Perrier's 100% loss of market share in early 1990 provides an extreme example of a `firm-specific business conditions shock'. Such shocks clearly affect the stock market value of firms, but their effects on employees will depend on the institutional characteristics of the labour market. Legislation and labour contracts protect employees from losing their jobs because of adverse developments in product markets, but the degree of job security this affords varies across countries, time, firm sizes and sectors.

Larger turnover costs reduce the variability of employment at a given firm, and they also reduce the value of firms in a dynamic environment because firms seeking to maximize their value need not take fully into account the value attached by workers or by society to the stability of employment relationships.

In Discussion Paper No. 422, Research Fellow Giuseppe Bertola presents a dynamic general equilibrium model that allows for random factors at the firm level, ex ante investment decisions, and ex post employment and production choices.
Bertola notes the importance of accounting for growth and technological progress endogenously, which allows different institutional settings to be modelled in terms of constraints on a firm's `flexibility' in its ex post decisions. Bertola allows explicitly for fluctuations in business conditions at the level of the individual firm, but not at the aggregate level. He assumes labour supply to be exogenously fixed, that all firms require labour, and that output in general equilibrium may be either consumed or invested in setting up a new firm. New goods are endogenously introduced over time by entrepreneurs, the creation of new firms increases the overall productivity of the economy by more than the monopoly rents captured by entrepreneurs, and these productivity spillovers permit steady, endogenous growth in equilibrium.

Bertola first discusses the solution under perfect flexibility, before introducing various imperfections reflecting the influence of institutional constraints on individual decisions. Real resource costs of labour reallocation across firms reduce welfare and growth and may fall on either `labour' or `firms', and labour reallocation may be initiated by firms or workers. The variability of wage rates over time and across firms, and of firms' asset values, depends on the labour contracts, but the dynamic allocation of labour and the equilibrium growth level are not affected by institutional factors, provided all agents are wage-takers and capital markets are perfect.

Bertola notes, however, that wage earners may find it optimal to lobby for stringent job security provisions if they are to pay some or all of their match-specific training and set-up expenses. In this case, turnover costs may be partly due to institutional rather than technological features of the economy. Higher productivity and faster productivity growth will reward countries with greater flexibility of production.

Flexibility, Investment and Growth
Giuseppe Bertola

Discussion Paper No. 422, June 1990 (IM/AM)