A hell of an incentive for murder: a rationalist’s guide to suicide

Mike Pottenger


October 28th, 2012

A morbid example of when incentives produce gruesome behaviour.

Republished from The Conversation
By Mike Pottenger, University of Melbourne

The Holy Roman Empire, 1704: Agnes Catherina Schickin slits the throat of a seven-year-old boy.

1746: Johanna Martauschin smashes the skull of a small child.

1753: Sophia Charlotte Krügerin cuts the throat of a nine-year-old boy.

1761: Eva Lizlfelnerin steals a baby and throws it into a river to drown.

These tales are true, blood-curdling, and perhaps more chillingly still, are the stories of people whose behaviour can be considered rational.

They are four of 300 “suicides by proxy” (the bulk of which occurred in the 1700s), researched and described by Professor Kathy Stuart in her article on the topic, as well as her interview on This American Life, and upcoming book Suicide by Proxy: Crime, Sin and Salvation in Enlightenment Germany.

Kathy’s article is subtitled “The Unintended Consequences of Public Executions in Eighteenth-Century Germany”. I prefer to think of it as an example of the unintended consequences of incentives.

What’s rational?

Let’s get something out of the way. When I say these acts were rational, I don’t mean rational as in “reasonable” or “sensible”, as the term is often used colloquially. I mean rational in the way economists or rational choice theorists use it.

If you’ve ever taken an economics course, or read an economics text for thrills, or been pinned down at a party after unwittingly asking an economist what they do, you’re probably familiar with the idea of people being “rational” as meaning that they weigh the marginal costs and benefits of each possible action before choosing the one that gives them the largest benefit relative to the cost. Criminologists will recognise the approach in Becker’s work on rational crime, and political scientists will likely be familiar with rational choice theory.

And if you were taking a good course, reading a good book, or chatting with smart economist? Well, then you’ll know that while it’s assumed people will respond to incentives that affect these costs and benefits, the result might be different than you would imagine.

The law of unintended consequences

Suppose you face a rat plague. Getting private citizens to help catch rats might help keep numbers down, but they might be reluctant to help given the risk of disease. So, if you increase the benefits of hunting rats by, say, offering a bounty on rats, the rational response would be for more people to go out and catch rats, right?

When this was tried in Australia, Vietnam and Peru, amongst other countries, some people responded rationally, but not in ways that reduced the rat population; they bred rats. This ensured they had a ready supply to cash in for payment.

Another well known example relates to car accidents in the US. In the late 1960s, legislation was passed to make a range of safety devices (including seatbelts and padded dashboards) mandatory in all cars, in the hope of reducing deaths from car accidents. In 1975, Charles Peltzman proposed that although these measures would make passengers safer in the event of an accident, they technically reduced the costs of driving recklessly, making it more likely that some people would take more risks while driving. His research suggested that while there were fewer deaths per car accident, there were more car accidents, and that these effects had offset each other.

Of additional concern was that the number of pedestrians killed in car accidents went up; as Steven Landsburg puts it in his accessible overview of Peltzman’s research “pedestrians, after all, gain no benefit from padded dashboards”.

Such behaviour is rational and the consequences, while not necessarily predictable, can be explained or understood after the fact in terms of rational responses to incentives.

Religion, rules and rationality

What, then, made the behaviour of Stuart’s child-murderers rational, in an economic sense? Stuart explains that in the Holy Roman Empire, suicide was a mortal sin. Kill yourself, and there would be no confession, no salvation – just eternal damnation. But this rule, with this most substantial of spiritual costs, didn’t have quite the intended effect.

Kill someone else knowing you will be executed for murder, turn yourself in, ask forgiveness before your execution, and you could effectively kill yourself and still get to heaven. And if you kill an innocent child, who hasn’t yet sinned, the child goes to heaven too, before growing up to “learn damnation”, as one 18th century jurist put it.

So, regardless of the intentions, the religious rule didn’t put a stop to suicides among the pious, and wrought terrible consequences for hundreds of children. The seven-year old killed by Agnes Schickin to bring about her own hell-free demise is analogous to the innocent pedestrian in this more twisted and macabre example of unintended consequences.

The problem for policy-makers

These suicides by proxy also demonstrate the difficulty policymakers can face when trying to tweak incentives. Responding to this spate of child murders, Stuart notes that policymakers made the executions more painful, but cases persisted. Even when the death penalty was removed, cases continued for some time before petering out.

Why the slow response to the changed incentives? Was it just that old habits die hard? Was information about the change in the laws slow to travel, or undermined by rumours that execution could still be expected?

I’d wager that what remained clear was the strong religious rule that suicide meant eternal damnation, that murder did not, and that if there was even a chance of execution then people would still be willing to chance it.

I’d also wager that if the church had preached far and wide that they’d no longer grant absolution to murderers, it might have stopped people from killing the children (though not from killing themselves). But then, I’d be making a prediction about how a rational agent will respond to a change in incentives. And this can be a fool’s errand.

The views expressed within this article are those of the author and do not represent the views of the ESSA Committee or the Society's sponsors. Use of any content from this article should clearly attribute the work to the author and not to ESSA or its sponsors.

  • Monika

    Does this article suggest the rational actor must be secular?

    In the case of persons trying to cash in on rat bounties, or persons in safe cars feeling a false (ish) sense of security with the pedestrian the ultimate loser, altruism has not prevailed and we would assume rationality. Their rationality has unintended consequences. But policymakers can learn from this and policies can be adapted accordingly.

    Is ever it sensible to think of the religious person as a rational actor? Using the logic applied to suicides by proxy, a radical Islamist suicide bomber is a rational actor. As religion relies on faith and the underpinning foundations cannot be validated by empiricism, by definition an individual whose actions are based on religious belief is not a rational actor. Therefore where religious belief forms the context for an action, providing sanctions and incentives to alter behaviour (ie making execution more painful) based on assumptions of rationality are doomed to failure. There is no room for learning for the policymaker. God/s trumps earthly incentive always.

    Perhaps traditional economic incentives only apply in essentially secular societies.

    Is religion then the enemy of the policymaker?

    • http://@Mike_Pottenger Mike

      “As religion relies on faith and the underpinning foundations cannot be validated by empiricism, by definition an individual whose actions are based on religious belief is not a rational actor.”

      Interesting thought, though this depends on how we define rationality. In economic and rational choice theory, an individual is said to be rational if they choose courses of action that maximise their expected utility, based on preferences determined by the consideration of the marginal costs and benefits associated with each action.

      Suppose a person is faced with a choice between two beverages, neither of which they have tried before. They’re deemed rational according to the above definition if they choose whichever one they expect will maximise their utility. Faced with imperfect information, there’s no guarantee that they will actually maximise their utility, but rationality is defined based on the process,
      not the result. If they end up choosing one beverage over the other for no better reason than the fact that they like the colour of the label more than the other, and hope or believe this means they will enjoy it more, then they can still be regarded as rational.

      If a person’s perception of a cost or benefit is a function of their beliefs, this doesn’t mean they aren’t still behaving as a rational agent in that sense, in the context of those beliefs.

      What I’m pointing out in the article is that the suicides by proxy actually were rational, according to this definition.

  • Monika

    Found out today that a rat bounty policy (free mobile phone for 60 dead rats) has just been instigated in South Africa.

    Brings to mind the famous quote “history repeats itself, first as tragedy, second as farce”

    • http://@Mike_Pottenger Mike

      Thanks – good to have another example to add to the folder for class!

      The bounty problem is sometimes called the Cobra Effect, with regard to a similar problem with cobras in India under British colonial rule.

      Disappointing that there was no mention of it in the article – a quick search for “rat bounty” quickly reveals examples of cases that backfired.

Founding sponsors




Gold sponsors


Silver sponsors