“What If” Questions, Bond Villains, and Depressing Data about Durable Inequalities

“One of my favorite blogs out there is “What If?” It’s made by the guy who does XKCD and its tag line is “Answering your hypothetical questions with physics, every Tuesday.”  There have only been twenty Tuesdays since the blog started, but he’s already covered such pressing topics as “How much Force power can Yoda output?“, and “What if everyone who took the SAT guessed on every multiple-choice question? How many perfect scores would there be?“, and “How long would the sun last if a giant water hose were focused on it?“, and “What would the world be like if the land masses were spread out the same way as now – only rotated by an angle of 90 degrees?

He calculates each one out to the best of his ability, making clear what assumptions he’s making.  For example, for the rotated land masses one, he attempts to figure out what the weather would be like in the new world and such.

Randall’s “wild guess” about where hurricane basins would be.

I just last week wrote one of my friends and said “Imagine how sweet it would be if there were a social science version of this.”  I brainstormed a few questions I’d want to ask (my friend works for an economics blog which is why my questions have a lean towards economics instead of sociology or one of the other social sciences based I actually read a lot of):

  • What would happen to domestic labor markets if all countries eliminated work visa requirements and everyone could work in any country they’d want to?
  • What would happen if China traded all its dollars for euros in one day?
  • What if all racism disappeared tomorrow, how many years would it take for black-white income disparities to disappear?

I happened across something cool today: “Ask an Economist: Which Bond Villian Plan Would Have Worked (and Which Would Not) [sic]”.  Uhh… despite the poor grammar of the title, it’s pretty awesome.  Oversimplistic, sure, but still pretty awesome.  See, I thought Bond villains were primarily out just for some weird kind of revenge or some interstate ideological victory (Golden Eye was the first Bond movie I saw, and the one that made the most lasting impression).  Turns out, logically, a lot Bond villains just want to make a little paper.  And a lot of Bond writers have no idea how the economy works.

The actual article is clearly comprised of about twenty minutes or so of describing Bond plots to an economist (Jean-Jacques Dethier, a development economist at the World Bank) and getting his off-the-cuff reactions.
For example:

Plot: Gold tycoon Auric Goldfinger’s (Gert Frobe) plan is quite simple: He wants to attack the U.S. Bullion Depository in Fort Knox and detonate an atomic bomb, thus irradiating the gold stored there, rendering it worthless for decades. This will in turn increase the value of Goldfinger’s own gold and cause economic chaos in the Western world.

Plausibility: “This looks plausible to me,” says Dethier. “If you irradiate the gold, you can’t touch it — which will effectively reduce the gold supply, at a time when the United States currency was still on the gold standard.” However, there is one potential problem — the vast majority of the gold in the U.S. wasn’t in Fort Knox — it was (and remains) in the basement of the Federal Reserve Bank of New York, in downtown Manhattan. (But most of the gold in New York belongs to other nations, so Goldfinger’s evil plan is still fairly solid.)

Not as detailed as I’d like (there’s no attempt to calculate the change in the value of gold this would bring, and there’s no attempt to anticipate complications), but still really fun and worth reading.  Spoiler alert: most Bond villains turn out to be idiots.

Which brought me back to thinking about the questions I asked my friend. The last question is probably the easiest to answer and the most sociological.  To phrase it differently, what if race disappeared and black socio-economic mobility became like white socio-economic mobility?  How long before the original “black” group had same income distribution as the “white” group?  Evidence from Sweden is: umm, it’s going to take a while.  There was a really interesting study of accumulated advantage and social mobility based on Swedish last names (Slate.com’s summary here, UC Davis Econ Professor Gregory Clark’s full paper, “What is the True Rate of Social Mobility in Sweden? A Surname Analysis, 1700-2012” here).

See, Sweden stopped minting new noble houses in the 17th century so anyone with a “noble” last name probably has been rich for a while.  Likewise, anyone with a last name ending in “-sson” was probably on the lowest strata of society right around 1901, when the surnames got fixed (before that -sson surnames were purely patronymic, like Iceland’s still are: Peter Andersson’s Sven would have been Sven Petersson, not Sven Andersson).  There are several distinguishable groups (titled and untitled noble families, latinized names [-ius/-eus], Lund, -berg, all above the most common -sson) but, more or less, your last name hasn’t meant anything for the last couple of generations.  A noble like Adler, a “latinizer” like Linnaeus, and a former commoner like Andersson have no different rights or legal advantages in contemporary Swedish society–they’re all equal under the law.  And, at least equally importantly, this is Sweden there’s no social or job market discrimination  against dudes named “Andersson”.  However, between a guy named “Andersson” and “Adler”, there would likely be differences in “accumulated advantage” (also known as “the Matthew effect” or, simply, “the rich get richer”).  The former nobles, without any legal advantage, still have material, social and cultural advantages and, even in a meritocratic system, these advantages prove durable.  Think of it this way: my father was a university professor–you better believe I had certain advantages over my peers with less cultural capital as I moved through the education system (nevermind the SAT class my dad paid for when I was 17 and all those expensive extra-circulars I had, all of which helped me get into college.  In fact, come to think of it, every job I had until I graduated college was also set by through my familial or social network).

Clark’s paper doesn’t document those advantages, but he doesn’t need to.  He shows that several generations after nobility meant very little in terms of societal advantage, the former nobles still have it pretty good and the former commoners still aren’t doing nearly as good.  Let’s take a look at doctor’s in Sweden for example:

In Sweden today there are only half as many Andersson doctors as you’d expect if doctors were spread throughout society completely by chance, but roughly twice as many people from untitled aristocratic families and latinized families, and four times as many as would high aristocratic families.  Sure, the well-to-do families have lost some of their advantages of 100 years ago, and while it might look like the process is slowing down when you eye ball, but really it’s progressing at about the same pace: 70% of the last generations advantage is still being inherited (that’s what the “b=.70” means).  Draw imagine the line below continues on, when does it “1”, the point where people from elite families are no more likely than chance to become doctors.A similar pattern exists for lawyers: For top academics: And university students:

And probably another elite field you can get data on.  Okay, so what is all this?  What’s going on in nice, socialized, egalitarian Sweden?  (A note on “egalitarian”: in Sweden, a university professor makes 140% of a bus driver’s salary.  In the U.S., it’s 160%. in Sweden, a doctor earns 230% of a bus driver’s salary.  In the U.S., it’s 600%.  This is one of the reason Sweden has better social mobility than the U.S.–it has a smaller gap to begin with, see Clark pg.31-32).  Well, remember how I said earlier, “70% of the last generations advantage is still being inherited (that’s what the ‘b=.70’ means).”  A perfectly meritocratic system where we assume aptitude is spread equally throughout the population (epigenetics might mean this is an inaccurate assumption, but let’s avoid that whole discussion) would have a b=.00.  An ultra-rigid caste system (think Brave New World) would have close to b=1.00.  Here’s what Clark found:
Two things to notice here: 70% of inherited advantage is a low estimate, and, even more troubling,  social mobility does not seem to be increasing in our lovely post-War era.  And though some of the charts above show straight lines, really remember that everything is like the second graph I showed you with a curved line with a flatter and flatter slope as it approaches 1 (those are the “fitted” lines you see on the graph–they show what we think it would look like if we had observations for ever moment, not just one at the beginning and one at the end and maybe one in the middle).  But this is depressing: in a society like Sweden’s there’s not much more mobility today in many fields than there was roughly a hundred years ago.  Clark ends the paper depressingly with:

Generalized and long-term social mobility in Sweden in recent years is much lower than the rates reported in standard two generation studies of intergenerational correlation of income or educations.  Rates of long run social mobility are indeed so low that the 18th century elite in Sweden have persisted to the present as a relatively advantaged group.  There is little evidence that intergenerational mobility rates have increased within the last 2-3 generations, compared to rates in the pre-industrial era.  The b for underlying social status may indeed be as high as 0.70-0.80.  Such mobility rates are the as we observe for underlying social status in a variety of other countries such as the UK, USA, and even India and Chile [see the appendix].

The strong intergenerational persistence of status in a country with many year of generous public provisino of opportunities and funding for education, at rates similar to other countries without such equalizing expenditures, suggests that the formation and functioning of families.  These may be forces that are impossible to change with public policy.

It’s too late at night to try to calculate exactly how long it would take for people from elite and non-elite families have exactly the same distribution in society, but it’d be possible to find a plausible b for America (notice there are several ones here, depending on the profession), find the current difference between blacks and whites in income, education, occupational prestige, whatever (things that proxy for the unmeasurable “social status”–Clark suggest though that “social status” changes more slowly than mere income or education), and plug it into the equation.  Maybe I could even find some to check my math.  But, to go back to our original question, I can confidently say without doing all of the math, even if all racism magically ended tomorrow, it would take a bunch of generations and definitely longer than 100 years for blacks and whites to actually live in an equal society.

Look at Sweden’s current distribution of income: It’s not that there are no rich Andersons, and no poor people from noble families (just a note: this is based taxable income so poor people making no income are apparently not included, so Clark thinks this underestimates the low distribution of Anderssons; also this is income, not wealth–wealth tends to actually matter more and show an even clearer pattern of accumulated advantage), it’s that distributions are roughly the same, the nobles are just moved over a a few steps towards the richer side (glance at the left and right extremes, of course).  This is like the classic metaphor of, if life is a 100m race, some people have a 10, 20, 50m head start (or, to switch sports metaphors, a lot of people said of Mitt Romney “He was born on third base and thinks he scored a home run.”)And the two curves are moving steadily closer together, evidence suggests, but they started off so far apart in the first place it’s going to be a while until they’re perfectly lined up.  That’s assuming inequality continues to decrease–which is not what we’ve seen in the U.S. recently (though this process is slightly different: this is driven by mass gains at the very, very top of the distribution).  I guess this answers why there are fun physics “What Ifs” (Yoda can power this many windmills!  Soul mates are fake!  Meteors explode!) while people don’t normally try to answer sociological what ifs–the answers would just be really depressing.

If someone (preferably someone with stronger math skills) wants to partner up and do social science what if’s, please let me know!


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: