Workplace relations reform: examining the economic data

What light (if any) does the economic evidence shed on the arguments for or against the Government’s proposed changes to Australia’s industrial relations system, especially to “unfair dismissal laws”, and the minimum wage?

As Mark Latham noted in his Diaries, “wages … are such an emotional issue it is impossible to shift opinions, even based on sound empirical research”.

By way of background, there has been a significant amount of change in the direction of “greater flexibility” in the Australian labour market over the past decade, largely as a result of reforms introduced by the Keating Government in 1993, and by the Howard Government in 1996. In particular:

• the proportion of employees whose pay is set by awards only has dropped to 20 per cent, from an estimated 68 per cent in 1990;
• fewer people are members of a trade union than at any time since 1958. The proportion of employees who are trade union members has dropped from 37.6 per cent in 1993 to 22.7 per cent in 2004. Only 17.4 per cent of private sector employees are members of a trade union, down from 27.5 per cent in 1993;
• more people are now self-employed (as owner-managers of incorporated or unincorporated enterprises) than are members of trade unions; and
• more people now directly own shares than are members of trade unions.

While Australia’s once highly centralised and co-ordinated wage-fixing arrangements are still more centralised than in other English-speaking nations (other than Ireland) or the Asian members of the OECD, Japan and Korea, Australia now sits towards the lower end of the OECD spectrum on both these scores.

Since the 1993 Keating Government reforms, Australia’s labour market has delivered strong jobs growth (albeit at a slightly lower rate than during the previous 12 years); falling unemployment and under-employment; rising productivity (at least until the end of 2003, since when productivity growth has gone into reverse) and real wages; well-behaved real unit labour costs (rising by 1.3 per cent a year, on average, over the past 12 years); and (at least by Australian standards) low levels of industrial disputation.

How much of this is due to changes in industrial relations arrangements, as opposed to 14 years of more or less continuous economic growth - the longest period unpunctuated by at least two consecutive quarters of negative growth in Australia’s history - is difficult to ascertain. Indeed to the extent that changes in industrial relations arrangements have lessened the probability of a “wages break-out” as the economy has approached “full employment” over the past year - as has occurred at the same stage of each of the three previous business cycles - then they may have contributed to prolonging the current expansion.

International data on the impact of industrial relations arrangements on labour market performance is not entirely clear. However it is clear that some countries are able to combine relatively centralised wage setting arrangements with rapid employment growth and low unemployment

Employment protection regulation and labour market outcomes

Australia’s employment protection legislation is “one of the least restrictive” in the OECD. At the time of its most recent survey, only the US, Canada, the UK, Ireland and New Zealand had less strict employment protection legislation than Australia.

The Government nonetheless argues that Australia’s “current unfair dismissal laws not only discourage job creation across businesses, but impose costs on businesses - small, medium and large alike”. The Prime Minister has described “a culture of complaint and litigation that loads extra costs onto those who society relies on to create wealth and jobs by taking risks with their own livelihoods” and said these laws “hurt the good staff, they discourage small firms from taking on more people and they are a prime example of where over-regulation has worked to the detriment of both business and also employees”.

The Howard Government is not the only contemporary government which takes this view. The Socialist Government of Spanish PM Rodríguez Zapatero is also seeking to relax rules for dismissal of workers.

The idea that legislation which increases the costs (both pecuniary and in terms of management time) of dismissing employees has an adverse impact on hiring enjoys fairly widespread support among economists. Earlier this month the Chairman of the US Federal Reserve, Alan Greenspan, said:

Many working people, regrettably, equate labour market flexibility with job insecurity. Despite that perception, flexible labour market policies appear to promote job creation, not destroy it. An increased capacity of management to discharge workers without excessive cost, for example, apparently increases companies’ willingness to hire without fear of unremediable mistakes.

Economic theory generally suggests that employment protection legislation tends to depress both hiring and firing. This prediction is generally supported by empirical evidence. That is, employment protection legislation does protect existing jobs, at least to some extent, while also adversely affecting new job creation.

It may well be that employment protection legislation inhibits the ability of businesses, and the economy more broadly, to respond to technological change (which may have to be accommodated through worker turnover): although it could also be argued that protected workers are less resistant to technological change and more willing to re-train.

The finding that employment protection laws reduce both hiring and firing also means that any effect on unemployment depends on which of these is affected more.

The evidence is not clear. However, studies do provide some support for the notion that employment protection legislation may adversely affect the prospects of young people and women - who at any given time are more likely to be seeking to enter or re-enter the workforce - while positively impacting the stability of employment of prime-age men.

It’s worth noting that the Government proposes to retain the existing unfair dismissals laws for employers with more than 100 employees (who account for about 38 per cent of employees), whilst abolishing them for employers with fewer than this number - even though these laws impose costs on large businesses as well as small ones.

If the Government accepts there is a case for employees to be protected against unfair dismissal - which it presumably does since it is not abolishing the law for all employees - then it is a little strange that the employees who are to remain protected from unfair dismissal are actually those least likely to be unfairly or capriciously dismissed. That’s because large corporations usually have HR departments whose functions include ensuring that their managers adhere to “procedural fairness” when dismissing staff, and because large corporations are generally more conscious of the reputational risk involved in unfair dismissals cases.

That being so, a more creative response on the part of the Labor Party to this element of the Government’s proposed reforms might have been to replace the word “under” in the relevant section of the legislation with the word “over”: so that those working for employers with fewer than 100 employees continued to be protected against something that was more likely to happen to them, while larger employers were exempted from those provisions.

New procedures for setting the minimum wage

Australia’s minimum wage, ($484.40 per week or $12.75 per hour for a 38-hour week) is high, relative to median wages, by international standards. According to the OECD, Australia’s minimum wage (in 2002) was equivalent to 58 per cent of the median earnings of full-time adult employees, higher than any others except France. Britain’s Low Pay Commission reported that Australia’s 2004 minimum wage was 58.8 per cent of median full-time adult earnings, highest of the 14 OECD economies covered.

As the Melbourne Institute’s Mark Wooden points out, “for most economists, minimum wage to median earnings ratios of close to 60 per cent are indicative of a system that prices many of the unemployed out of the labour market”.

David Card and Andrew Krueger, authors of the now well-known study ten years ago which suggested that increases in the minimum wage did not have the adverse impact on employment traditionally ascribed to them, acknowledge that their findings only applied to levels of the minimum wage that existed in the US (a little over 30 per cent of median full-time adult earnings) and that, beyond some point, minimum wage increases must harm employment.

The UK Low Pay Commission believes that the British minimum wage - first introduced at £3.60 per hour (for persons aged 22 and over) in April 1999, and since increased by 35 per cent to £4.85 per hour, equivalent to 43 per cent of median full-time adult earnings - “does not seem to have had any significant negative impact on the labour market any significant negative impact on the labour market”, although two research projects which they commissioned found “small negative employment effects in those sectors and among those most affected by the minium wage”. The commission also found that the minimum wage has “had a major beneficial impact on the aggregate earnings of women” without “harming women’s job prospects”.

In the Australian context, recent studies suggest increases in the minimum wage have adversely affected employment. A 2003 study of the impact of six increases in the WA minimum wage between 1994 and 2001 by Andrew Leigh suggested that a 1 per cent increase in the minimum wage leads to a 0.15 percentage point fall in employment. Don and Glenys Harding, who were commissioned by the Commonwealth Government to survey the impact of “safety net” adjustments on small and medium businesses, found that a 1 per cent increase in the minimum wage reduced employment by about 0.2 per cent. Both of these studies were the subject of methodological criticisms and were given short shrift by the Australian Industrial Relations Commission in this year’s safety net review, as was most of the international evidence on this subject.

Nonetheless, the commission has allowed the minimum wage to decline from 60.6 per cent of median earnings in 1996 to 58.4 per cent in 2004, or from 51.9 per cent to 49.2 per cent of full-time adult average weekly ordinary time earnings over the same period.

In some ways, the debate over whether the minimum wage should be set by the AIRC, or by the Fair Pay Commission envisaged by the Government, seems to be a cover for a debate about whether this trend should be accelerated or not.

Of course the Australian minimum wage has its origin, nearly a century ago, in the notion of an income sufficient to allow a man to support his wife and three children in “frugal comfort”, which may explain why the Australian minimum wage has historically been higher relative to median earnings than in other countries.

(It is interesting to recall that the 1907 Harvester Judgment was premised on a creative extension of the Commonwealth’s excise power, just as the Howard Government’s WorkChoices package is partly grounded on an extension of the corporations power.)

However, the labour market is no longer comprised predominantly of married men supporting stay-at-home wives and dependent children. In reality, more than half of low-wage earners are located in households in the top half of the income distribution, while the bottom quintile is dominated by those in receipt of social security payments. For this reason, many economists (myself included) believe that support for low-income earners is more appropriately provided through the income tax and social security systems than via the industrial relations system. In that context, I think it is unfortunate that the Government’s WorkChoices proposals haven’t been accompanied by reforms in these areas.

Summing up

My interpretation of the economic data is that it lends some support to the Howard Government’s proposed reforms, but that support is neither unequivocal nor incontestable. The economic evidence also suggests that goals of increasing participation in the labour force, reducing long-term unemployment and reducing poverty require reforms in other areas in addition to workplace relations, including tax, social security, and education and training.

In the end, attitudes to the Government’s proposed reforms are probably informed more by politics than economics, and I doubt there is anything any economist can say to alter that.