One of the most famous graphs in the behavioral economics community — from Johnson & Goldstein’s 2003 paper in Science — depicts the rates at which people in various countries consent to donate their organs. It shows that consent rates are dramatically lower in countries that require people to opt-in to donate, as opposed to those that automatically register everyone and then make them opt-out if they don’t want to participate.
Dan Ariely says that the graph is one of his “favorite graphs in all of social science,” and that it demonstrates “the influence of rather ‘small’ changes in the environment (opt-in vs. opt-out) on our decisions.” Richard Thaler’s recent book, “Misbehaving,” claims that the paper shows “how powerful default options can be.”
Yet, does the number of registered organ donors, in and of itself, really matter? After all, there are a lot of things that people neglect to opt out of, including, for example, the terms and conditions of every website they regularly use. (For all I know, when I signed up for Facebook, I inadvertently agreed to donate one of my kidneys.)
Moreover, it is not clear that doctors even pay attention to who is “signed up” to be an organ donor. In opt-in countries, doctors routinely seek and receive permission from family members to donate a deceased relative’s organs, regardless of whether the person is officially registered as a donor. And the same is true for opt-out countries: a PBS report noted that doctors in such countries still check with family members before carrying out the donation process.
Thus, organ donation consent rates may not be all that important. What really matters is how often people actually donate their organs. Here is a graph of the actual donation rates in those same countries in 2003 (drawn from the data here).
There’s still a bit of a difference, but it’s not nearly as pronounced. In fact, Poland and Sweden had donation rates that put them squarely in the middle of the opt-in countries.
To be fair, Johnson and Goldstein’s original paper went on to explain that an opt-out law moved the actual donation rate by only about two donations for every million people — a fairly small effect that is nowhere close to the dramatic differences depicted in their chart on consent rates, and that is similar to the effect found in a recent systematic review of 13 studies on the impact of opt-out legislation.
But even that modest effect might be an illusion, because none of these studies can properly account for selection bias. Perhaps the countries in which people view organ donation more favorably are also more likely to pass presumed consent laws in the first place, and so the higher actual donation rates in those countries may be due, entirely or in part, to the overall more favorable attitudes.
I would therefore suggest that, while default options certainly can influence our decisions, we shouldn’t give too much weight to a chart that suggests that people in Belgium donate 23 times as many organs as their neighbors in Denmark. The true effect of opt-out laws, if any, is surely quite a bit smaller than that.