Economics is often abstract. Economic models are formulated in general terms, without reference to any particular economy. Theorizing involves assumptions that are patently unrealistic. Laboratory experiments involve artificial environments that may be quite different from real-world situations. Econometric studies often exploit specific data sets that allow the researcher to identify certain effects. But we are not interested in models, experiments, or econometric regressions for their own sake, but for the insights they provide about the economy at large and for the uses to which they may put in dealing with real-world problems. Relying on the experience of the participating laureates, the panel will discuss why and under what conditions the insights gained from models, experiments, or econometric regressions can be used outside these narrow domains and how they contribute to dealing with real-world problems.