Consider a wide definition of a conspiracy as a secret plan between two parties or more to do something, legal or illegal, that hurts somebody else’s interest for the purpose of furthering their own mutual interests. This definition is close to the dictionary’s (“A secret plan by a group to do something unlawful or harmful”), although it focuses less on secrecy. For instance, a firm is a general conspiracy to wage competition against other firms’ interests and it manages small conspiracies, only some of which are secret. Government actors are the masters of conspiracies.

The problem is to distinguish good from bad conspiracies, and plausible from implausible conspiracies (or “conspiracy theories,” as specific conspiracy claims are often called). A bit of economics is especially useful to make the latter distinction.

To be engaged in by an individual, a conspiracy must promise a good likelihood of higher benefits than costs for the individual himself. If this is not true for the minimum number of conspirators required, the conspiracy will not happen. Focusing on the costs, the difference between a plausible and an implausible conspiracy is often determined by whether it is legal or illegal since an illegal conspiracy that is uncovered brings severe punishments, that is, higher costs.

There are conspiracies in any spontaneous order, in any system of individual relations without a central and effective coercive authority. Subgroups of individuals always engage in more or less secret plans to further their own interests. In the case of illegal conspiracies, the higher the expected punishment (cost of punishment times the probability it will be metered), the less frequently they will happen.

Some spontaneous order also exists in any conspiracy. Conspirators don’t necessarily act in a way that is consistent with the conspiracy. In a firm, individuals sacrifice more or less the company’s interest to further their own interests. In an army, the foot soldiers at the end of the chain of command don’t do exactly what the faraway general intended them to do (see Gordon Tullock, Bureaucracy, in The Selected Works of Gordon Tullock, edited by Charles K. Rowley, Liberty Fund, 2005).

There is some spontaneous order in illegal conspiracies. Spontaneous rules developed within groups of pirates to help further their objectives (see the work of Peter Leeson). But in some circumstances, the self-interested behavior of individual conspirators leads them to act against the conspiracy. A conspirator will rat on his co-conspirators if he thinks that the conspiracy is on the verge of being unmasked and that he can thereby reduce his own punishment. And he is even more motivated to spill the beans as he knows that his co-conspirators are equally motivated. If a co-conspirator is going to denounce me, I better denounce him first. The famous Prisoner Dilemma models such interactions.

The more numerous the conspirators need to be and the longer and more complex their plan, the less likely each one will participate, because he knows he can be betrayed by any of the others. The less likely it is that such a conspiracy will materialize.

Conspiracy hoaxes are implausible conspiracies precisely because they ignore these individual incentives. Given the decentralization and complexity of the American election system, a large election fraud—one that has some chance of changing the election result—would need much complicity, including from government actors. The government conspirators’ individual benefits in terms of future power, perks, and money can be large, but their costs, if the conspiracy is uncovered, will also be large. In a country with severe penalties for interfering with voting or tampering with ballot boxes, with a host of independent investigating authorities and judges, with a free press and a pack of Pulitzer-chasing journalists, the cost will likely be too high for any individual to engage in a vast electoral fraud.

A conspiracy to “steal” the recent presidential election is not completely impossible, but it is highly implausible. It is not more incentive-compatible than, say, Pizzagate (an imagined conspiracy of Democratic pedophiles in a DC pizzeria) or than the claim, pushed by Alex Jones, that the Sandy Hook murders did not really happen. If my argument is not true, the difference between the United States and banana republics would be blurred.

These explanations by incentives overlap Ockham’s razor: in explaining election results, choose the theory that is the simplest—although determining what is simpler is not always simple.

If the foregoing is true, we would expect that most if not all election conspiracies would be run by foreigners under the umbrella of their own governments. In such cases, the individual cost of being unmasked is relatively low and may well be lower than the benefits conferred by the government. That the rulers of Russia would mount a conspiracy to influence the result of an American election, especially to favor a narcissist and potential puppet, is certainly incentive-compatible, although it is not in itself proof of the conspiracy.

There is a demand for conspiracy theories, as co-blogger Scott Sumner recently argued and, in response, there is a supply of them, a supply that has dramatically increased with the reduction in their production (including diffusion) cost. In this post, I looked at the supply side.

An extreme example of conspiracy theory is the one peddled by the president of an organization called Judicial Watch: “elite units of the National Guard” fed the ballots of several states into a quantum computer (nothing less!) and were also able, with a GPS technology and the ballots’ “ink made of corn,” to follow their travel; the conclusion is that Trump won reelection with “over 80%” of the legal vote (see In that particular conspiracy theory, everything is nonsense, from the beginning to the end. The criminal actions of the guardsmen and their superiors and accomplice are not incentive-compatible. Time is scarce: don’t waste yours with that sort of conspiracy theory.