Imperfect Labour Markets

In 1971, Doeringer and Piore produced their influential book on internal labour markets in the US. Robinson, often in collaboration with his colleague at the Institute, Kenneth Knowles, had been independently developing his own ideas in this area and on the workings of local labour markets more gen­erally.

Robinson's detailed studies of mainly manufacturing firms in Coventry and Liverpool had led his thinking in the same direction. For the time, Rees and Shultz had an unusually rich data set of individuals describing their character­istics and those of the firms that employed them. Such data were not available to UK researchers, who often had to rely on broad averages of earnings by occupation and firm. Even when individual pay data could be extracted from employers, there was little or no further information on the individuals con­cerned.

However, Robinson went beyond the reporting of data and started to explore why firms were not simply wage takers but had some discretion in the wages they offered. Important here is the inability or unwillingness of workers to respond to signals from alternative employers. Embedded in Robinson's writings is discussion of why such a supply response might be limited—lack of information, mobility costs and efficiency wages. His investigation of local labour markets made it clear that workers often lacked basic information about starting salaries for apparently equivalent jobs in the same town. However, perhaps more important than this were two other sources of infor­mational imperfections. The first relates to Smith's concept of compensating differentials. Workers found it difficult to assess the non-wage and non­monetary benefits of working for a different employer. Even more significant was a time dimension. Even if an individual believed that she could assess the immediate impact of a potential job move, it was difficult to assess what the future in the new firm would hold for her. This depended so critically on things which could only be ascertained once a job had been taken up—how well she got on with her managers and fellow workers, how ruthless or other­wise her managers were when she made mistakes, how readily good perfor­mance was recognised in the form of extra pay or other rewards or promotion. Faced with such uncertainties, it might be quite rational for an individual not to respond to the lure of higher starting pay elsewhere—the rationality of irrationality or, as it became known, “bounded rationality”.

At first sight, the issue of mobility costs might not seem relevant for studies of local labour markets. Yet, moving from one firm to another even in a mod­erately sized town, let alone a large city, can involve increases in travel to work expenses and complicate childcare and other family responsibilities. Over and above these costs, Robinson and others emphasised the “psychic” costs of changing firms. People contemplating such a move might be concerned about losing contact with friends and colleagues at their current workplace, about leaving their social circle for a new and largely unknown community. In other words, he saw the firm as a social community as well as an economic entity. The third constraint on supply responses was the consequence of efficiency wages. Employers differentiated between workers in the same occupational category, recognising that some were more capable and productive than oth­ers and being willing to pay them “over the odds” in order to attract and retain them. Taken to the extreme, this implies that the only truly homogenous unit of labour is a single person, throwing the then typical textbook model into some confusion.

This imperfect supply response frees employers from the constraints of a perfectly competitive local labour market but, if that was all there was to it, the employer would, at least to an extent, be a monopsonist. However, Robinson was also aware of the literature on labour as a quasi-fixed factor of production, as pioneered by Walter Oi (1962). Employers may train workers in skills that are to an extent specific to his firm. They can pay (at least in part) the costs of that training because they calculate that the worker will be reluc­tant to leave the firm. Imagine someone is hired whose market value of mar­ginal product (VMP) is 100. After training, that person's VMP rises to 150 but only in that firm. Therefore, the employer can start to recoup the costs by paying the person less than 150 but more than 100 and the worker is likely to stay.

Robinson understood the fundamental imperfections in the labour market long before they became so central in academic thinking. Like Doeringer and Piore, he envisaged internal labour markets as administered markets often characterised by limited ports of entry, rewards for length of tenure and well- defined promotion ladders. Importantly, he understood that unions would have had relatively little room for bargaining or bargaining power over wage- related matters at the workplace but for the existence of internal labour mar­kets. Often, unions used this bargaining power to shape the institutions of the internal labour market—for example, by encouraging limited ports of entry or rewards for seniority. Although Robinson worked mainly on the internal labour markets of firms (horizontal internalisation), he also stressed the importance of similar forces at work to create and intensify occupational internal labour markets (vertical internalisation).

3