Alex expressed his frustration with the discussion of the murders in the Charleston event as evil, that it allows us to shrug and move on, rather than consider why the event happened.  And this got me thinking while I biked along a particularly beautiful part of Ottawa about evil and the agent-structure problem of Racism in the U.S.

As a quick caveat, I am not an expert on any of what I am about to discuss.  I am starting to realize how apt the name of my blog is as something that is spewed is not filtered.  Something that is semi-spewed is only partially filtered.

I don't want to be too social science-y, but when I think of the Charleston event, I don't want to let anyone off easily.  That is, if we blame society or politics, we focus on the social/political structures at the expense of agency--that this one individual had free will and chose poorly, so very badly.  If we focus on the individual, that the perpetrator was insane or racist or whatever, we tend to gloss over the conditions that enabled him and even encouraged him to kill nine innocent people.  The reality is that both matter and both... can be understood and perhaps even treated via social science.

Observers/pundits/folks trying to make sense of events like this prefer to privilege one side or the other, the actor or the structural conditions, and get upset when people focus on another aspect.  Nobody is wrong for focusing on the agent or on the social/political structures, but they are wrong for dismissing the arguments of others who focus on the other part of the problem.

I do think that focusing on "evil" is a problematic approach because it takes us away from agents and structures.  There can be "evil" within individuals and within institutions, but we need to explain why there is variance--some people and institutions are more "evil" than others--as the variance provides us with leverage so we can figure out causes and conditions that facilitate the bad stuff and then develop ways to mitigate/reduce. 

Why focus on agents?  Because lots of people exist within a set of norms (what is appropriate behavior, what is expected behavior) and institutions and yet few pick up a gun and shoot a bunch of people (too many, but still not only a few).

Why focus on social structures and political institutions?  Because an individual who is somehow messed up may act out in any number of ways.  Why target African-Americans?  Something in this guy's social scene/growing up/whatever helped direct the blame for his circumstances towards African-Americans.  Why was he able to kill so many people so quickly?  He lives in a society where access to guns is easy and where the focus is on gun rights and not on gun responsibility.

Obviously, there is more to both sides of this problem, and that is where social science comes in--figuring out as much of this as possible not so that we can eliminate evil (even I am not that naive or optimistic) but so that we can reduce the harm caused by these agents and these structures.

And to both illustrate the agent-structure problem and to have something pretty to offset the ugliness of racial divisions, I present a picture from the aforementioned bike ride.

Social Science Done Right

Erica Chenoweth is seeking to do good stuff and is asking for our help.  She wants to understand how to communicate about non-violent resistance better.  Given that her work shows that nonviolent resistance works better than violent resistance, this makes a bunch of sense.  So, check out the survey, which will take 15 minutes or so, here.

Ignorance is Bliss, Canadian Edition Continued

I posted last week about the strange thing that Canadian Prime Minster Stephen Harper said about committing sociology as if it were a crime.  Well, this sociologist has produced an incredibly biting, satirical response: that he is a homegrown sociologist.


Probably will not change anything as this government seems impervious to such stuff.  I truly doubt that it will face much of a challenge in the next election, so we are probably stuck with the forces of ignorance for quite some time. 

Of course, Canada is hardly unique in all of this.  Ron Rogowski just published a pretty trenchant piece noting that the problem with political sicence is not that it is not policy relevant but that we are too policy relevant--we find things that are inconvenient truths.  The only time politicians want to hear what we have to say is when we can give them advice on how to win, not on how to govern good.

So, in sum, we social scientists and society in general are just screwed.

Most Similar Secessionists

Watching the news about the Scottish referendum on independence, and it is almost stunning how much the stuff parallels Quebec.  Let me count the ways:

1. Neither Quebeckers nor Scots have a majority that want complete independence.
2. The Scottish National Party today and the Parti Quebecois in 1995 don't want a single clear question, and certainly don't want the national parliament having a say over the question.
3. The SNP and PQ are both musing about changing the election laws to allow 16 and 17-year olds to vote.  A wonderfully clear admission that they are desperate for votes.  One only adds new voters to the pool if one cannot get over a threshold without them.
4. Speaking of which, what would be enough votes?  Fifty percent plus one is the Quebec "standard."  But Scotland is in a country that is in the EU (while the EU is still around), and the EU applied 55% as the standard for Montenegro.  Hmmm.

However, I need to be clear that these cases may be less than fruitful for social science--there needs to be some variation between two cases to have a useful comparison.  Either two very similar cases have a dissimilar outcome, or two different cases have a similar outcome.  And I would like to place a bet on an independence referendum failing, just as Quebec fell short in 1995 and is nowhere close today.*

*  Indeed, this morning's paper was chock full of stories about the PQ trying to figure out how to get folks enthused about the party and about independence.  That they must find ways to get the separatist mojo going provides significant evidence that the PQ and Quebec have won the big battles, making independence or even a confusing referendum that mixes and matches autonomy, confederation, and iendependence unnecessary.  Well, unnecessary for being the PQ's raison d'etre.  Well, other than running against Montreal.  This weekend's meetings has heaps of proposals for shifting yet more resources and power away from Montreal.  

Residual Categories Make for Tough Social Science

One of the basic realities in any effort to categorize things is that a bunch of stuff is left over.  They are put into a separate category of other stuff.  Any effort to give it a sense of unity is likely to stretch the concept so far that it bursts.  The exemplar would be Clash of Civilizations where "Sub-Sarahan Africa" does not fit into other categories and is thus considered being one civilization.  The problem is that it does not actually have any real unity is akin to the stuff that unites all Hindus, Westerners, Muslims (Clash is mostly about religion), and so forth, opening up the author (Sam Huntington) to charges that he is was racist.  That and Central and South America are seen as a civilization distinct from other Catholic countries (Italy, Portgual, Spain, etc) because the folks there are .... um, brown.

Anyhow, this clip illustrates the challenges of the residual category:

Reality Provides Social Science Experiment

Is it Harper or is it about minority government?  That is the question of the day/week/month/year.  Prime Minister Stephen Harper was adamant that there would be no boots on the ground in Afghanistan after the Canadian mission in Kandahar ends in July.  Then, he came up with a new training mission that was to be Kabul but is now Kabul-centric since there were not enough slots in Kabul itself.  And now the allies are seeing the new majority government as being perhaps more willing to stick around in Kandahar.

So, we have a nice apples and apples or apples and oranges comparison since we have the same leader, the same foreign policy challenge and so on, and only one big difference--going from minority government to majority.  The Auerswald and Saideman book in progress (July as an end-date?) does focus on such things as the nature of the government, so on one hand, we expect that this might make a difference.  However, we also find in the book that minority governments are not as paralyzed as people would aver, as it really depends on who bears the onus of cooperating.  That is, does the government need to get some part of the opposition to cooperate?  In which case, the government is weak and restrictions are likely.  Yet if the opposing parties must cooperate among themselves, then the government may actually have a free hand.

Thus, it should make a difference if the government is majority or minority.  BUT our book has a second step in it--when individuals are empowered (as in Presidential systems and single party parliamentary governments), the personality of that individual matters.  Perhaps Harper himself is not a big fan of the combat mission, especially since his control freak-iness is constrained when it comes to the fighting.*  So, to have more control, he might be inclined to stay out of combat.

*  There are big risks if a leader does micro-manage a war effort--as he/she then becomes directly responsible for anything going wrong, and Afghanistan is a place where lots of things can and do go wrong--IED's, prison breaks, corruption, etc.

So, I am less sanguine about the chances of Canadians being sent to southern Afghanistan than the allies.  But as a social scientist, I am keen to see how the experiment plays out.

Crowdsourcing Human Rights

I have always admired not only the academic work of Christian Davenport (everything I know about repression comes from him and Will Moore), but his data determination.  He went to great lengths to determine how Rwanda's genocide played out, at the risk of being called a holocaust denier for simply stating the truth that more Hutus died than Tutsis.  He then developed the definitive repression website (a site for researchers of repression, not for repressers to pick up tips): www.staterepression.com  (I reviewed that site a while back).  Now, he is trying to get folks to report human rights abuses and direct data his way to crowd-source via this site: https://iii.nd.edu/. 

People wonder how the internet has changed what social scientists do.  Davenport keeps teaching us that we have only scratched the surface.  In addition, he demonstrates how important it is to be patient.  His Rwanda project took quite some time and resources to generate the findings.  This new project, relying on crowdsourcing, is a very much a long-term project as the crowd will take a while to gather.  It will take more than just a minor blog like mine to get enough folks to gather virtually and provide the data that Davenport needs to get the crowdsourcing to work.  Spread the word.

Get 'Em Early

As always, if you measure what you need to measure rather than what you can measure, you actually might learn something.  This article shows that better kindergarten experiences may not produce long-lasting improvements in test scores, but might actually produce better life outcomes.

As Raj Chetty, a Harvard economist, says: “We don’t really care about test scores. We care about adult outcomes.”

So, what is going on?

The economists don’t pretend to know the exact causes. But it’s not hard to come up with plausible guesses. Good early education can impart skills that last a lifetime — patience, discipline, manners, perseverance. The tests that 5-year-olds take may pick up these skills, even if later multiple-choice tests do not.

MC Exams are best at testing MC exam taking skills.  The other thing going on here is that you enjoy your first tastes of the educational system with great kindergarten teachers (like my daughter did in our one year in Virginia), it might create more enthusiasm and interest in learning.

As an educator, I, of course, buy into these results due to self-interest and identification.  The broader context?

Barely a week seems to go by without a newspaper or television station running a report suggesting that education is overrated. ....  But the anti-education case usually relies on a combination of anecdotes and selective facts. In truth, the gap between the pay of college graduates and everyone else grew to a record last year, according to the Labor Department, and unemployment has risen far more for the less educated.

When I was teaching American (and Texas) Public Policy, I would always show some graphs on education levels and pay, and the gap between college graduate and high school graduate was always quite steep.  So, I would point out that my class, required for graduation, was going to make all the difference between different lifetime outcomes.  Good to know I am still right about that.

Of course, the bigger problem is not necessarily measuring the outcomes of the kids but the quality of the teachers and the programs.

But more data, studied by folks with good educations, help to make sense of what is going on.  Well, as long as it confirms my biases.

Engineering Terrorism

This piece builds on the recent failed attempt and an old paper that I saw presented a few years ago--that Islamic terrorists are disproportionately engineers.  The Slate review lists the usual suspected explanations: lack of good engineering jobs in most of their home countries (Saudi Arabia excepted); the skills issue; that engineers are more conservative and rigid, so they adapt poorly to change; and the like. 

But the article omits a key issue--the data that Gambetta and Hertog use relies on whose college majors?  Those terrorists that have either been captured or killed.  So, this raises some questions--does being an engineer make someone more likely to get killed or captured?  Well, in the original paper, Gambetta raises an issue that might be relevant here--engineers are more likely to have autistic children and are more likely to have borderline social skills.  While they spin that as meaning that these folks adjust poorly, I take another reading--that their dataset may include more engineers because engineers may be more likely to get caught due to poor social skills.  Either they reveal themselves or they may be bad at detecting infiltrators.... 

Anyhow, read the original piece, as it is a model of interesting social science--taking a correlation and then working backwards to figure out all kinds of possible explanations.

The Benefits of a Long Drive Home: Ask the Reader for Syllabus Ideas

Well, one of them is new blog ideas. So, hopefully, I can return to my pre-vacation pace. First post is inspired by a book I was reading over the vacation (or holiday as they call it up here): Traffic: Why do We Drive the Way We Do by Tom Vanderbilt [hat tip to Steve Greene for referring it to me].

I am far from done, but the book is great and has already changed one driving habit--I am now a late merger--and we experienced the signs and merging experience discussed in the book driving through Pennsylvania.

Anyhow, the book and the drive made me realize that it would be fun to teach an undergraduate seminar on "pop social science." I have belatedly become a fan of this genre or this genre has recently bloomed--I have not done the research to figure out how new this literature is. Anyway, there are courses on political science in film, but I don't know if there is anything like pop social science.

The three books thus far for such a course would be: Tipping Point by Malcolm Gladwell; Freakonomics by Stephen Dubner and Steven Levitt; and Traffic. Since I am too tired to do research now from the long drive, I am asking my readers for ideas--any good books that present some sort of social science (preferably but not exclusively political science) to wider audiences to popularize some set of ideas?

And what would you do with such a course?

Creatures of the Flame: Cooking and Evolution

Very interesting piece in Salon today on the role of cooking in evolution. The big difference between us and the chimps are not so much the thumbs but the cooking of meat. Cooked food is denser in nutrition, so the efficiencies of cooking gave our predecessors advantages over the non-cookers and facilitated the development of our brains (imagine zombie reference here).

Yes, you do quite a convincing job of arguing that a purely raw diet cannot sustain an active human. Do you believe that we have evolved to a point where a raw diet is fundamentally against our biology?

Yes, I suppose I do. If I hesitate, it is because I certainly recognize that raw foodists who live in an urban area of a well-to-do nation can make it work, so it's not that much against our biology. But I do feel very confident now that going off into the wild and living like a hunter-gatherer on raw food is not possible. People who switch to a raw diet report feeling constant hunger and lose large amounts of weight, even when they are careful to take in at least the nutritionally suggested number of calories a day for an adult. Basically, all the studies show that over the long term, a strictly raw diet cannot guarantee an adequate energy supply for our bodies. In other words, raw foodism is against our biology in a state of nature.

It gets even more interesting/controversial--after some snippingl:

So the concept of marriage began fundamentally not as about power or sex, but food?

Yes, though that would mean that women always do the cooking, and when I first started down this path, I wasn't at all sure that was the case. So, I went to the anthropological literature, and sure enough, I found reports of societies where men did the cooking. But then I dug into it more carefully -- and I discovered that, in the cases where the anthropologists claimed the men had done the cooking, the scientists had been wrong. In every single society women cook for men. And, what's more fascinating, in many societies you can really say that food or domestic promiscuity is far more serious than sexual promiscuity. In other words, it's more of a breach of social convention for a woman to feed the wrong man than it is for her to have sex with him.

Hmmn. Suggests all kinds of new plot twists in the average soap opera or drama. But what I really want to know now is why do men do most of the barbecuing these days? Anyone?

And, what does this say about the current generation of men-folk? During grad school, it seemed that in every couple, the man was as good as the woman in the kitchen and even better in some cases. It would be immodest for me to say that I am a better cook than my wife, but not necessarily untrue. So, does that make me un-evolved? Regressing or post-evolution?

Blink on the Battlefield: Spidey-Sense

Interesting lead article in NY Times today on intuition and the battlefield.

The men and women who performed best in the Army’s I.E.D. detection study had the sort of knowledge gained through experience, according to a preliminary analysis of the results; but many also had superb depth perception and a keen ability to sustain intense focus for long periods. The ability to pick odd shapes masked in complex backgrounds — a “Where’s Waldo” type of skill that some call anomaly detection — also predicted performance on some of the roadside bomb simulations.

It reminds me of Malcolm Gladwell's Blink. I did find it a bit frustrating because it essentially argued that we have instincts that are often quite sharp and accurate, except when they are not. It was a fascinating book with lots of good pop social science, but not quite as strong as his first book, Tipping Point, or his more recent one, Outliers.

The article goes back and forth from stories about soldiers/marines in Iraq to the science of perception/intuition. A fascinating read.

Sometimes I wonder what would have happened if I had not been scared off from switching into Psych 101 two weeks into the term when I realized that chemistry was out and poli sci was in my second year in college.

Measuring Progress and Policy

The new headline reads: 1 in 7 released detainees rejoins the fight. Obviously, we would prefer if none of the folks released ever fought again. But, the recidivism rate for criminals in the US is somewhere around 60-70% and perhaps a bit lower in the UK [Wikipedia as the source, but other studies seem in line]. So, is one in seven good or bad for those detained? I would say it is much better than we have a right to expect, given that an unknown number were exposed to some abuse and rehabilitation was not the aim of incarceration. Of course, critics will look at the numbers and assert that one in seven is too high. But then again, where do we put the detainees if we want to keep them forever? Not in my backyard apparently. There are no easy solutions to this problem, one that was hard enough even before the Bush Administration screwed it up beyond belief. That only one in seven are getting back into the fight is actually good news.

And this gets us back to the recurring theme: cognitive consistency. Ok, the process in play is not just psychological but political--that opponents of Obama will want to take the darkest possible interpretation to undermine him and his supporters (perhaps even myself) will always take the most positive to try to win the political battle of the moment.

The measurement problem--how do we know we are making progress, how do we know we are making good decisions--is incredibly hard. Did we overreact to H1N1 or not? No pandemic thus far, but the effects of preventive efforts, like deterrence, are really hard to measure. And we are really impatient. I remember being asked by the media days after the bombing of Serbia started about the policy's failure, and folks were quick to judge the invasion Iraq to be a failure because of the sandstorms. In both cases, the objectives were reached in a reasonable time frame, but one that was longer than expected. So, expectations drove the perception of success. More problematically, the real question for both campaigns is not whether US military could achieve its goals, but whether the effort was politically and strategically sound? What are the goals, what are the costs, and what are the second and third order consequences of the policy?

One of the common threads in most of the discussions I have had with miltiary types in Canada, the US, Brussels and Afghanistan is that progress in counter-insurgency campaigns is incredibly difficult to measure. Body counts clearly do not suffice, but neither do number of clinics or schools opened. The best way to measure is whether locals, Afghans in this case, are willing to bet on the government or the insurgents. Are they willing to provide action-able intel to NATO, ISAF and the Afghan government, do they sit on the fence, or do they help the insurgents? Unfortunately, this information is classified.

Even if we had such a measure, that indicated trends in where the Afghans stand, we could still debate about how much progress is appropriate given how much time has passed and how much has been expended. Unfortunately, the debate would probably not turn so much on the facts, but on how it could best be played. Again, 1 in 7 recidivists among the detainees is really not bad. Perfection is not an option.

Sorry for the rambling, but then again, you can best measure for yourself whether this post was progressive (illuminated) or regressive (confused or even destroyed knowledge).

Geek Squared! Star Trek and Public Goods

I have a modest spoiler so don't read below if you care about the content of the curriculum at Vulcan Junior High.





Ok, not to give anything away in the movie, but in an early sequence where the various Vulcan kids are being tested on all kinds of knowledge--math, biology, physics, etc--there was one clear social science reference--non-rivalry & non-excludability. These are the defining characteristics of Public Goods, and are central concepts in Mancur Olson's classic Logic of Collective Action. Olson famously (famous enough to make it into the 23rd century and all the way to Vulcan) argues that it will be difficult to cooperate to provide Public Goods--when one's usage of a good does not reduce its availability for others and where non-contributors cannot be excluded. In such cases, free-riding is not only likely but the logical course of action.

Olson's insights here, while often contested, have shaped how we think about political mobilization, the need for government/coercion, and especially the environment. Why recycle? If I do it and no one else does, my effort is wasted. If everyone does it but I do not, then the planet is clean and my marginal waste is irrelevant.

It is especially ironic that Public Goods appears in the new (and rockin') Star Trek because the logic of collective action largely produces a pessimistic, cynical outlook towards human nature and the possibility of cooperation but ST is all about an optimistic vision of the future. Perhaps the new writers of ST mythology, JJ Abrams and the gang, are trying to square the geek circle--that we can overcome the collective action problem and that we will so that we get this amazing future of the Federation of Planets, in which Starfleet is explicitly referred as a force for humanitarian efforts and peacekeeping. Indeed, for a minute, it sounded like the conventional perception of the mission of the Canadian Forces before Operation Medusa in Afghanistan in 2006 (first major combat for Canada since the Korean War).

Of course, Olson and his successors provide a variety of ways out of the collective action problem: small groups can mobilize as their members perceive their contributions as mattering; individual entrepreneurs can see it in their interests to facilitate and even underwrite cooperation; and the like.

To sum up, it was just a fun movie that made a very interesting and deliberate reference that is far more intriguing to me than the volume of a cylinder--another example in that sequence in the movie. This is not the first time that Star Trek has overlapped with social science, and it will almost certainly not be the last, given the box office success it is having. Live long and prosper, indeed.

Wolverine, Reality TV and Social Science

• Wolverine
  

People who know me know that I am big fan of movies, TV, and books, including comic books. I used to collect, especially X-Men and Spiderman related titles, until the choice became diapers or comic books. It was easy to make the break at the time as the plots of the books at the time went deeply south and Marvel was proliferating titles so it really became hard to follow a single group or character without buying4 or 5 books a month. Anyhow, the past several years have been a bonanza for those with my interests--heaps of good to excellent comic book-based movies: X-Men, the new Batman, Iron Man, etc. So, we saw Wolverine today, and I will not talk about it much for about a week so that my three or four readers are not spoiled. I will say two things:

1. definitely a fun movie to see in the theatre with all of the noise and effects; and
2. After seven years in Canada, it has struck me how un-Canadian Wolverine is. Yes, his roots in the books and the movies are in Canada, but the opening montage shows him fighting in American (and Union) uniforms. Wolverine is one of the most aggressive superheroes in the Marvel universe--he has killed; he was prone to beserker rages early on (I don't know how he is portrayed in the books these days); he is a loner, etc. Canadians are surprised that their military is trained to fight and kill; always seek consensus (hence the eh? that is either implicit or explicit); and are pretty calm if they are not Montreal drivers. Perhaps that is why Wolverine spends most of his time in the US.

Reality TV and Social Science:
While biking today, I was listening to my favorite podcast: the Sports Guy, a columnist for ESPN and former poli sci major, about 4-6 years my junior with a similar passion for pop cultural references. His conversation about a favorite Reality TV show reminded me of something I have been thinking about for a few years. I wish I had a research agenda that required experimenting on people. That's right--on people. Malcolm Gladwell's Tipping Point reminds us of the dark ages of social science, with various conversation experiments on people--the Milgram authority/shock experiment; the Stanford Prison experiments; etc. Now, each university has a research ethics board to approve any funded research that involves experiments or people. I have had to fill out the forms so that I could interview military officers about their experiences. My students have had to do the same. So, it would be very difficult or impossible to get approval to do morally questionable research on people today.

And that is why we might want to turn to reality tv game shows. Rather than getting funding from an established foundation with procedures for treatment of human subjects, if one has a question that requires putting people through troublesome situations, there is now an alternative. It does not take much viewing of the various reality TV shows to see that young and not-so-young people are willing to do pretty much anything for a modicum of cash and/or their 15 minutes of fame. We can already learn much from the first season of Survivor about how people react to competition with few rules and modest manipulation. We can learn from the subsequent seasons how people learn and imitate as the strategies of the first season, such as alliances and duplicity, are applied anew or not so anew.

If only I had testable hypotheses about individuals or small group behavior?!