Wednesday, March 27, 2013

Demography and the Limitations of Empirics

I won't mention his name (as he gets called out far too much on this blog as it is), but a certain semi-prominent libertarian defector has left a partially puzzling response to a commenter in regards to a post on gay-marriage. The blogger criticizes the move towards changing institutions as fundamental as marriage. The commenter replies that maybe such change is part of a longer, slower general evolution of the institution; ie, not brought on the hastened wings of recklessness but rather a more lengthy process of osmosis. The blogger responds with two points. It is the second one that I find at least mildly baffling. I'll paraphrase here:

"It seems to me that the societies that embrace this institutional change the most have suffered for it. Look at Western Europe's demography. Their population has been dwindling for some time now. That doesn't seem like an insignificant consequence to me!"

I think statements like this embody the raison d'etre for schools of thought similar to the camp of Austrian economics. It illuminates the limitations of empirics generally in such matters. As the blogger in question is a defector of such a camp (Austrians), I can't help but be sent reeling at such comments. Let's look at this statement on a couple of different levels.

The first presumption, to my mind, lies apart from the more general truth-value of the statement, which hinges on the connection between this particular set of social institutions and demography. The first premise is unspoken - it is, rather, smuggled in. That premise being that declining demography is, as a matter of fact, bad or at the very least undesired. Now, it's true enough that, given the structure of government(s) across Europe, declining demography will pose various problems regarding their structures of taxation and benefits - even though they aren't demonstrably insurmountable. That's certainly a good argument for maintaining positive population growth. On the other hand, there are certainly (strong) arguments for negative growth; namely growing limitations on material resources. I'm not particularly endorsing either of these views. Instead, my point is simply that asserting that negative population growth is bad assumes a certain optimum population level. And, given that this discussion has not been had nor its conclusions justified, we would probably do well to give those particular justifications if we're going to treat it as a given.

The second way in which the blogger's statement disturbs me is the more obvious one - what is the nature of the connection between such institutional changes and demography? Is there an actual co-variance  or is it mere correlation? We're given no particular justifications. It's precisely this kind of political proposition that drives the Austrians mad in the realm of economic analysis.

What we have here, generally, is a mental mapping of two trends that happen to currently enjoy a similar trajectory. But justifying their connection is a matter that's beyond the empirical nature of those trajectories. The data must be interpreted. Even though, as humans, we are inclined to see causal relationships and patterns in everything, that does not mean that all things that appear to have connection actually do. If we can discern the nature of things through a tandem use of empirics and theory, we can avoid both the accusations of scientism and anti-science.

To peel all this back a little bit, let's look at the thrust of that statement again. The context is gay-marriage. A connection is implied between the push for such changes and declining demography. But how does this make sense in the context of gay-marriage itself? The implication seems to be that if this trend reversed direction we would see more positive demographic growth. But how could this be the case given that we are talking about individuals who identify themselves as being homosexual? Do homosexuals, en masse, suddenly adopt heterosexual behavior when marriage isn't an option for them?

What could make such an assertion make sense? Are we envisioning going back to a state of affairs that's so oppressive to homosexuals that we will socially brow-beat them into heterosexuality? Surely that's not what our author quite has in mind. Maybe our author believes that sexual orientation is largely a matter of personal choice, and thus such legal discrimination will provide the incentive to make the "better" choice? At least that justification is a somewhat plausible one on its face. But of course that assumes the highly contested (and I'd say, at this point, refuted) idea that gender-orientation or sexual orientation is not biologically driven. I'm not saying there aren't good arguments for such ideas. But, again, I don't think it's something we can just presume either.

Of course, after all is said and done, our blogger could have perfectly defendable justifications for his presumptions - even if they are are also perfectly wrong. But the larger point here is that things that seem like straight-forward motions toward empirics need to be dissected and questioned. It happens far too often that someone will throw a statistic out in the course of an argument and then it is merely glossed over in subsequent argument. But ALL empirics require substantive theory. And if the logic underlying the empirical implications are not sound, then what are we to say of the conclusions drawn from them? I would hasten to say, "Not much."