The relevance of empirical work is determined largely by its ability to resolve questions in theoretical debates, and hopefully move those debates forward. I'm encouraging comments from others on areas that they believe have been largely resolved, or where empirical work has helped move the theoretical debate forward. (Or maybe where we've been spinning our wheels?)
To get things started, one area of particular interest to me has been the rapidly growing body of work on the effect of exceptions to employment at will. Through the 1970s and 80s, many states recognized exceptions to the employment at will rule. Some scholars (notably Stewart Schwab) suggested that this trend may have been motivated by the desire of courts to police long-term employment contracts for opportunism. Others (notably Richard Epstein) worried that these exceptions would simply increase firing costs. In this case, present employees are benefited by the rule that makes it harder to dismiss them, but future workers face costs as employers are now more wary of hiring anyone.
Miles (2000), Schanzenbach (2003); Autor (2003); Kugler & St. Paul (2004); and Autor, Donohue & Schwab (2006), all found remarkably consistent results: exceptions to employment at will appear to have raised firing costs, and employers responded by turning to temporary help, retaining current employees longer, and (perhaps) employing fewer workers period. These authors used different data sets, different outcome variables, different methodologies, and different dating techniques for deterring the recognition of exceptions. Given this, the outcomes are remarkably consistent.
A new paper by McLeod & Nakavachara (available here) refines the previous analysis a bit by suggesting that workers for whom joint investments are important may have benefited from employment-at-will exceptions (specifically the "good faith exception"). Unfortunately, lower-skilled workers were adversely affected. The authors say the results are preliminary, but the work seems to be taking the next logical step in the literature.
This may revive Schwab's theory, but raises a difficult public policy question. If the results are correct, the rule works out to a wealth transfer that is probably pretty regressive (and given the previous work mentioned above, probably a net negative)--unless the courts could refine the rule to only protect certain classes of employees, which is likely beyond the ability of the common law.
So here we have an example of consistent work, informed by theory, and now moving theory forward--with some implications for public policy.
Are there areas where we have been spinning our wheels? One candidate that comes to my mind is empirical work on tort reform. It's not my area, but my impression is that people find no effects, some effect, opposite effects, and so forth.
[Exact citations for the articles referenced here are available in the linked paper.]