Labour market flexibility and employment adjustment: micro evidence from UK establishments.

1. Introduction

The issue of labour market flexibility has recently become prominent in the debate over labour market policy, amid evidence suggesting labour markets have become more flexible in many developed economies in recent years (OECD, 1986, 1989, 1994; Standing 1986, 1988; de Neuborg, 1990; Lilja et al., 1990; Walterskirchen, 1991; de Luca and Bruni, 1993; Bentolila and Dolado, 1994; Jimeno and Toharia, 1994; Beatson, 1995; Bertola and Ichino, 1995). It is argued that flexibility has beneficial effects on employment, output, and prices, making the economy less inflation-prone and improving the prospects for job creation.

However, there is no clear definition of flexibility and the term has been used in a variety of ways. For example, there is what Beatson (1995) refers to as the intensive margin of flexibility. This describes the ability of a firm to deploy its workforce as it wishes; a more flexible firm in this sense is less constrained by manning and demarcation agreements and is more able to re-deploy labour when faced by shocks (Cross, 1988; OECD, 1989; Marsden and Thompson, 1990; Machin and Wadhwani, 1991a, b; Lorenz, 1992; Booth and Chatterji, 1995). The recent OECD Jobs Study uses flexibility in this way when it argues that '... traditional patterns in the organisation of work and working time... hinder labour market flexibility and, indirectly, job creation' (OECD, 1994, p. 34). There is also the extensive margin of flexibility. This denotes turnover costs; a more flexible firm in this sense finds it less costly to adjust the size of its workforce (Emerson, 1988; Grubb and Wells, 1993). Flexibility has also been used to describe adjustment in the wider labour market; flexibility in this sense refers to a greater responsiveness of real wages to unemployment (OECD, 1986; de Luca and Bruni, 1993; Bentolila and Dolado, 1994; and Jimeno and Toharia, 1994).

Despite its prominence in the current policy debate, there is little direct empirical evidence on the effects of flexibility. Most studies use proxies such as trade union presence (which can be thought of as a proxy for flexibility on the extensive and intensive margins) or the presence of temporary workers (which can be regarded as a proxy for flexibility in the wider labour market; see Bentolila and Dolado, 1994) to infer the effects of flexibility. Although there is some value in this, there are also clear dangers. For example, unions may impose restrictive practices that increase the marginal cost of labour and so reduce flexibility along the extensive margin. But, they may provide a channel of communication and so increase flexibility along the intensive margin.

This paper differs from earlier studies in that we use direct data on flexibility along the intensive and extensive margins to study the effects of flexibility at the microeconomic level. Using data on 396 establishments drawn from the 1990 Workplace Employee Relations Survey (Millward et al., 1992), we study how flexibility affects the way that firms respond to demand shocks. This addresses an issue at the heart of the policy debate, namely whether more flexible firms are more likely to increase employment and less likely to increase price following a positive demand shock.

We have three main findings. First, very few firms choose to respond to a demand shock by adjusting price, whereas 92% of firms choose a quantity response such as adjusting employment, hours, or capacity. There is some evidence of asymmetry in the response to demand increases and decreases in that employment and capacity adjustment are more common with a positive shock. Also there is more adjustment of employment and capacity in the long run than the short run but more adjustment of hours in the short run than the long run. Second, firms which are more flexible on both the extensive and intensive margins are more likely to respond to a demand shock by adjusting employment and hours of work. Third, the effect of unions is to encourage adjustment of hours rather than employment.

Our results can be seen as offering a microeconomic explanation for recent macroeconomic evidence of changes in the way the business cycle affects the labour market. Our second main result suggests that labour input will be more responsive to demand in a more flexible labour market, so both employment and unemployment should be more closely aligned to the business cycle. As Bertola and Ichino (1995) document, these changes have been observed in recent years in many OECD countries (for more specific UK evidence, see Blanchflower and Freeman, 1994; Beatson, 1995).

Our results also predict that productivity growth should become less sensitive to the business cycle; this has also been observed in the UK over recent years (Beatson, 1995). The remainder of the paper is structured as follows. In Section 2 we discuss our data and define the variables we use; in Section 3 we discuss our econometric technique and present our results; finally, Section 4 concludes and assesses the policy implications of our results.

2. Data

Our basic data set is the 1990 Workplace Industrial Relations Survey (WIRS). This is a nationally representative survey of some 2,000 UK establishments which has become a primary data source in the labour economics and industrial relations literatures (Millward, 1993).

The variable we study concerns the way firms respond to demand shocks. We use responses to the following question, asked to financial managers:

Suppose there were a substantial and sustained rise in demand for your product or service. What would be management's most immediate response?

They are asked to choose one of the following responses: to (i) do nothing; or adjust (ii) price; (iii) employment; (iv) hours; (v) output; (vi) capacity; (vii) subcontracting; (vii) wages; or (viii) another response (we do not observe what this is). They are then asked the same question about responses to a reduction in demand and are again asked for one response (the 'output' response is replaced by 'increasing …