How do we know that real wages are too high?

It is common belief that the wages that we observe are above the level that would prevail in a competitive labour market and also it is common to believe that wage moderation should be encouraged as a way to keep unemployment down. This paper considers whether we can have confidence in these beliefs. It presents a number of models designed to cast doubt on the conventional wisdom. First, we show that in an efficiency wage model in which there is involuntary unemployment, a binding minimum wage may increase employment. We then present a general equilibrium matching model in which there is involuntary unemployment but wages are below market-clearing levels and raising wages can reduce unemployment. We then consider the empirical evidence on employment and wage determination to argue that it is just as consistent with this model as with models in which wages are at or above market-clearing levels.