How Game Theory Can Transform Your Kid’s Halloween Candy from Your Worst Enemy to Your New Best Friend
As a parent, you will dread these days after Halloween. That stash of Halloween candy sits in your kid’s bedroom. “Bit-size” Butterfingers wait to stick to molars and drill their little cavity holes, the hard brown goo in the Tootsie Roll wrappers await their contribution to childhood obesity, and the BlowPops prepare for bouncing your child off the walls in a frenzied sugar high. Maybe the pile has been moved to a kitchen shelf for “safe-keeping,” regulated at a piece or two a day. Still, you wish it would go away. You have come to hate the days and weeks after Halloween.
But there is good news: Deep in that Halloween candy lies a responsible parent’s best friend. A simple bit of game theory sheds light on this serendipitous bounty that lies in your child’s Halloween bag.
One challenge every parent faces is the loving, yet firm, discipline of children. Along with proper communication, parenting books are virtually unanimous in urging parents to address a child’s bad behavior with negative consequence, especially if the behavior is willful, and even more if it is repeated AND willful. The problem with this necessary but tiresome aspect of responsible parenting, is that it is exasperating, often inconvenient, and unpleasant.
Because of this, statements 
articulating the usual consequences for inappropriate behavior are often not credible. Consider the diagram in Figure 1. The game consists of a child choosing between a good action and bad action, and a parent choosing a response in the event that the child chooses the bad action. First, suppose the child chooses the good action. In this case, everything is fine. The child (player 1) has a payoff of, say, 3. The parent (player 2) earns a payoff of, say, 10 because everybody is happy, and no sibling is having her hair pulled.
But what happens if the child decides to pull his sister’s hair. Perhaps the parent threatens a “time out”; but the family has plans, or for some other reason the time out is inconvenient. Maybe all this happens at the grocery store (a time out in the vegetable section?). Some parents not wishing to “spare the rod” (or at least the hand) might even threaten a spanking. But while a spanking is no fun for the child, hopefully it is no fun for the parent either. Faced with the decision to follow through with it, many parents would rather not. These days most parents feel bad about spanking children, and kids are quick to pick up on this vibe. So at least sometimes, time outs and spankings may constitute what game theorists call “non-credible threats.” They fail to pass the hurdle of what is called a “subgame-perfect Nash equilibrium.” And when a child begins to learn that these threats are non-credible, or at least the specific instances when they are non-credible, the bad behavior may be “on the equilibrium path” as shown by the arrows in Figure 1. 
At this frustrating juncture lies the nearly messianic wonder of the Halloween candy bag. It provides a new tool in the arsenal of the parent that possesses some magnificent game-theoretic properties for “helping children make good choices.” Foremost among these is the basic fact that even though the trick-or-treater likes the candy, it is not good for them. Moreover, the parent knows it’s not good for the child, and the child knows that the parent knows that it’s not good for them. This makes the threat of taking a piece of Halloween candy away from the child eminently credible, as seen in Figure 2. The child knows that if the parent threatens to take away a piece of Halloween candy, that she will follow through because the child knows that the parent knows it’s bad for the child anyway. (Plus perhaps the child suspects that the parent may sneak off to a corner of the house and eat it herself, transforming a potentially distasteful punishment into a significantly more appealing one.) But either way, seizing a piece of the Halloween candy is a highly credible threat that in the example in Figure 2 gives a payoff of -8 to the child, a negative payoff that here lies somewhere between a timeout (-5) and a spanking (-10). But it’s not the payoff that counts, it’s the credibility of the threat. And moreover, the value of Halloween candy option is augmented by what behavioral economists call loss aversion: People hate losing something they possess more than they like receiving the thing in the first place. All of this just simply piles on to the credibility and consequence of the Halloween candy seizure threat.
I must add that this theory does not lack for empirical substantiation, as it has been rigorously and repeatedly tested in the context of our post-Halloween household. The Halloween candy warning is generally met with a (picture slow motion movie drama) NNNNOOOOO!!! followed soon thereafter by “good choices.”
So three cheers for the Three Musketeers, for Baby Ruth, for Pez, Sugar Daddies, for the whole ensemble of Halloween sweets—a responsible parent’s new best friend.
Follow Bruce Wydick on Twitter @BruceWydick.
Spank or Treat
How Game Theory Can Transform Your Kid’s Halloween Candy from Your Worst Enemy to Your New Best Friend
As a parent, you will dread these days after Halloween. That stash of Halloween candy sits in your kid’s bedroom. “Bit-size” Butterfingers wait to stick to molars and drill their little cavity holes, the hard brown goo in the Tootsie Roll wrappers await their contribution to childhood obesity, and the BlowPops prepare for bouncing your child off the walls in a frenzied sugar high. Maybe the pile has been moved to a kitchen shelf for “safe-keeping,” regulated at a piece or two a day. Still, you wish it would go away. You have come to hate the days and weeks after Halloween.
But there is good news: Deep in that Halloween candy lies a responsible parent’s best friend. A simple bit of game theory sheds light on this serendipitous bounty that lies in your child’s Halloween bag.
One challenge every parent faces is the loving, yet firm, discipline of children. Along with proper communication, parenting books are virtually unanimous in urging parents to address a child’s bad behavior with negative consequence, especially if the behavior is willful, and even more if it is repeated AND willful. The problem with this necessary but tiresome aspect of responsible parenting, is that it is exasperating, often inconvenient, and unpleasant.
Because of this, statements
articulating the usual consequences for inappropriate behavior are often not credible. Consider the diagram in Figure 1. The game consists of a child choosing between a good action and bad action, and a parent choosing a response in the event that the child chooses the bad action. First, suppose the child chooses the good action. In this case, everything is fine. The child (player 1) has a payoff of, say, 3. The parent (player 2) earns a payoff of, say, 10 because everybody is happy, and no sibling is having her hair pulled.
But what happens if the child decides to pull his sister’s hair. Perhaps the parent threatens a “time out”; but the family has plans, or for some other reason the time out is inconvenient. Maybe all this happens at the grocery store (a time out in the vegetable section?). Some parents not wishing to “spare the rod” (or at least the hand) might even threaten a spanking. But while a spanking is no fun for the child, hopefully it is no fun for the parent either. Faced with the decision to follow through with it, many parents would rather not. These days most parents feel bad about spanking children, and kids are quick to pick up on this vibe. So at least sometimes, time outs and spankings may constitute what game theorists call “non-credible threats.” They fail to pass the hurdle of what is called a “subgame-perfect Nash equilibrium.” And when a child begins to learn that these threats are non-credible, or at least the specific instances when they are non-credible, the bad behavior may be “on the equilibrium path” as shown by the arrows in Figure 1.
At this frustrating juncture lies the nearly messianic wonder of the Halloween candy bag. It provides a new tool in the arsenal of the parent that possesses some magnificent game-theoretic properties for “helping children make good choices.” Foremost among these is the basic fact that even though the trick-or-treater likes the candy, it is not good for them. Moreover, the parent knows it’s not good for the child, and the child knows that the parent knows that it’s not good for them. This makes the threat of taking a piece of Halloween candy away from the child eminently credible, as seen in Figure 2. The child knows that if the parent threatens to take away a piece of Halloween candy, that she will follow through because the child knows that the parent knows it’s bad for the child anyway. (Plus perhaps the child suspects that the parent may sneak off to a corner of the house and eat it herself, transforming a potentially distasteful punishment into a significantly more appealing one.) But either way, seizing a piece of the Halloween candy is a highly credible threat that in the example in Figure 2 gives a payoff of -8 to the child, a negative payoff that here lies somewhere between a timeout (-5) and a spanking (-10). But it’s not the payoff that counts, it’s the credibility of the threat. And moreover, the value of Halloween candy option is augmented by what behavioral economists call loss aversion: People hate losing something they possess more than they like receiving the thing in the first place. All of this just simply piles on to the credibility and consequence of the Halloween candy seizure threat.
I must add that this theory does not lack for empirical substantiation, as it has been rigorously and repeatedly tested in the context of our post-Halloween household. The Halloween candy warning is generally met with a (picture slow motion movie drama) NNNNOOOOO!!! followed soon thereafter by “good choices.”
So three cheers for the Three Musketeers, for Baby Ruth, for Pez, Sugar Daddies, for the whole ensemble of Halloween sweets—a responsible parent’s new best friend.
Follow Bruce Wydick on Twitter @BruceWydick.