The subject of technological unemployment is used to illustrate how accepted theory leads to wrong answers because the technique used is that of partial equilibrium analysis. On this basis one rules out the investment needed to make the change in production method. In a dynamic context, in which such change continues, the employment effects of this investment should be allowed, and they can be seen to be substantial relative to the loss in employment from the installation of the new capital equipment. The result is that technological change conceived in this broad context is expansionary, and the implications for industrial relations and other policies is very important.