The Tragedy of Political Science

If it's popular, it probably sucks.

A few things that are true about political science:

1. Good theories are concise.

2. Concise theories are simplified models of reality—they are abstract.

3. The more abstract a theory, the less likely it is to accurately describe any particular case in the real world.

4. The more you want your theory to explain, the more abstract and/or less concise it has to be.

We have a trilemma here. Pick two: concision, accuracy, explanatory breadth. You can have a concise theory that works really well, but it’s only going to explain some very specific phenomena. You can have a theory that explains a lot of things while only making a few assumptions, but it’s going to commit a lot of errors. You can have a theory that explains a lot of things, and gets most of them right, but it’s just going to be a description of the world and won’t teach us much that we can’t learn through direct observation.

A few more things that are true about political science:

5. The theories with the greatest policy relevance are concise and accurate. They explain things very well within a narrow policy domain and don’t waste the time of policymakers and their staff.

6. Academics prefer theories that are concise and either accurate or broad. Today, most political scientists work on middle-range theories, which privilege accuracy over breadth, since the grand theoretic debates of the 1970s through 1990s are widely considered obsolete.

7. The preferred theories in the mass media are highly concise and broad. A successful pop political scientist is someone who can make their case to a lay audience in a single news segment and speak with authority about lots of issues.¹

Now, here’s where it gets concerning:

8. The mass media want commentators to make policy prescriptions. Academics who commentate want to exert policy influence.

9. When an academic’s theory isn’t very accurate, the academic can just make up their analysis or policy recommendations and the lay public will be none the wiser.

I suspect that a lot of these things are true in the other social sciences as well. If so, then we should expect the most influential ideas from social science to (1) be unrepresentative of what’s most influential within the disciplines, and (2) make claims that are poorly supported by science. I don’t know what goes on in the other disciplines, but as Paul Musgrave has pointed out, many of the most memetic political science ideas of the past 30 years have been dangerous nonsense.²

Take Samuel Huntington’s “Clash of Civilizations” argument, which predicted that conflict after the Cold War would break down along cultural lines:

Huntington’s thesis was not a conjecture based on careful empirical study—it was a speculation looking forward based on some cherry-picked contemporaneous examples. Many academic articles that sought to rebut Huntington by testing his hypothesis fell into this trap, attempting to show him wrong with sometimes quite impressive tests. But Huntington could not be disproved by mere facts. His idea was primed to thrive in the wild, free from the confines of empirical reality.

Or the democratic peace, which observes that democracies rarely go to war with each other:

A long debate within political science concerns why this correlation might hold. International relations graduate students studying for comprehensive exams have to keep straight numerous subdebates: whether the causes of the peace stem from the incentives of democracy for leaders or the deep normative underpinnings of liberalism; whether the real cause is capitalism and prospects for trade instead; whether political scientists have cooked the books by redefining U.S. adversaries as nondemocratic even when they have had representative governments; and how methods and measurements confirm or complicate the story.

[…]

As the democratic peace concept raced away from serious and conflicting academic debates, it simplified and evolved. […] In a new, more transmissible form, the democratic peace became part of the justification for the 2003 invasion of Iraq. A new variant emerged in neoconservative circles: If democratization yielded a more peaceful world, then it naturally followed that promoting democracy was a means to democratization. For muscular conservatives of the Bush administration, the implication was obvious: The Middle East thus needed to be forcibly democratized. Secretary of State Condoleezza Rice—who holds a Ph.D. in political science—argued that the democratic peace, and even forcible democracy promotion, was thus the “only realistic response to our present challenges.”

Look closely and you’ll notice that Huntington’s Clash of Civilizations and the D.C. Beltway, funhouse mirror version of the democratic peace have at least two things in common: Both are simple enough for a lay person to understand, and both claim that everything worth knowing about international politics comes down to a single variable. Concise and broad, but not accurate.

1. Noam Chomsky (not a political scientist) has complained that concision is required of commentators in the mass media. You might think Chomsky lacks concision because he doesn’t believe in social science and opts instead for broadness and accuracy:

The very idea of a ‘theory’ in the social sciences is therefore highly suspect in Chomsky’s politics and must be critically examined in the light of power dynamics within the academy, as well as the status anxiety of social scientists relative to natural ones. Philosophically, Chomsky is a rationalist as opposed to an empiricist; he does not gather and manipulate empirical facts using experimental methods in order to arrive at general conclusions (although his work is grounded in a wealth of empirical data and exposes the weak empirical foundations of much IR theory), but assumes that empirical phenomenon in the social world are the result of an order of relationships or principles that may be hidden but are substantially ‘given’ and will become transparent once brought to light. “In the analysis of social and political issues it is sufficient to face the facts and to be willing to follow a rational line of argument,” Chomsky declares. “Only Cartesian common sense, which is quite evenly distributed, is needed […] if by that you understand the willingness to look at the facts with an open mind, to put simple assumptions to the test, and to pursue an argument to its conclusion.”

2. Ironically, the one popular political science idea from after the Cold War that actually did get it right, Francis Fukuyama’s “End of History” thesis, is also the one that gets the most hate.