Wednesday, September 21, 2016

The Government Made Me Do It

Welcome friends!

I read an interesting article the other day about what I think must be one of the central mysteries of our time.  Why are undereducated underemployed economically struggling older rural while people so attracted to spoiled self-obsessed fabulously wealthy jet setting big city liar and all around shyster Donald Trump?  Talk about your strange bedfellows.  I know a lot has been made of Trump’s efforts to tap into every low sentiment one can find in small town America: racism, nationalism, nativism, religious intolerance, misogyny, hatred of sexual minorities, and so on.  I get that part.  He has clearly tapped into a previously unused political resource.  But it’s not like he’s some kind of political genius.  We all knew such deploring sentiments were out there.  It’s just that no one else was sufficiently disreputable to go there.  But is there something else this segment of conservatives finds appealing about the man?  Something a bit more positive perhaps?  Something people who don’t live in Oklahoma or Texas or other such places might hope to understand?  Well, much to my surprise I’m starting to think there may be.

The article I was reading discussed an apparently common attitude among such people that the US government is the source of all that ails them.  Yes indeed.  They don’t attribute their financial difficulties on anything you and I might think about: technology, industrial organization, trade, demand, unemployment, commodity bubbles, and so on... Sorry but only selected archived (previous year) posts are currently available full text on this website.  All posts including this one are available in my annual anthology ebook series available at the Amazon Kindle Bookstore for a nominal fee.  Hey, we all need to make a buck somehow, right?  If you find my timeless jewels of wisdom amusing or perhaps even amusingly irritating throw me a bone now and then.  Thank you my friends!

Thursday, September 8, 2016

Profiling and the Ethics of the “Free Market”

Welcome friends!

I was just reading an article on what sounds like a rather interesting book with the whimsical title Weapons of Math Destruction by Cathy O’Neil.  It got me thinking not only about racial profiling but also and more generally about the ethics of unregulated markets so I thought I’d say a few words about that this week.  I suppose I may be neglecting the antics and inanities of conservative presidential candidate Trumpo the Clown, who apparently is now tied with Hilary Clinton in the polls.  Remarkable in a way but just nothing very interesting.  I suppose it’s entertaining in the way reality TV is entertaining but it certainly presents one with precious little to think about.  It’s not like the man represents anything in particular.  He doesn’t really have a serious platform as far as I can tell.  Just some things he sometimes says and sometimes denies plus a lot of pouting and posturing.  Actually I suppose it’s more like a cheap horror movie.  I see stupid people.  Wait, that was actually a pretty good movie.  Well, you get the idea.  How about this one?  Only two things scare me!  One is nuclear war.  What’s the other you ask?  Stupid people!

Anyway getting back to my topic for today the subject of Ms. O’Neil’s book is the way certain formal or informal mathematical algorithms and models can have a rather large effect on one’s life even though one may not realize they even exist or know how they work or might find their use morally repugnant.  She’s talking about things like the models some banks use to set rates for loans that often include all manner of considerations such as one’s neighborhood, the models some employers use to determine suitable candidates for promotion based on personality tests and various other quantifiable characteristics, and the models some judicial systems apparently use to determine appropriate prison sentencing based on various characteristics of convicted perpetrators including not only prior convictions and previous police encounters but again their neighborhoods and the criminal records of family and friends.  Ms. O’Neil finds these common uses of statistical algorithms and models unfair and believes they perpetuate poverty along racial lines because they stack the deck against people living in certain neighborhoods.  Of course, she notes these algorithms can theoretically be used for good as well as bad so we can’t really fault mathematics or model building in general.  No, the problem lies in how we choose to use them. Basically the same situation we have with technology and other artifacts of human ingenuity.  We might have a hundred hate filled carnivals of stupidity in the conservative blogosphere but then we have my own blog in the liberal blogosphere like a shining beacon of light in the darkness.  The problem isn’t the existence of blogs; it’s what we make of them.

What we’re talking about here seems to me to be simply another form of the “profiling” we discussed previously in a law enforcement context.  (See my post Race and Bad Cops from July 14, 2016.)  The fascinating thing about this issue to me is that as Ms. O’Neil points out the issue is not that the statistical inferences are wrong.  I assume they’re probably true.  If for example we’re just talking about banks making up numbers and willy nilly charging people in some neighborhoods unnecessarily high rates for their loans one wonders why some other banks or other financial institutions don’t find it worthwhile to enter those markets and undercut those banks.  The interesting issues occur when the inferences are spot on but we just don’t like when people use them.

Now I should say having had the good sense to be born with a suitably pale complexion and having never travelled to the Dark Continent or spent much time in the various other darker hued portions of the world I’ve never really faced racial profiling per se but I’ve certainly faced the general concept in other contexts.   One example off the top of my head is that I recall as a young man who had finally attained whatever the age of majority was at that time, 18 years I suppose, I was legally entitled to enjoy a nice beer with my pizza but after a year or two of gastronomic bliss some policy wonk discovered adults of my age were statistically more likely to be involved in alcohol related car accidents than were older people and I was unceremoniously dumped back into the category of underage drinker.  Had there been a  problem with my drinking and driving in particular?  No, not at all.  The problem was I had something in common with some people who did, specifically that we shared the same age, and I paid the legal price for that unfortunate chronological affinity.  It was profiling of course but in a guise that passed largely without comment in the social discourse of the time.  And the interesting thing is without a crystal ball or some other special technology to assess my individual drinking and driving habits the statistical inferences were assuredly correct.  Based solely on my age I’m sure I was more likely to have been involved in an alcohol related accident than some other people.  I’m sure I can find similar arguments that could be used against me today but I suppose that’s one of the benefits of age.  Try that sort of thing on me now and one might just find oneself falling into the very same category one day.  You know, I’m sure everyone is better off not drinking but if we’re going to start prohibiting people from doing things I’d much rather it was you than me.

So if the statistical inferences are correct what’s the problem with profiling anyway?  Well, the problem is clearly it contradicts the common ethical principles of individual responsibility and innocent until proven guilty.  As a young man I didn’t appreciate losing my ability to have a glass of beer because some other young idiot drove his or her car into a tree.  Along similar lines I doubt any of us feel very comfortable penalizing someone who scrupulously pays his or her bills on time because he or she has the misfortune of living in a neighborhood with a bunch of deadbeats.  The information may be fine.  The statistics may be correct.  We just don’t like using information that way.  I basically feel about profiling the same as I did when I wrote about it in a law enforcement context.  Haven’t changed my opinion at all.  I have no problem with profiling per se if what we’re talking about is using whatever information we have to stop crime.  The police should pay more attention to people who are statistically more likely to commit crimes for whatever reason.  However, that doesn’t extend to police actions that negatively affect innocent people who fall into that category.  I don’t support police stopping people or harassing people or behaving in any way differently toward people based on their statistical likelihood of being a criminal.  Innocent until proven guilty.  That’s my motto.  But if you want to investigate a robbery it probably makes sense to start with the unemployed young man with the gun in his car before moving on to grandpa sitting on the porch swing reading the evening news.  We can’t banish common sense but we can prohibit treating people like criminals when we know they might not be and in fact most likely are not.

This time out since we’re discussing these issues in an economic context as well it occurs to me the business about the loan rates illustrates a general point I’ve tried to make about so-called free markets in many other contexts.  Markets do what they do.  If there’s information relevant to making a buck then businesses in an unregulated market will use it.  Normally that’s fine but in some funny cases like the one we’re discussing right now it’s not.  Markets aren’t magic.  There’s no invisible hand making everything come out right.  If one doesn’t see any ethical issues with what’s happening in a particular market then great; no need to do anything.  If one does see some ethical issues then one will want to get in there and regulate or do something to make it come out in a way one finds more ethically appealing.  It’s the same general sort of issue I usually discuss in the context of distributional issues.  The market will certainly distribute good and services.  Don’t worry about that.  But what matters for the market doesn’t necessarily correspond or anyway fully correspond to what one may feel ought to matter.  I guess my point is only a fool bases one’s ethics on what happens in a market rather than assessing what happens in a market using one’s ethics.  I know we all have different ethical opinions and so on but can we at least agree we’re going to discuss ethics and not get sidetracked into supposing the issue involves some technical aspect of economic theory?  How about this plan?  Why don’t you jot down what you find ethically appealing or repugnant as the case may be about how our labor market and other relevant markets such as the stock market and so on distribute wealth in this country?  I’ll do the same and discuss it in a future post.  We can compare notes.  It’ll be fun.  Let’s see how far apart we really are on these issues.  Plus I won’t have to talk about why economic theory is not a suitable basis for a coherent theory of social ethics yet again.  Happy days!

References

Math is racist: How data is driving inequality.  Aimee Rawlins.  CNN Money.  September 6, 2016.  http://money.cnn.com/2016/09/06/technology/weapons-of-math-destruction/index.html?iid=TL_Popular.