Karthik Tadepalli

Economics PhD student at UC Berkeley

How to Win Elections and Oppress People

Published July 15, 2020


A pessimistic view of politics is that politicians are evil people wreaking havoc on the lives of ordinary people for their own rapacious benefit. According to Bruce Bueno de Mesquita and Alastair Smith, the truth is even more pessimistic than that: politicians are ordinary people wreaking havoc on the lives of ordinary people for their own survival. The gloomy truth is that all governments, democratic or authoritarian, are subject to the same logic of political survival: reward your supporters or they will depose you.


The Dictator’s Handbook expounds a political Theory of Everything known as selectorate theory. It posits that all political action is taken by leaders to preserve their own power. Furthermore, political power comes from three concentric circles:

  1. Interchangeables (the nominal selectorate) - all people who could potentially support a leader, e.g. all adult citizens of America.
  2. Influentials (the real selectorate) - the subset of interchangeables who actually exercise influence in supporting a leader, e.g. all Americans who vote in a given election.
  3. Essentials (the winning coalition) - the subset of influentials that a leader absolutely needs to stay in power, e.g. the set of Americans across different states whose votes will carry a leader to 270 electoral college votes.

Leaders stay in power by maintaining a winning coalition of supporters. They keep their winning coalition loyal with rewards like money, favors, and positions of influence. But rewarding followers is expensive: the corollary is that leaders try to maintain a minimal winning coalition - the smallest coalition they can have and still keep power, balancing their desire for power with their limited capacity to reward.

Notice that this categorization is not specific to democracies or autocracies. As BdM&S point out, “democracy” and “authoritarianism” aren’t really defining features of countries - they are simply consequences of different coalition makeups. Consider an equivalent characterization of the selectorate, now applied to China:

  1. Interchangeables - all Communist Party members
  2. Influentials - the CPC Politburo
  3. Essentials - the Politburo Standing Committee

The difference between China and America, according to BdM&S, is that a minimal winning coalition in America has more than 60 million people, whereas the MWC in China could have just 7 people. What we call “democracy” is just a political system where the MWC is very large, whereas “authoritarianism” is a political system where the MWC is comparatively small.

The size of a MWC matters because 7 people can easily be bribed, but 60 million people cannot be. In other words, an American presidential candidate would love to win by simply buying off 60 million voters, but this just isn’t feasible for them. The only way to win over 60 million people is to provide public goods that benefit them, like roads they can use or hospitals they can go to. Thus, democratic leaders provide public goods to their people, not because they are inherently more benevolent but because they cannot maintain a MWC without good governance. Furthermore, this creates a race to the top: opposition parties gain a MWC by having better ideas and more effective governance, leading democracies to have the celebrated governance that democrats tout proudly in their reelection campaigns.

This theory neatly explains the middle of the democracy-autocracy spectrum, a collection of weak democracies and benevolent dictatorships: India, Brazil, Singapore, China seem to be defined by more than just whether they are superficially democracies or dictatorships. It explains why institutional corruption is higher when voter turnout is lower (smaller selectorate means higher returns to corruption), or why political parties in weak democracies mix legitimate campaigning with outright bribery (e.g. distributing food before elections).


The feature that differentiates selectorate theory from other political theories is the principle that power comes from influential supporters. Once you hear this, it becomes so obvious that you can’t imagine having ever thought differently. But if that’s too obvious for you to consider it an insight, BdM&S draw interesting implications from the selectorate principle.

The first insight is that if your essentials give you power, they can take it away by supporting your political rival (c.f. every military coup in history, or revolutions that the military refused to suppress). The only way to prevent this is to give them better rewards than your rival can credibly promise: seizing more resources from the people and redistributing them to your supporters. The flip side of this carrot is the stick: if the consequences of not being in your coalition are dire, then even the greediest influential will accept a modest reward for being in it. BdM&S put the implications more bluntly:

People outside your coalition are irrelevant at best, and threats to your authority at worst. Rewarding them over your own essentials is a path to losing those essentials, and thus losing power.

In the past, the dominant People’s Action Party in Singapore has punished districts that didn’t vote for the PAP by withholding housing improvements from them. If you think that kind of suppression wouldn’t happen in a “real” democracy, then I have a bridge to sell you.

The second implication of selectorate theory is that if your coalition is too large, the rewards you can give them will be spread too thin to retain loyalty. Therefore, you as a leader don’t just want a MWC, you want to make your MWC as small as possible at a systemic level. The easiest way to do that is to minimize the set of influentials. This can happen in authoritarian regimes through busting powerful oligarchs or “anti-corruption” campaigns to remove bureaucrats. And of course, democratic leaders also go to extreme lengths to minimize the set of influentials.


TDH is a two-idea book. The first idea is that political leaders stay in power by maintaining the support of a winning coalition, and differences in winning coalition size/composition explain the democracy-autocracy spectrum. This is the most interesting claim, and it’s pretty convincingly argued. It’s a relief to read a theory of political behavior that focuses on incentives, as well as identifies the actors who make decisions (the leader level, rather than the country/party level). That’s the idea that I’ve focused on here, and that I’ve taken away most strongly from this book.

But the second idea, much more implicit, is that politics is largely opportunistic. The influentials want to be paid in material rewards, and the leader wants only to stay in power. In other words, political actors—both leaders and influentials—are non-ideological. This implicit claim, which is deeply, deeply underanalyzed. It’s claimed mostly by implicit example, with cases of governments rewarding supporters with money and leaders flip-flopping on ideology to stay in power.

This may make sense when all the examples of influentials are superficial examinations of bureaucrats and oligarchs. But you know who else could be an important part of the selectorate? Religious leaders. Islamic empires like the Safavids and the Mughals had powerful ulema, councils of Islamic scholars who legitimized rulers. The Holy Roman Empire relied heavily on the Catholic Church. Even today, religious leaders are intertwined with politics in countries from America to India to Iran to Ireland. This theory is blunted severely when considering ideological influentials like religious leaders.

But it’s worse than that, because it’s not just specialized types of influentials who can be ideological. Why would generals and bureaucrats care only about monetary rewards? For example, British colonial rule in India created a bureaucracy that was largely English-educated. I would imagine that after Indian independence, this bureaucracy held more Western-friendly attitudes. At the same time, many soldiers in the Indian army had been forced to die for Britain in World War 2 despite their ongoing independence struggle, which could make Indian generals hate the West. How might this clash influence the Indian government’s diplomatic relations with the West?

But according to BdM&S, these generals, bureaucrats and religious leaders don’t care at all about their stated beliefs or past experiences, and are simply maximizing private utilities that happen to be denominated in dollars.

If we have learned anything… it is to be suspicious of people’s motives. Appeals to ideological principles and rights are generally a cover. J.P. Morgan had it right: there is always some principled way to defend any position, especially one’s own interests.

It is trivial to say that people can lie about their motivations, and as an economist I am amenable to private utility maximization. But utility isn’t money! It’s just a representation of people’s preferences, which can value many different things. And what is BdM&S’s conception of “one’s own interests?” Power and money, then more power and money.

There is nothing wrong with money as a motivation in general, but it falls apart pretty quickly in context. The US would happily give Kim Jong-un an island in the Bahamas if it made him stop playing frisbee with his nuclear controls and abdicate instead. Money can’t explain the “interests” of many leaders, although of course many leaders are simply rapacious thieves. Even power doesn’t square the circle, because why do leaders want power? Either for money (already addressed) or to advance some state goals - but then you haven’t rid your theory of ideology!

I think this baffling omission reflects something broader about the book and the flip side of its simplicity. The Dictator’s Handbook takes its title way too seriously, leading to cringe-inducing sections like:

Bravo to Senior General Than Shwe of Myanmar, who made sure following the 2008 Nargis cyclone that food relief was controlled and sold on the black market by his military supporters rather than letting aid go to the peopleā€”at least 138,000 and maybe as many as 500,000 of whom died in the disaster.

BdM&S devised a pretty interesting theory of political behavior. But then for some reason they decided to crank the shock factor to up to a hundred, and brand their book as the One Weird Trick to Make All Your Political Rivals Hate You. “Maybe if we call it a dictator’s handbook and repeatedly comment on how you should steal and oppress the life out of your people, then people will read it!” Evidently, it attracts attention: CGP Grey’s explainer of the theory got over 8 million views, and sparked responses, reaction videos, and even a parody. This is no typical reception to a political theory!

The result of this Buzzfeedization is a theory that explains quite a lot about politics, but leaves a lot more untouched in favor of keeping it absurdly simple.


So on both sides of the coin, simplicity is the defining feature of The Dictator’s Handbook. Selectorate theory is compelling in its simplicity and appeal to incentives, but simplistic enough to raise eyebrows. I enjoyed it a lot, but you could probably watch CGP Grey’s explainer and lose almost none of the value.