It is argued that policymakers, macroeconomists and microeconomists should all take high unemployment more seriously. The shortcomings of existing theories of unemployment are discussed, and a new definition of involuntary unemployment is proposed. A model is sketched in which falling aggregate demand leads to "Keynesian" unemployment because labor is heterogeneous and relative wages matter. Microeconomic theory is criticized for assuming away unemployment and, in the process, radically changing the answers to some basic questions in trade theory and public finance. Finally, some speculative explanations are offered for the low unemployment now found in states like New Jersey and Massachusetts.