A fascinating paper, Why Do Voters Dismantle Checks and Balances? by Daron Acemoglu, James A. Robinson and Ragnar Torvik (h/t FE). There is a pile of maths in there, but you don’t really need it and I only skimmed it. From the conclusions:
In many weakly-institutionalized democracies, particularly in Latin America, voters have recently dismantled constitutional checks and balances that are commonly thought to limit presidential rents and abuses of power. In this paper, we develop an equilibrium model of checks and balances in which voters may vote for the removal of such constraints on presidential power. Our main argument is simple: checks and balances are indeed effective (at least partially) in reducing presidential discretion and prevent policies that are not in line with the interests of the majority of the citizens. This naturally reduces presidential rents, which is however a double-edged sword. By reducing presidential rents, checks and balances make it cheaper to bribe or influence politicians through non-electoral means such as lobbying and bribes. In weakly-institutionalized polities where such non-electoral influences, particularly by the better organized elite, are a major concern, voters may prefer a political system without checks and balances as a way of insulating politicians from these influences. In consequence, voters may dismantle checks and balances, implicitly accepting a certain amount of politician rent or politicians’ pet policies that they do not like, in order to ensure redistribution when they believe that the rich elite can influence politics through non-electoral means.
They provide examples from Venezuela and Ecuador to illustrate their ideas. The basic idea, restated, is that the poor get to elect the pols, but the rich get to bribe them, so the poor may prefer pols too rich to bribe.
Note that their model is a steady-state one; it includes nothing about effects such as the country falling apart as megalomaniac Prez’s ruin the economy. Nonetheless, that process is slow enough, and deltas from the current state are what count, so voters might, in a prisoners-dilemma sort of way, choose to ignore that problem.
This has eerie echoes of Hobbes (which the paper fails to cite, tsk tsk, young folk nowadays):
The difference between these three kinds of Commonwealth consisteth not in the difference of power, but in the difference of convenience or aptitude to produce the peace and security of the people; for which end they were instituted. And to compare monarchy with the other two, we may observe: first, that whosoever beareth the person of the people, or is one of that assembly that bears it, beareth also his own natural person. And though he be careful in his politic person to procure the common interest, yet he is more, or no less, careful to procure the private good of himself, his family, kindred and friends; and for the most part, if the public interest chance to cross the private, he prefers the private: for the passions of men are commonly more potent than their reason. From whence it follows that where the public and private interest are most closely united, there is the public most advanced. Now in monarchy the private interest is the same with the public. The riches, power, and honour of a monarch arise only from the riches, strength, and reputation of his subjects. For no king can be rich, nor glorious, nor secure, whose subjects are either poor, or contemptible, or too weak through want, or dissension, to maintain a war against their enemies; whereas in a democracy, or aristocracy, the public prosperity confers not so much to the private fortune of one that is corrupt, or ambitious, as doth many times a perfidious advice, a treacherous action, or a civil war.
In fact Hobbes has a somewhat different intent there, so I’m being unfair to the authors. He’s explaining why Monarchy (effectively, Presidential rule is Monarchy) is better than other forms of government. Elsewhere, he does say that if the pols are corrupt, you’re better off with only one of them. I don’t think he ever mentions the too-rich-to-bribe argument, though.