Submission 167 Australian Council of Trade Unions Workplace Relations Framework Public inquiry

The FW Act marked a return to a more open and transparent methodology for minimum wage fixation, which explicitly referred to “fairness” as a consideration. Since this institutional change has taken place, the wage fixation decisions of FWA and the FW Commission have been modest and predictable. Notwithstanding this, the minimum wage is often the target for derisive commentary on the basis of its supposed employment effects or that the framework for determining it and applying it is somehow outmoded. We disagree with both of those views.

Whilst we regard the institutional and procedural framework as appropriate, we are of the view that it could be made more effective by a clearer articulation of its re-distributional purposes to more squarely concentrate on reducing inequality and the incidence of low pay.

The minimum wage and employment

The problems of entrenched unemployment or withdrawal from the labour market amongst the lowest skilled has been a focus for policy makers for some time and resurfaces in the recent McClure report. Having summarised the problem in his own analysis (outlined earlier), Bob Gregory commented on possible policy responses. Across the board, he argued, most policy interventions had failed to solve the problem. His list included real wage freezes, expansions in education and training, changes in immigration to favour skilled immigrants, labour market deregulation and the weakening of trade unions. He also concluded that ‘The labour market bifurcation problem cannot be cured by Keynesian economic expansions of the sort that economists of my generation believed in for so long’.¹⁷⁷ With a heavy heart Gregory turned to a micro-economic solution that entailed cuts to relative wages for the unskilled—to create jobs—and cuts to welfare payments–to encourage labour supply. As he admitted himself, these solutions had been lifted out of Economics 101.

While Gregory was able to highlight the problem—the loss of full-time jobs for unskilled workers over a long period of real wage constancy—his attempt at a solution ran aground. It ended up repeating the familiar complaint against minimum wages, that they are job destroying. In this case, it simply invoked the inverse: that low wages create their own jobs. As many other economists not in thrall to neo-classical economics have argued¹⁷⁸, a race to the bottom by cutting wages does not work. More realistic models of the labour market, based on a dynamic monopsony framework rather than perfect competition, are more useful for thinking about how wages and employment are related.

The familiar invocation of the US labour market, with its higher employment to population ratios for the unskilled, is misleading. Labour markets are not separate from society, but an integral part of the social fabric. Along with these higher employment population ratios goes a large population of working poor: full-time workers who still require food stamps to survive; high levels of job turnover, such that permanent ‘Vacancy’ signs are a feature of low wage workplaces; and a host of social problems that extreme levels of inequality generate.¹⁷⁹

The fallacy behind the Economics 101 prescription is that it repeats the logic of Say’s law, that supply creates its own demand. If there are hoards of very low paid workers clamouring at the door, employers will let them in.

Conceptually this is unsound: as long ago as 1926 Sraffa pointed out that firms do not face diminishing marginal returns and that their employment levels are determined by the demand for their products and services¹⁸⁰. In other words, recruiting labour hinges on the current and anticipated state of customer orders, not on some mythical point on the curve where marginal cost equals margin revenue. As Joan Robinson put it more bluntly in her review of Sraffa’s 1960 book: ‘the marginal productivity theory of distribution is all bosh’.¹⁸¹

It might be argued that Say’s law is irrelevant and that relative wages are what matter. There are several responses to this view which are based on empirical data. Taking the very long-term perspective—from 1914 to 2010—Bray showed that the minimum wage as a percentage of average earnings has steadily declined for nearly all of that period, with the only exception being the 1970s and 1980s. During that period that percentage nearly halved, from over 80 per cent to just over 40 per cent.¹⁸² Moreover, over the last ten years in particular, the ratio of the minimum wage to the median has dropped from 48.2 percent to 43.3 percent.

Secondly, when Cully examined what employers look for when recruiting low skilled workers he found one major element in their decision-making, something which was likely to be inversely related to the cost of labour: experience. Indeed when employers used the term ‘productivity’ they did not mean marginal productivity. Rather they used it as a ‘short-hand term which refers to a set of characteristics, such as effort, loyalty, likely tenure and pattern of absence’.¹⁸³ Again, all of these elements are inversely related to the cost of labour. The advocates of Economics 101 ignore labour market frictions, even though the concept is buried somewhere in that tome. The high costs of labour turnover, an enduring feature of low wage markets, are ignored in the prescription for cutting the real wages of the low skilled.

Finally, Bruce Chapman’s simulation study of the strategy of upskilling the unemployed to increase their employment outcomes suggested poor outcomes, a result which had implications for the notion of cutting wages to solve unemployment:

the answer to Australian job creation, at least in the short to medium term, cannot rely on increasing the skills of the unemployed … That decreasing real unit costs in the ways considered is not the panacea for employment creation has two significant, albeit speculative, implications. One is that the other way of decreasing labour costs—wage cuts—might similarly have little effect on employment creation in the short to medium term.¹⁸⁴

The naive importation of the American enthusiasm for reducing the minimum wage as an answer to unemployment fails to acknowledge several stark realities about the Australian labour market. As Chapman also observed: ‘few employees are now earning a minimum safety net’. The operation of the minimum wage in Australia—as a set of pay scales across the award system—makes any simplistic analysis of wage elasticities at the bottom of the labour market largely meaningless.

Furthermore, would reductions in pay rates make any sense in those sectors of the economy where conditions like minimum wages and penalty rates most apply? Reducing labour costs to gain international competitiveness has been an industrial strategy in some countries, but the sectors of the economy where these conditions apply in Australia are largely service sector work which is not trade exposed. For coffee shops or restaurants facing only local domestic competition, or for public sector workplaces with night shifts, there would be no net benefit to the economy in reducing their labour costs. Instead there would be a drop in aggregate consumer demand as workers’ take-home pay fell. Some individual employers might make super-profits and some State Treasurers might smile more often but these are not net benefits for the economy as a whole.

Gregory’s dismissal of macro-economic solutions reflects a broad sea change in the economics profession over the last 30 years, which has seen micro economic thinking come to dominate policy formulation. As James Galbraith has observed, this has led down many fruitless avenues, such as the ‘skill biased technical change’ explanation for the growth of wages inequality.¹⁸⁵ The greater availability of large-scale unit record datasets, such as HILDA and the ABS CURFs, has encouraged this shift to micro-economic thinking. While the availability of such data is to be welcomed, its slant toward framing problems in terms of labour supply is one of its most unfortunate side-effects. Until the recent AWRS there had been no large scale survey of workplaces in Australia since the last AWIRS in 1995. Datasets which might encourage researchers to frame problems in terms of capital investment and labour demand are notably lacking.