The finding is clear-cut. People that buy life annuities live longer than people that don’t. Therefore a theory that buying an annuity will help you live longer is supported…or is it?
Probably not. A more likely reason is that the people buying an income that lasts as long as they live are in good health and come from a long living family. It could also be that people concerned with not running out of money are more risk averse, and the reason life annuity buyers live longer is because risk averse people live longer. Or it could be that since getting a meaningful life annuity income requires a fair amount of premium, more wealthy people buy life annuities than poor people–and maybe wealthy people live longer. Or maybe education is a factor–better educated people buy more annuities, because they are more likely to be aware of them, and better educated people live longer.
We have a case where we believe the conclusion is well supported, but the “why” is not. The reason is that there is more than one possible explanation, or variable, to account for the longer lives of life annuity buyers. These multiple variables, or covariates, make it difficult to determine the real cause of the outcome because they interact.
When there is a news story about some scientific finding, the reporting usually presents it as black and white–”Scientists Find Dodo Bird Eggs Cure Cancer”–but the study’s actual written conclusion tends to be more mealy-mouthed and use a lot of qualifiers because of the possibility that other variables played a part in that cancer cure. This lack of directness and finality can be frustrating when you want a black and white answer, but the reality is most answers are a shade of gray.
For the life annuity question you could try to account for or eliminate (control) other variables. For example, if you looked at groups of only college graduates and only high school graduates–and they both preferred life annuities with equal gusto–then you might be able to rule out education as a variable. Then you might have the subject groups complete a risk aversion test too, and this might rule out risk aversion as a variable. If you can isolate the variables you have a better chance of coming up with the right answer.
How does one deal with this uncertainty? The way I handle it is by accepting that every belief I think is correct today may be proven false tomorrow. It’s kind of like renting your house of beliefs rather than owning it. This makes it easier when a belief you hold dear is overturned, because you simply move to the next rental. It also makes it easier to challenge your own beliefs. I can’t begin to count the number of beliefs I once thought were true that have been replaced by new information.
This does not mean you blindly accept the belief of the day–I wouldn’t run out and stock up on Dodo eggs–but it also means if you do find credible support for the Dodo egg hypothesis, that you don’t ignore it. Maybe having scrambled Dodo eggs at Sunday brunch is worth a try. You simply hold onto to the belief unless and until it is disproved and replaced.
