Does Democracy Make Dictators Crazier?

By Mike Dorf

Most analyses of Kim Jong-un's recent round of aggressive statements and acts offer three non-mutually-exclusive explanations: 1) There is an internal power struggle in North Korea, in which he needs to show extreme external bellicosity in order to ensure support from the military; 2) He is trying to come across to the U.S., to South Korea and to the world as crazy in order to gain maximum leverage in future negotiations; or 3) he really is nuts.  Here I want to focus on explanation number 3), and propose a research project for an enterprising political scientist.

I begin with the following observation: Many dictators throughout history have been crazy, but lately it appears that any given dictator is more likely to be crazy than a random dictator of the past.  I have a hypothesis that would explain this phenomenon, if it turns out to be real.  The hypothesis is that once democracy catches on as the only presumptively legitimate form of government, the sort of person who is able to obtain and hold power as a dictator needs to be so ruthless that the job will select for crazy people, whereas in the past, a benevolent dictator was seen as more of a blessing.  My causal hypothesis can only get off the ground if we have verified the descriptive hypothesis.  And to test either, we would need to specify a few terms more clearly so that we can "code" various cases.

a) Who counts as a "dictator"?  I would define dictator in terms of the system of government in which he (and it's almost always a man) operates, rather than in terms of how he exercises that power, because how the dictator exercises power is part of the determination of whether he's crazy.  To be sure, it will be impossible to fully escape the endogeneity problem because dictators often change the system of government.  I'll leave the details to be worked out by the scholar who carries out the research project, but I would want the definition to say something like this: A system of government is a dictatorship if and only if a single individual has the practical (as opposed to formal) power to dictate all major rules, policies and actions of the state.   I'm not committed to that particular definition, however.  The main thing is to make sure that paradigm cases end up on the right side of the line.  The list of dictatorships would include the Roman Empire under the Caesars; many medieval European kings; Communist heads of state like Lenin, Stalin, Mao,  and Pol Pot; and Nazi Germany.

b) What counts as crazy?  There are easy cases, like Caligula (crazy), Idi Amin (crazy), and Charlemagne (not crazy).  But what about dictators who were undoubtedly ruthless but appeared to harness their ruthlessness in rational ways?  Many conquerors would appear to fall into this intermediate category: Alexander the Great; Genghis Khan; and Napoleon are examples.  Here again, I leave the harder definitional question for whoever might be bold enough to undertake this research program.

c) How do we break down history into periods?  Because my hypothesis is that dictators get crazier as democracy spreads, we need an independent variable for democracy.  Do Athenian democracy and Rome in the Senate period count?  If I were designing the study, I think I would just look at slices of time (with periods getting shorter as we get closer to the present, as the human population increases and our records vastly improve).  Then, if the data show a correlation with ancient periods of democracy, we can decide whether they count.

Bonus: A related research project would try to answer the question whether you need to be crazy to become a dictator or dictatorial powers make people crazy (because absolute power corrupts absolutely).  We can think of examples of each but it would be interesting to know the proportions.

So, if you are a political scientist looking for a research project, let me just say this: You're welcome.

Postscript: No, this is not a one-week-late April Fool's joke.