Free Riding in the Lab and in the Field

We run a public good experiment in the field and in the lab with (partly) the same subjects. The field experiment is a true natural field experiment as subjects do not know that they are exposed to an experimental variation. We can show that subjects' behavior in the classic lab public good experiment correlates with their behavior in the structurally comparable public good treatment in the field but not with behavior in any of two control treatments we ran in the field. This effect is also economically significant. We conclude that a) the classic lab public good experiment captures important aspects of structurally equivalent real life situations and b) that behavior in lab and field at least in our setting is driven by the same underlying forces.