Uri Gneezy and John List write:
Recent discoveries in behavioral economics have led scholars to question the underpinnings of neoclassical economics. We use insights gained from one of the most influential lines of behavioral research — gift exchange — in an attempt to maximize worker effort in two quite distinct tasks: data entry for a university library and door-to-door fundraising for a research center. In support of the received literature, our field evidence suggests that worker effort in the first few hours on the job is considerably higher in the "gift" treatment than in the "non-gift treatment." After the initial few hours, however, no difference in outcomes is observed, and overall the gift treatment yielded inferior aggregate outcomes for the employer: with the same budget we would have logged more data for our library and raised more money for our research center by using the market-clearing wage rather than by trying to induce greater effort with a gift of higher wages.
In other words, people in the real world show behavior much like that of traditional economic agents. Here is the paper. Have I mentioned that John List is one of the most important young economists? He has jumped from a U. Wyoming Ph.d. to a U. Maryland job to the notoriously-stingy-to-tenure Department of Economics at the University of Chicago. If you want to see a tough skeptic about many commonly accepted research results, especially in the realm of economic experiments, read some of John’s other papers. John is developing more finely grained methods of discovering when we should believe laboratory experiments. Are you surprised he puts greater trust in market data?