Productivity : the capital stock and profit sharing in UK firms 1974-86 : a micro-data study
The work described in this thesis is concerned with empirically testing four hypotheses about the observed behaviour of the labour market over the 1974-86 period. The data used for these tests is drawn from the published accounts of a panel of UK manufacturing firms. Specifically these four hypotheses are:- 1. That the oil shocks of 1973 and 1979 caused a great deal of 'economic' scrapping of capital equipment. Such a fall in the capital stock would not be picked up in the official figures, and could be a cause of the fall in measured TFP growth after 1973. Using data on 'current cost' valued assets to inflation adjust the 'historic cost' figures contained in accounts we conclude that the official measure of the capital stock does overstate the 'true' figure for most of this period, but that this overstatement is by much less than suggested by other authors and could not account for more than a fraction of the productivity 'slowdown.' 2. That the introduction of profit related pay will lead to desirable economic outcomes including higher productivity and perhaps a cure for stagflation. Specifically we test the models of firm and worker behaviour that are due to Martin Weitzman. We test a key proposition of Weitzman by exploring it's implications for empirical employment and stock returns equations. We find no support for Weitzman's proposition. We also find evidence that profit sharing will be wage inflationary. However there is some evidence that profit sharing causes higher productivity. 3. That payment of 'efficiency wages' enhances productivity. The efficiency wage hypothesis suggests that firms will pay high wages because they more in the way higher, productivity than they lose in terms of increased wage costs. This could explain why High wages persist with high unemployment. We test for a direct effect of relative wages on productivity. We find such an effect and carry out further experimentation to test whether this result points to an efficiency wages or some competititive explanation of wage setting. We also find that high unemployment leads to higher productivity. This finding provides some evidence for efficiency wage models. The finding that high unemployment boosts productivity could be an explanation of the high productivity growth of the 1980s. 4. That firms that recognise unions will have lower productivity growth then non-union firms. Unions are supposed to adversely affect the economic performance of the firm by defending restrictive practices and resisting technical change. Using a data set that includes firm level data on union status we find that there is no evidence to suggest the union firms have lower productivity over our sample period. We find that unionised firms grow at the same rate as non-union firms over the 1975-78 and 1985-86 periods, but grow faster over the 1980-84 period. We feel this effect is consistent with unions being unable to defend restrictive practices over the eighties, but that the probable cause of this is the 'shock' effect of the 1979-81 recession rather than the change in trade union legislation. We also assess whether unions act as a deterrent to investment. This would be the case if investors expected unions to capture the quasi-rents from capital. We show that union density does not effect investment and that union firms do not invest less than non-union firms. By estimating a wage equation we also show that unionised workers are not more effective than non-unionised at capturing the quasi-rents from capital.