231 It should be noted that this theoretical model operates on the assumption that the employers minimise costs both before and after the minimum wage adjustment, and that workers' skills and level of effort are identical and given exogenously (Brown et al. 1982 at 488). It also proceeds upon the assumption that all workers in the market are covered by the minimum wage. The model has been expanded over time to take in additional considerations or assumptions such as the effect of monopsony, shock effects (in terms of productivity) and two sector models (in which the coverage of the minimum wage is incomplete or where there is queuing for covered-sector jobs). There has also been a theoretical development of the effect of introducing heterogeneous workers into the equation. Doubt has been cast, even at a theoretical level, as to whether labour markets reach a general equilibrium and, if they do, if all parts of the labour force participate (see the discussion of imperfect or segmented labour markets: Gerard Adams, (1987) at 10).
232 These theoretical constructs have been tested over time by empirical studies which, as the AIRC identified, are often directed to the analysis of particular sectors or occupational groups. The empirical studies are directed to testing the conventional proposition that an increase in the minimum wage moves equilibrium backwards along the demand curve for labour (depending upon elasticity for employment) thus potentially reducing the employment of workers affected by the minimum wage adjustment. Nowhere has there been such empirical research undertaken directly on the effect of safety net adjustment changes in the Australian context, let alone an adjustment of the kind sought in this case which may be confined to unincorporated bodies or non-constitutional corporations.
233 As the AIRC has noted in its safety net adjustment decisions for 2003 to 2005, the studies have produced some mixed results which have led to some quite strenuously contested policy questions in economics. It is not without significance that the accepted view of the impact of minimum wages on employment has shifted over time depending upon the development of this empirical research and the academic discussion that is derived from it. A useful snapshot of some of the literature is given by Mark Stewart in his paper "The Impact of the Introduction of the U.K. Minimum Wage on the Employment Probabilities of Low-Wage Workers", Journal of the European Economics Association, March 2004 (at 68) as follows:
A consensus seemed to have emerged by the 1980's that the effect of minimum wages on employment in the United States was negative although probably fairly small. Most of the evidence was based on time-series estimation and much of it on teenage employment, where the effects were felt to be largest. Research findings in the 1990s have blown this consensus apart. On the one hand a growing body of research finds zero or positive employment effects (e.g., Card and Krueger (1994, 1995, 2000) for the United States, the U.S. results in Abowd et al. (2000) and Machin and Manning (1994) and Dickens et al. (1999) for the United Kingdom). On the other hand there is also a body of recent research that finds significant (both statistically and numerically) negative effects (e.g., Kim and Taylor (1995), Currie and Fallick (1996), Burkhauser et al. (2000), Neumark and Wascher (2000), and Neumark et al. (2000) for the United States, the French results in Abowd et al. (2000) and Machin et al. (2003) for the United Kingdom). Thus the employment effect of minimum wages remains a highly contentious issue.
234 We have set out the full references given in that extract (which we have reviewed) in Appendix C. This is not to suggest that the references given by Stewart are exhaustive of the literature on the topic. For example, Charles Brown (1999, at 2154) suggests that a reading of the "new and old evidence" indicates that the short term effect of the minimum wage on teenage employment is small with time-series estimates that centred on elasticity of -0.10 moving closer to zero in samples that included the 1980s. He suggests that one possible reason for this outcome is that the demand for low-wage labour is "just not terribly elastic in the short term" (2157) although some studies do show more significant adverse effects for this group.
235 For his part Stewart made the following conclusions as to the lack of impact of the introduction of the U.K. minimum wage in April 1999:
This paper uses individual-level longitudinal data from matched Labor Force Surveys, the British Household Panel Survey, and the New Earnings Survey panel to estimate the impact of the introduction of the U.K. minimum wage in April 1999 on the conditional probability of subsequent employment among those whose wages would have to be raised to comply with the new minimum. A difference-in-differences estimator is employed using individuals from slightly higher up the wage distribution as the comparison group. The estimated impact of the introduction of the minimum wage on the probability of remaining in employment is insignificantly different from zero for all four demographic groups (male and female adults and youths) and all three datasets. This finding is robust to an extensive range of modifications considered.
The estimated effect is also found to be positive (although insignificant) for both male groups and for young women in all cases. The estimated effect is negative (although insignificant) for adult women in the LFS, but this depends on the construction of the hourly wage rate using usual hours, and is no longer the case if actual hours in the reference week are used. It is also negative for this group in the NES, although even less significant. In the BHPS the estimated effect is positive for this group (but insignificantly different from zero). These findings too are robust to the modifications considered. In summary, the evidence presented in this paper indicates no significant adverse employment effects of the introduction of the U.K. minimum wage in any of the four demographic groups considered or in any of the three datasets examined.
236 As Stewart acknowledges, those conclusions (and, for example, those of Brown (1999)) may be contrasted with yet other studies of recent times which have come to contrary conclusions showing more significant employment impacts of minimum wage adjustments, for example, see the paper by Taylor, L and Kim, T "The Employment Effect in Retail Trade in California's 1988 Minimum Wage Increase" Institute for Research on Poverty, September 1993.
237 Butler, J in "Minimum Wage Laws and Wage Regulation: Do Changes to a Minimum Wage Affect Employment Levels?" (2006) 29 (1) UNSW Law Journal 181 discusses minimum wage impacts in the context of the establishment of the Australian Fair Pay Commission and the Workplace Relations Amendment (Work Choices) Act. Much of the discussion in the paper focuses upon the prospect of decreasing minimum wages (in both real and absolute terms). Despite the appearance of down playing controversy over the impacts of increases in minimum wages the paper, in fact, describes the range of differing views on the topic but concludes, particularly in relation to more vulnerable workers such as teenagers, that significant negative outcomes will be found. The paper does not deal with most of the articles referred to in Appendices B, C and D of this decision and whilst acknowledging components of the AIRC's 2005 Safety Net decision (as to the comparative position of minimum wages and median wages) does not deal with the discussion by the AIRC of academic studies on the impact of minimum wage adjustments. The paper does, however, acknowledge some factors that we would highlight:
1 In classical economic theory, an increase in the minimum wage will only result in disemployment if the resultant wage is greater than the marginal product of the work in question. It is difficult to know to what extent low paid workers in Australia are above the market clearing rate (at 199).
2 In Australia, the coverage of the minimum wage may be heterogeneous. If a minimum wage rises in a particular sector, then economic theory indicates that jobs will be higher in that sector (although there may be some levelling out of the system as a whole (at 194).
3 The greatest controversy lies in the degree of any impact in minimum wage movements on employment.
4 Other labour market regulation measures will have an impact on the effect of minimum wage adjustments, as will general high levels of economic growth or low levels of unemployment in an economy mute any negative effects of the minimum wage adjustment on employment. In this respect, it should be noted that the paper was written before the recent OECD study we shall refer to below which discusses the effects of labour market programs on employment growth, identifies the ongoing controversy over minimum wage effects on employment and suggests a contrary view as to the significance of those adjustments on employment.
238 The most recent discussion of the topic was undertaken by the OECD in its publication OECD Employment Outlook: Boosting Jobs and Incomes 2006. In that publication the OECD reviewed its policy recommendations to reduce unemployment, raise employment and increase prosperity given in its 1994 publication OECD Job Study (see The OECD Job Study: Facts, Analysis, Strategies, Paris, 1994; The OECD Job Study: Evidence and Explanations, Part 1: Labour Market Trends and Underlying Forces of Change, Paris, 1994, and The OECD Job Study: Evidence and Explanations, Part II: The Adjustments Potential of the Labour Market, Paris, 1994. In particular, the OECD adjusted its conclusion in relation to the economic effect of minimum wage adjustment.
239 The OECD first concluded in relation to "in work benefits" programmes that, under certain conditions, a minimum wage, set at an appropriate level, can be one of the options to prevent employers from pocketing the earnings' subsidy introduced by "in work benefits". Secondly, and more significantly, the OECD concluded that the evidence concerning the impact of minimum wages on employment was ambiguous; that the impact of minimum wages on employment may be modest or non-existent, depending upon the level set; and that minimum wages could, in certain circumstances, encourage higher participation in the workforce. There was particular caution expressed about the setting of minimum wages for young employees. The relevant conclusions of the OECD are as follows (at 86-88):
Minimum wages. Simple economic reasoning indicates that a statutory minimum wage or labour costs set at too high a level will become a barrier to employment for low-productivity workers, reducing national output while also frustrating the equity goals motivating these measures. However, pinning down the size of the employment losses that result from minimum wages has proven to be difficult and there is considerable uncertainty concerning how many jobs might be lost due to minimum wages set at the levels actually observed in different countries. Indeed, the empirical evidence concerning a negative impact of minimum wages on employment is mixed, with some studies finding evidence of significant effects, particularly for youth (Neumark and Wascher, 1999; OECD, 1998, Chapter 2), while others do not detect any effects (Card and Krueger, 1995; Dolado et al., 1996; Elmeskov et al., 1998). In Bassanini and Duval (2006), no significant impact of the minimum wage on the aggregate unemployment rate is found. However, some evidence does emerge that higher minimum wages may lower the employment rate of youth ( i.e. the 20-24 age group).
The ambiguous evidence concerning the impact of minimum wages on employment means that it is important for governments making use of a statutory minimum wage to monitor closely whether it is resulting in significant job losses. The fact that a considerable number of studies have found that the adverse impact of minimum wages on employment is modest or non-existent, also suggests that there may be scope to use minimum wages as one part of employment-centred social policy, intended to mitigate poverty while fostering high employment rates (OECD, 1998, Chapter 2 and 2003a, Chapter3). A minimum wage could encourage higher participation, by helping to make work pay for the low skilled. But it probably can only play a supporting role in a broader anti-poverty programme, due to the need to avoid setting it at too high a level. Another important limitation is that a substantial proportion of the workers in minimum-wage jobs are not poor ( e.g. because other family members have earnings). In-work benefits can be much more tightly targeted on low-income families than can a minimum wage, but have other drawbacks ( e.g. their budgetary cost and possible stigma effects). Furthermore, as emphasised in Sub-section 2.2, a modestly set minimum wage may be a useful supplement to in-work benefits, since it limits the extent to which employers can appropriate that benefit by lowering pay levels (Gregg, 2000; OECD, 2005d).
An important consideration in setting the level of the minimum wage is how it interacts with the tax system, since there is evidence that an overly high minimum wage magnifies the negative impact of the labour tax wedge on employment (see Sub-section 3.2). An employer considering hiring a low-skilled or inexperienced worker is likely to compare the worker's expected productivity with the sum of the minimum wage and employer-paid social security contributions. Table 3.10 shows that the minimum cost of labour, as a percentage of labour costs for the average employee, differs significantly across countries with statutory minima, ranging from under 20% in Mexico in 2004 to over 50% in Australia, Turkey and three EU countries. In most cases, the relative cost of employing a minimum wage worker did not change much between 1997 and 2004.
...
Lessons
Recent experience confirms the importance of policies to assure that wages adjust flexibly in response to supply- and demand-side pressures, so as to support high levels of employment in a constantly changing economic environment. The detailed policy recommendations for reforming wage-setting institutions remain largely valid, but there appear to be grounds for introducing some modifications:
...
Minimum wages. Recent experience suggests that a moderate minimum wage generally is not a problem, but that adequate allowance for sub-minima for youth and possibly other vulnerable groups is essential. Another insight is the potential for a well designed minimum wage to contribute to a broader strategy to foster higher employment by guaranteeing that work pays better than remaining on social benefits. However, the danger posed by negative policy interactions has also been confirmed, particularly that between a too-high minimum wage and high rates of labour taxation.
240 We have set out in Appendix D the references referred to in this quotation that have not otherwise been listed in Appendix B.
241 In addition to the theoretical and empirical works we have referred to, we have also had regard to the evidence of Dr Hughes on the question of employment effects. His evidence was given at both the macroeconomic and microeconomic level. At a macroeconomic level, Dr Hughes gave evidence that the claim by Unions NSW was affordable having regard to Reserve Bank of Australia policy. He also concluded that, from a microeconomic viewpoint, the claim was affordable and that no adverse consequences would likely eventuate from the grant of the claim in full.
242 As to the first proposition, Dr Hughes gave evidence that, for over a decade, the Reserve Bank has maintained the view that general wage movements beyond around 4.5 percent a year were inconsistent with its conduct of monetary policy (consistently with an overall CPI inflation within an annual range of 2 to 3 percent on average over the course of the business cycle). He based this conclusion upon a unit labour cost arithmetic, with the trend in underlying non-cyclical labour productivity growth lying at around 2 per cent per year. Thus, a "wage pace" around 4.5 percent would leave unit labour cost growth around the middle of the inflation range. Any higher "wage pace" would push unit labour cost growth towards the high end and eventually beyond the target range.
243 Dr Hughes thereby concluded that a 4 per cent wage rise would be consistent with the Reserve Bank's "instructions from the Commonwealth Government" leading to a consumer price inflation around the low end of the target range unless offset by other non-wage influences. He contended that " the bank can be presumed to provide sufficient liquidity in other monetary support to accommodate inflation within a 2 to 3 percent target range and therefore the Unions NSW claim was consistent with Reserve Bank macroeconomic control of the economy ".
244 The cross-examination of the witness on this aspect of his evidence was essentially to suggest, erroneously in our view, that his statements as to the policy and position of the Reserve Bank could not be substantiated either in writing or otherwise. His evidence, in this respect, should be accepted.
245 His microeconomic analysis was more controversial. Dr Hughes addressed squarely the issue that, by the wage claim, workers " might be pricing themselves out of the market". He described that issue in this way: "If substantially higher than average wage gains were to be awarded it is possible that consequent relative price changes for the products they produced might lead to a lessened consumer interest and the decline in demand for their services". However, Dr Hughes rejected that outcome for the present proceedings as he considered that the claim sat " right in the middle of a very narrow range within which wage fixing has been occurring". In this he was referring to the latest wage price index (ABS Cat No.6345), which showed that the Australia- wide private sector wage rates gained 4 percent over the year to the March quarter 2006. He concluded that unlike periods of higher unemployment earlier in the decade there has recently been a " considerable evening up in the pace of wage increases. Thus over the year to the March quarter 2006 the lowest increase amongst the 15 industry groups into which the private sector index is disaggregated was 3.01 percent (personal and other services). The largest gain was 5.0 percent in the construction industry ".
246 Dr Hughes then expressed why this outcome would support his conclusion that there would be no adverse microeconomic consequences as follows:
Among private sector industrial groupings suggested by the Commonwealth government as most likely to be affected by the outcome of this case the retail trade wage index grew 4.05 per cent, that for the hospitality activities of accommodation, cafes and restaurants by 3.19 per cent, property and business services by 3.96 per cent, health and community services by 4.29 per cent, education by 4.05 per cent and manufacturing by 3.94 per cent.
Thus the 4 per cent claim appears unlikely to disturb industrial wage relativities being extremely close to the centre of the concentrated distribution of wage gains now underway.
247 The primary thesis developed by Dr Hughes was that the economic discussion of the impact of the minimum wage adjustments to employment had concentrated upon relative changes in the price of labour or wages and not as he contended would occur in the present circumstances where, because of shifts in the wage price index, the grant of the claim would produce no relative price adjustment.
248 The essential contradiction proposed in cross-examination of Dr Hughes' evidence and in submissions put by Employers First was that the wage adjustments reflected in the wage price index for the relevant period merely reflected the safety net adjustments that had been awarded by the Commission in the previous period. Thus, there would be a relative wage or price movement.
249 We propose to set out part of the cross-examination of Dr Hughes to illustrate the approach taken:
Q. Having regard to that, you would agree at least part of the percentages you refer to there take into account the wage increases awarded by the state or Federal Commission over the year?
A. Part of those increases indeed would be reflecting arbitrated increases and my understanding is however the arbitrated increases from the safety net minimum wage changes were very much in the minority.