“Social Science” in Times of Change

This is a fascinating tweetstorm by a bright political scientist, as is this retort by a famous ex blogger.

Quite frankly, I am not surprised that political science, along with the media, political elites, etc. are willfully blind to the Trump phenomenon:  they are not designed to learn new things in a time of rapid changes efficaciously.  I do not mean anything disparaging by this:  scientific enterprise is, as I keep repeating, an inherently conservative process, where changes are achieved incrementally by convincing illustration, built on backs of sufficient data, supporting modifications to existing theories.  In absence of sufficient data, old theories win out.  By nature, new, transient, and rare phenomena lack sufficient data to be convincing.  Theories in social science are not sufficiently precise so as to make predictions with very small expected variance.  That means that a lot of old theories persist as “accepted conventional wisdom” only because we lack enough data to displace them.

This is a dangerous proposition for social sciences to be in.  Krugman famously likened the progress of social sciences to mapmaking:  as the standards of proper cartography increases, maps become increasingly empty as the alleged features in terra incognito fail to meet the standards required by mapmakers.  For those staying around Genoa, say, perhaps a very precise map covering only Genoa is useful.  For those who are sailing around the southern tip of Africa, a map showing fanciful creatures along with the Cape of Good Hope might be better than nothing.  But an empty map is not empty–as per the Computer Science koan that I stole and posted the other day, an uninformed prior, too, is still a prior.  In the case of social sciences, in particularly,  we are actively giving the prescription that terrain around the Cape is just like the terrain around Liguria, which, even if we didn’t know what things are really like in South Atlantic, we should know is utterly false.

Since 1990s, ironically, since political science has started getting embroiled in “do parties matter” or “what makes parties ‘strong'” debates, there has been very little attempt at defining what a “strong” parties that “matters” is supposed to be doing.  We “knew” what a strong party is supposed to be:  a party that has some well-defined goals and goes about systematically achieving them, by strong arming those whom it can and rigging institutions as necessary.  We “knew” this had to be the case because that’s why we imagined a strong party should be like.  This is reminiscent of the Atlantic article that criticized many “theories of mind” as being worse than wrong pseudosciences:  we are concocting theories that “feel” right to our normative ideas (even if they are not exactly “political” in the usual sense) and willfully refusing to approach them from an alternate direction.  As much as Nyhan bemoans the inability of people to consider how unstable and vulnerable political institutions are, we also know that there was as much incentive to examine such potential problems as there was for vulnerabilities in the financial sector:  most of the time, they work fine, and besides, people are profiting (in multiple senses) handsomely from their normal operations.  Proposing that the insiders, both of the “politics as usual” and of the financial institutions, forego these profits for merely “theoretical objections” that the lunatic socialist loafing about at the Cafe Central might one day lead a revolution to topple a might empire is worse than silly–it is dangerous to your career.

Yes, Trump is destroying the Republican Party.  Perhaps Trump is as dangerous a man as many think he is, or perhaps he is not.  Regardless, the entire episode should be been foreseen–there was enough to theorize how it might come about, simply by extending the “conventional theories” of why parties are so “strong” and by thinking through the consequences.  (A useful analogy would be the “crumple zone” in automotive safety–too strong a chassis is conducive to greater, not lesser danger to the passengers, while a chassis that is designed to collapse in a certain fashion is far safer, coupled with a forest fire analogy–the stress from too brittle institutions accumulate over time, rather than shatter all at once in reaction to a single accident.)  I will confess to feeling a certain degree of schadenfreude that the mainstream is so confused by this phenomenon.


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 )

Twitter picture

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

Facebook photo

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

Google+ photo

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

Connecting to %s