Growth and Employment
Innovative search

Keynesian short-run analysis implies that a rise in growth that results from increased aggregate demand may reduce unemployment; but the long-run effect of growth resulting from the advancement of knowledge is less clear. If creating job openings is more profitable in a faster-growing economy, the unemployed may find jobs more quickly; but technological innovation may also raise the rate of job destruction. In Discussion Paper No. 577, Research Fellow Philippe Aghion and Peter Howitt examine these long-run effects for an economy consisting of a continuum of perfectly competitive sectors. They use a variant of the Pissarides search model, in which unemployment is due to the intersectoral reallocation of labour, to examine industrial innovations in the form of productivity improvements. If the current workers of a firm that experiences an innovation cannot work with the new technology, it can hire new workers that match its new requirements only through time-consuming search. Firms with the newest technology pay the highest wages: if workers' search activities are unrestricted, all workers therefore search for new jobs even when employed.
Aghion and Howitt find that the overall effect of growth on unemployment depends on its source. Higher growth may result from an increase in the size or frequency of innovations. Increased size reduces the likelihood that incumbent workers can adapt to the new technology; while increased frequency raises unemployment directly, by increasing the job-destruction rate, and also indirectly, since the expected duration of any newly created jobs is shorter. Increased growth will also raise the interest rate, which will discourage the creation of job openings and thus raise overall unemployment, if the intertemporal elasticity of substitution in consumption goods is small. Also, if demands for different sectors' output are complementary, demand for output (and hence job creation) will increase with growth in the others. Innovations in a given sector also reduce the relative price of its output, however, which discourages job creation to an extent that increases with the complementarity of intersectoral demand.
Aghion and Howitt also extend the model to allow for endogenous growth, and they show that reducing the variable costs of research has similar effects to an exogenous increase in the size or frequency of the innovations it produces; while reducing its fixed costs always reduces both unemployment and growth. Introducing `learning by doing' with positive external effects across sectors creates a feedback from unemployment to growth that amplifies the above effects. If this is strong enough to produce multiple equilibria so that high (low) growth and low (high) unemployment can sustain one another government intervention may be required to avert coordination failure.

Growth and Unemployment
Philippe Aghion and Peter Howitt

Discussion Paper No. 577, September 1991 (IM)