Research on Human Honesty
I really don't think this study says much about humans (or economics or civics or anything else mentioned) at all. I'd argue that the setup is flawed by the condition "turned in . . . at public and private institutions". By doing that, you introduce all sorts of confounding factors that come with corporate policies and behaviors based on professionalism.
For example, it might simply be standard practice to drop most things into "lost and found" and wait for someone to come claim it. Beyond a certain value, though, it might make sense to spend valuable employee time trying to track down the owner. It'd be done because it reflects well on the corporation; the employee (in most cultures I'm aware of) should not for one second expect that they'll get a reward from the owner of a found item, wallet or otherwise. Either way, you don't just assume someone is being "dishonest" because you're not contacted.
Likewise, they should absolutely have expected a 98% monetary return for those who contacted them. Again, the economics of "feels like stealing" have a completely different baseline when the consequences of actual dishonesty can result in the loss of your job. Especially these days, as readers of this blog surely know, when cameras are increasingly in public spaces and corporations do security monitoring of all kinds (including, perhaps, even operational stings/tests that may be very similar to this experiment).
The better assumption is that there's an "uncanny valley" when it comes to job-related theft: things of lesser value go missing constantly, and then it drops off as the theft becomes less petty, shooting back up again when the value of the items that can be taken well exceeds the pay of the job. The fact that "non-experts and professional economists were unable to predict this result" just means that a lot of people don't think very hard about how reality works.
2672df5d8e0befe13b6eeb410d80169be80e2f425f4930d6c3ca8c23d5723507