Proposals for a Wage Freeze and Tax Credits: Will Subsidising Low Wage Jobs Solve Unemployment?

Research Paper 29 1998-99

Ian Watson
Economics, Commerce and Industrial Relations Group
29 June 1999


Major Issues Summary
A wages freeze

How it might work
The employment consequences

Aggregate wage restraint
Relative wage restraint

An earned tax credits system

How it might work
Has it worked elsewhere

The wider context- a discussion

The repercussions of reducing wages

An alternative approach for dealing with unemployment

Major Issues Summary

Despite six years of strong economic growth in Australia, unemployment remains very high and presents the most important social and economic policy issue facing the country. In recent times a group of economists, known as the 'Five Economists', have proposed that we solve the problem of unemployment by bringing in wage cuts for the low paid workforce combined with a tax credits scheme to compensate for the wage cuts. This research paper examines this proposal in detail and finds that:

  • The proposal largely ignores the problem of long-term unemployment. It overlooks the fact that only a small amount of any likely growth in employment which might result from wage cuts would lead to jobs for this disadvantaged group. Sound policy initiatives should specifically target assistance to the long-term unemployed, otherwise their prospects for employment grow worse over time.
  • Most labour market researchers recognise that the labour market in the 1990s is very different to that which operated in the 1980s. Since the recession of the early 1990s there have been major changes in how employers use labour. A large part of any likely increase in the demand for labour would lead to more hours of work for the existing workforce, either through work intensification or through under-employed workers taking on more hours of work. Moreover, many of the new jobs created would go to new labour market entrants, not to those currently unemployed.
  • The proposal amounts to an expansion of the low wage sector in Australia and runs directly counter to economic strategies which emphasise the 'high road' to the future, the importance of building the 'clever country', and a preference for high productivity workplaces. These approaches emphasise investment in training, and in research and development as the basis for economic progress. Whilst these strategies do not directly benefit the unemployed, particularly the long-term unemployed, they do generate wealth for the country and promise a more secure future. In turn, such wealth and security make possible direct labour market interventions which target assistance to the long-term unemployed.
  • The proposal entails restraint in both aggregate wages and relative wages. The former amounts to an 'incomes policy', but the mechanism proposed for implementing it is seriously deficient. The Five Economists suggest that curtailing award safety net increases would lead to a reduction in aggregate wages growth but this proposition is overly optimistic. It ignores the extent to which wages growth is largely unregulated amongst the two-thirds of the workforce outside the Award stream. Under these conditions, estimating the size of any reduction in aggregate wages growth is highly speculative.
  • Restraint in relative wages aims to make less skilled workers more employable by making their labour cheaper. This paper examines the logic of this proposition by looking at overseas research on the minimum wage. This debate has focused on whether raising the level of the minimum wage leads to increased unemployment for those workers on the minimum wage. This is basically the same logic which says that reducing wages for the less skilled leads to better employment outcomes. This paper finds that the theoretical debate in this area is inconclusive and depends heavily on what kind of assumptions are made about how the labour market works. In terms of empirical research, the conventional wisdom has been that raising wages does lead to increased unemployment, but research during the 1990s has seriously eroded confidence in this position. Reaching an unequivocal position on the relationship between relative wages and unemployment is just not possible, and Australian economists need to keep this in mind when proposing relative wage restraint in this country.
  • Were a large reduction in relative wages to take place, there is no guarantee that a large number of new jobs would be created because of 'displacement' effects. Employers would simply substitute 'subsidised' workers (those on the lower pay) for 'unsubsidised workers'. On the other hand, a large reduction in relative wages would lead to a major decline in living standards amongst the low paid workforce and an increase in the size of the 'working poor' in Australia.
  • In recognition of the problems of the 'working poor', the Five Economists have also suggested an earned income tax credits scheme to compensate low paid workers for their wage cuts. This paper assesses this proposal by looking at the American experience of the EITC. The lesson for Australia from the United States experience is that earned tax credits are a successful response to the problem of welfare poverty traps and may warrant further examination. Earned tax credits are not, however, a suitable partner for low wage jobs since they compound many of the problems to be found in the low wage sector. Instead, if earned tax credits can be used as a complement to wage improvements at the bottom of the labour market, then optimum outcomes all round can be expected. In this respect, the Five Economists are mistaken to link their tax credit schemes with an expansion in low paid work. There are too many serious shortcomings attached to low paid work for Australian policy makers to consider expanding the low wage sector here.
  • By way of an alternative to wage cuts, this paper proposes a more interventionist approach to the problem of unemployment. It examines the work of a number of economists who have argued for policies of public sector job creation and targeted labour market programs. The paper finds considerable merit in these ideas and suggests that on equity grounds these proposals are preferable to those of the Five Economists.


Unemployment in Australia remains the most important social policy problem of this decade. It also presents something of an economic puzzle: why have six years of strong economic growth not translated into strong employment growth? The unemployment rate has remained stuck above 8 per cent for most of the 1990s, and has only recently dipped below that. Public policy has focused again on the question of what can be done to reduce unemployment in Australia, preferably to levels below 5 per cent within a few years.

Over the last two years, various proposals have surfaced for reducing wages at the lower end of the wages spectrum in the hope that this will increase labour demand and boost employment growth. To offset the impact of such cuts on living standards, these proposals have included a system of tax relief for low income families. These proposals have been strongly associated with 'Five Economists': Peter Dawkins and John Freebairn (Melbourne University), Ross Garnaut and Michael Keating (ANU), and Chris Richardson (Access Economics). Some of these economists have penned various articles dealing with these proposals, and collectively they wrote an open letter to the Prime Minister in October 1998.(1) The letter has generated a level of interest from economists as it suggests the use of institutional control over (low) wages in an era when such control runs counter to current philosophies. However the proposal is worth evaluating as major players such as the Business Council of Australia have come to support the wage freeze proposal, affording the proposal extra weight. The full text of this letter can be found in the Appendix to this paper. The core suggestions in this letter were:

  • a freeze in award wage safety net increases ('Living Wage' adjustments) for four years, in the hope that the growth in average wages would decline and employment would thereby increase
  • compensation for low-income families using a system of earned tax credits.

In this research paper I look at both of these issues. I examine how a wage freeze might work, and what its consequences for employment would be. I also look at the tax credits proposal and examine its likely impact. Finally, I offer some suggestions for dealing with the problem of unemployment which do not entail reducing wages for those at the lower end of the wage spectrum. These views represent those of the author, and the intention is to put an alternative to the Five Economists' proposal.

A wages freeze

How it might work

To the layperson, the most puzzling aspect of the Five Economists' proposals is why they are asking the low paid workforce to shoulder the burden of solving unemployment. Australia's high levels of unemployment are largely the result of profound economic restructuring over the last quarter century. The unemployed are the 'losers' in this story. Yet there have also been 'winners', particularly highly skilled workers and managers who have seen their job opportunities and incomes increase considerably over this period. In terms of social equity, it seems obvious that the people who should be asked to 'pay' for solving the problems of unemployment should be amongst the 'winners'. Yet we are being asked by the Five Economists to focus on those at the bottom of the labour market, not those at the top. To understand this anomaly involves taking an excursion through some economic theory.

The theory behind the Five Economists' proposals is traditional text book economics: in a competitive labour market, a reduction in wages (the price of labour) should lead to an increase in employment (the demand for labour). Before we explore this idea more fully, it is important to draw a distinction between aggregate wages and relative wages. Aggregate wages reflect the average price of labour across the economy. Relative wages reflect the prices of labour for particular sub-groups in the workforce, such as less skilled workers, or teenagers, for example.

How might reducing aggregate wages growth lead to employment growth? There are two main reasons for why this can happen. Reducing wages growth changes 'factor shares' in the economy. This means that the owners of capital gain a greater share of the output than do the sellers of labour. This increased share may lead to an increase in productive investment, such as new plant and equipment, by the owners of capital, and this in turn may lead to an increase in demand for labour.(2) Secondly, more expansionary macro-economic policies by governments are possible when wage restraint is in place because there is less risk of inflation 'taking off'. Both of these reasons were evident in the thinking behind the 1983 ACTU-ALP Accord.

The case of relative wages relies heavily on textbook economic theory. Unemployment for certain groups of workers occurs because the market 'fails to clear' at prevailing wage levels. This argument suggests that low skilled workers, for example, may be selling their labour at a price which is too high. Consequently, reducing their wages will make them attractive to employers and result in an increase in their employment levels. An important sub-plot in this line of argument is that relative wages may be too high because 'institutional' factors-such as trade unions or industrial relations tribunals-prevent relative wages falling to a market clearing level. Labour market 'deregulation' aims at winding back these institutional factors. This is meant to make the labour market more 'flexible' in the sense that falling demand for certain kinds of labour should lead quickly to a drop in its price and a smooth readjustment in employment levels. It is important to note that this line of thinking entails large cuts in actual wage levels, not just minor reductions in wages growth.

The Five Economists' proposals aim to reduce aggregate wages growth, though there are suggestions that an impact on relative wages growth may also take place. They argue, for example, that:

Earnings would be suppressed more for low-skilled than for high-skilled workers. It is likely that lower growth in wages for the low skilled will be more helpful than lower growth in wages of the high skilled, as the unemployed are disproportionately low skilled.(3)

Commentators like Ross Gittins have embraced the Five Economists' proposals as an incomes policy, a kind of reincarnation of the Accord. On the other hand, proponents of labour market deregulation, such as the Business Council of Australia (BCA), have welcomed the Five Economists' proposals because they see wage freezes, coupled with earned income tax credits, as a vehicle for increasing wage flexibility without dramatically increasing poverty. The BCA looks to the United States as an example of a country with a low unemployment rate, deregulated labour markets, and high levels of earnings inequality and asks the question:

How far can we go in reproducing the best aspects of the American labour market-in particular its low unemployment rate-whilst protecting those on low incomes?(4)

The more extreme position on labour market deregulation is represented by Des Moore, who argues for the abolition of the Australian Industrial Relations Commission (AIRC) in its current form and for the removal of any impediments to the free operation of market forces.(5) He also sees a role for tax credits in alleviating poverty:

If wages were allowed to find their own level and the resultant labour market earnings proved to be socially unacceptable, they could then be supplemented through, say, tax credits to the extent that the existing means tested benefits did not provide appropriate top-ups.(6)

The Five Economists themselves propose a four year freeze in safety net increases (the AIRC's Living Wage adjustments). At present all award workers are entitled to receive these safety net increases, with those workers who are unable to achieve any wage increases through enterprise bargaining entitled to the full amount of the increase. Where enterprise agreement increases have taken place, the safety net increases are 'absorbed'. The Five Economists calculate that a freeze would lead to a reduction in the growth of real (inflation adjusted) award wages by about 10 to 12 per cent. They then assume that this would flow through to a reduction in average wages growth of about 3 to 4 per cent (than otherwise would have occurred). They make use of work by Reserve Bank economists, Guy Debelle and James Vickery, to argue that a reduction in wages growth of this amount would lead to a reduction in the unemployment rate of about 11/2 to 2 percentage points.(7)

Before looking more closely at whether such a wage freeze would have the desired employment outcomes, it is worth considering the issue of relative wages. The dilemma for advocates of deregulation is that a large reduction in relative wages-perhaps as high as 30 per cent-would be necessary for there to be any impact on labour demand. This would involve reductions in actual wage levels, not just reductions in wages growth. It is important to keep in mind that despite their passing reference to the low-skilled, the Five Economists are essentially concerned with aggregate wages outcomes, and they do not propose large reductions in wage levels. They do, however, discuss labour market deregulation and cuts in relative wages for the less skilled workforce.(8)

Unfortunately, the current debate constantly moves between the two extremes: an incomes policy focussed on curtailing aggregate wages outcomes, and a policy for labour market deregulation focussed on large relative wage reductions. The Business Council of Australia, in particular, has muddied the waters by promoting the Five Economists' proposals as a means of achieving 'relative wage flexibility'.(9) Consequently, when assessing the likely employment outcomes of cutting wages it is necessary to deal with both aggregate and relative wages.

The employment consequences

Aggregate wage restraint

Proposals to reduce aggregate wages growth in a co-ordinated fashion are essentially incomes policies, economic tools which have been around for many decades. The dilemma for advocates of any incomes policy in Australia in the 1990s is that the current system of enterprise bargaining means that the government can no longer control aggregate wages outcomes in a predictable fashion. Between one quarter and one third of the workforce is covered by awards, so reducing wages in that sector would have some kind of impact on average wages growth. But how big would the impact be? As mentioned earlier, the Five Economists proposed a four year freeze in safety net increases leading to a reduction in the growth of real award wages by about 10 to 12 per cent. This in turn was assumed to result in a reduction in average wages growth of about 3 to 4 per cent. In other words, they use the same ratio of award coverage to calculate the size of the aggregate wages impact.

Unfortunately, simple arithmetic may not be a good guide to follow. Since the decentralisation of wage fixing in the early 1990s, enterprise agreement wage increases have consistently run ahead of award wage increases. Between 1991 and 1997 the cumulative gain in wages outcomes for workers on enterprise agreements was between 26 and 30 per cent, whereas for those reliant on awards it was only about 8 per cent. What this means is that the Five Economists may be right in their calculations about how much wage restraint is possible in the award-only sector, but the relationship of this restraint to overall wages growth is largely unknown. Simply stating that average wages growth will be 3 to 4 per cent below what it would otherwise have been, is an insufficient basis for calculating the economic impact. If the non-award sectors grow by 12 per cent over the four years, then restraint in this range (3 to 4 per cent) will be quite impressive, but if the growth in the non-award sectors is 24 per cent, then this kind of award restraint will be marginal. The improvement in employment which might come about from wage restraint will be sensitive to the total wages bill in the economy, not from something happening on the margins.

It is worth mentioning at this point that this schism between the award and non-award sectors might make it impossible to achieve even modest aggregate wage restraint from an award pay freeze. Joe Isaac, for example, argues that the growing gap between award rates of pay and enterprise agreement rates of pay would widen considerably under the award freeze proposal. As a result:

Strong pressure to seek pay increases through overaward payments is likely to build up. No constraints are imposed by the Commission against such pressure; nor on those under enterprise agreements. Thus, the freeze and the reduction in real wages would be compromised to the extent that these pressures result in pay increases, especially when unemployment begins to fall.(10)

Nevertheless, if we assume that aggregate wage restraint is possible, and that it is sufficient to generate strong growth in labour demand in the economy, we still need to ask whether this would actually lead to new jobs becoming available. Most labour market researchers recognise that the labour market in the 1990s is very different to that which operated in the 1980s. Since the recession of the early 1990s there have been major changes in how employers use labour.(11)

One of the major spin-offs to enterprise bargaining has been a dramatic increase in what is called 'numerical flexibility', basically, increasing the utilisation of the existing workforce by modifying hours of work. This can include increasing the span of hours over which workers are expected to work, annualising hours of work so that overtime payments or other penalty rates are avoided, introducing 'split-shifts' and eliminating 'dead time' throughout the working day. This increase in numerical flexibility makes it possible for employers to meet an increase in their demand for labour by intensifying the workload of their existing workforce, rather than taking on new workers. It also means that employers have greater flexibility in the allocation of their labour force-deciding when they should work and for how many hours-and therefore less need to call upon new employees. Even for employers who still operate with traditional overtime systems, it is usually more economical for them to respond to increased labour demand by offering their existing workforce overtime, rather than taking on new workers.

The dramatic rise in non-standard employment during the 1990s has also had an impact on how employers utilise labour.(12) Contracting out work to labour hire agencies has become a widespread practice for dealing with peaks in product demand, or other fluctuations in work loads. As a result, there has been a decline in the recruitment of new workers, with firms instead using labour hire arrangements (often accompanied by a reduction in the existing workforce). While these developments have led to an increase in employment amongst the labour hire firms, the net impact is an overall reduction in employment opportunities. A further complication is introduced by the problem of under-employment, whereby members of the existing workforce want to work more hours than are on offer from their employers. In the late 1990s, about six per cent of the workforce are currently under-employed. In other words, it is open to employers to offer more hours to sections of their existing workforces.

Consequently, a reduction in aggregate wages growth in the late 1990s might well translate into increased investment and increased economic activity, leading to increased labour demand, but not all the extra hours created will be worked by new employees. The existing workforce will swallow up a large part of the extra hours of work that will become available. This issue of hours is ignored by the Five Economists in their calculations.

Finally, the last step in the logic of this argument concerns those workers currently unemployed. The question here is: how many of the new workers required will come from the ranks of the unemployed-particularly the ranks of the long-term unemployed-and how many will come from new entrants to the labour market? This proved to be one of the dilemmas of the 1980s. Successful employment growth in that decade did not reduce the unemployment rate as much as was expected. The participation rate-which measures how many people of working age are in the labour force-grew solidly during the 1980s because strong economic growth encouraged more people to look for work. As a result, the long-term unemployed found themselves at the back of an even longer queue. The Five Economists do recognise this phenomena, and in their calculations they assume that about half of the new jobs would go to new labour market entrants and half would go to the existing unemployed.(13)It is partly for this reason that the Five Economists also stress the importance of education and training and a 'systematic approach' to labour market programs.

The conclusions which we can draw from these sober facts about the labour market are:

  • the likely impact of a reduction in aggregate wages growth flowing from a freeze in award wages might be much weaker than that presumed by the Five Economists because of the influence of non-regulated sectors outside the award stream
  • a large part of any growth in employment (in hours of available work) would not necessarily translate into additional jobs for new workers
  • a considerable proportion-as much as a half-of any new jobs which are created would go to people currently outside the labour force, rather than to those workers currently unemployed.

What this means is that the Five Economists' proposals are very 'blunt' policy instruments for dealing with the core problem of unemployment: the long-term unemployed. Currently, about one quarter of a million people are long-term unemployed and about another 120,000 want to work but have given up hope of finding a job (so-called 'discouraged jobseekers'). These people remain the most disadvantaged group in our community. Their need for work, and for decent incomes, stands out as the most pressing social policy concern for the late 1990s. Most labour market researchers accept that sound policy initiatives should specifically target assistance to the long-term unemployed, otherwise their prospects for employment grow worse over time.(14)

Relative wage restraint

Does a lack of relative wage 'flexibility' in Australia contribute to unemployment, particularly amongst less skilled workers? The Five Economists assume that it does, but the evidence appears slim. Bob Gregory, for example, finds little evidence to support this view and presents data which suggests that the Australian situation parallels that found in the United States (where unemployment is at a record low).(15)

Even if Australia does not suffer from a lack of 'flexibility' in relative wages, isn't it the case that cutting wages at the bottom of the labour market will still improve employment opportunities for less skilled workers? The Five Economists and the BCA certainly think so. But answering this question is not as straightforward as it seems. Most of the recent economic debates concerning low wages and unemployment have actually explored the inverse proposition: that raising the minimum wage would lead to higher unemployment. Nevertheless, the logic of the argument-that there is a clear causal relationship between relative wage levels and employment levels-is the same. For this reason it is worth examining the minimum wage literature for the light it throws on this issue.

Argued at a theoretical level, changing wage levels can either increase or decrease employment. A great deal hinges on which assumptions of the textbook theory are applied. In the case of minimum wages, the 'pure' case tells us that if a minimum wage is fixed above the market-clearing wage, then labour demand will drop and employment will also fall. This 'pure' case, however, assumes that labour markets are competitive, that people have perfect information, that labour is homogeneous (all the same, and hence interchangeable), and that labour is perfectly mobile. These are obviously a set of assumptions which rarely apply in the real world.

Creating more realistic cases requires dropping various assumptions. If the labour market is not perfectly competitive and there is a monopsonistic employer (when firms have more power to set wage levels and workers have little bargaining power), then imposing a minimum wage can actually raise employment levels. This possibility essentially relates to staff turnover and shortages, and is in the context of a small wage increase.(16) If the assumption that all workers are homogeneous is relaxed, then reducing wages can have different outcomes depending on the skill levels of the workforce.

As well as relaxing assumptions, we also need to consider the wider context. If we take account of the consumer behaviour of the workforce, then raising wages can lead to increased employment because of greater consumer demand in the economy. Sometimes there is also a time dimension involved in this kind of modelling, with the short-term impact being different to the long-term impact.

The theoretical judgment about wage levels and employment levels is clearly inconclusive. The final position you arrive at depends entirely on what assumptions you work with. This has not stopped theoretical disputes from being influential. American economists, Card and Krueger, have observed that 'economists' views on the minimum wage are based largely on abstract theoretical reasoning, rather than on systematic empirical study'.(17) A group of European researchers have also noted:

Predictions of economic theories are almost always sensitive to assumptions. We are surprised by an unconditional claim [that raising the minimum wage automatically leads to unemployment] ... and sceptical that anyone actually believes it. Yet it pervades the analysis of the minimum wage.(18)

It seems clear then that we should examine empirical studies of the impact of minimum wages if we want to move beyond the impasse which purely theoretical debates engender. Unfortunately, the research literature is far from unanimous. The conventional wisdom in the minimum wage literature is that raising wages leads to increased unemployment, but research during the 1990s has seriously eroded confidence in this conventional wisdom.

One of the classic overviews on the impact of minimum wages carried out in the early 1980s concluded that a 10 per cent increase in the minimum wage would have an adverse impact on employment of between 1 and 3 per cent.(19) The authors of this study did concede, however, that their conclusion was not based on a large body of research because 'one could argue that there really are not 25 independent studies'. Their use of the same labour force data and the similarity of their methodologies meant that most of the studies simply replicated each other.

More sophisticated research began to emerge during the 1990s once better data became more commonly available. Researchers began to make use of longitudinal data to track the labour market fortunes of individual workers over time, testing to see whether those workers who gained minimum wage increases found their probability of future employment declining as a result.

Other researchers pioneered a 'natural experiments' approach which examined the impact of minimum wage on employment levels by comparing similar businesses in neighbouring states, where one state raised the minimum wage and the other didn't. These studies-which strongly rejected the conventional wisdom that minimum wages reduced employment-generated much debate during the mid 1990s and remain a source of contention.(20)

Recent research in Europe has also challenged the conventional wisdom. Studies of the impact of the minimum wage in France, the Netherlands, Spain and the United Kingdom failed to support the premise that raising wages invariably increases unemployment:

it is surprisingly hard to find strong evidence of any adverse employment effects of minimum wages in situations where many commentators are firmly convinced (rightly or wrongly) that the job losses exist.(21)

On the other hand, other research in France using longitudinal data found a strong relationship between minimum wage increases and declines in future employment probabilities.(22)

A recent OECD summary of the literature showed that over 20 major studies had been carried out during the 1990s examining the impact of the minimum wage on employment. The results of these studies varied considerably, but there were no clear, unambiguous findings. The OECD itself concluded tentatively that:

young workers may be most vulnerable to job losses at a high level of the minimum wage. There is less evidence available on the employment effects, if any, for other groups such as women and part-time workers.(23)

In summary, the theoretical arguments about wages and unemployment fail to resolve the issue, while the empirical research is divided.

One of the main problems with the minimum wage literature is that is seeks to find a generally acceptable and universally applicable answer to the relationship between wages and employment. Such a goal is elusive, if not utopian. There are so many unique factors in a local, or even national economy, that universally applicable findings are impossible to obtain. This does not stop economists trying to find the 'true' elasticity of demand for low wage labour (the responsiveness of employer demands for labour following changes in labour costs). In some cases, economists review the current literature and then average out the various elasticities to arrive at an estimate of what the 'real' elasticity might be: 'Taking all the studies for the many countries together, a "best guess" for the long-run constant-output labour-demand elasticity based on this literature is -0.30' (meaning that a 1 per cent rise in wages results in a 0.3 per cent fall in employment)(24) These kinds of guestimates are common in the economics literature, but they really have little credibility for understanding the real world, particularly labour markets in specific countries like Australia.

Despite these uncertainties about the elasticity of labour demand, it is possible that a very large wage cut for low-paid workers-perhaps over 30 per cent-could lead to an increase in jobs at the bottom of the labour market.(25) But would these jobs be 'new jobs', or would we see a displacement effect occur whereby employers simply substituted 'unsubsidised' workers with 'subisidised' workers? (We have to keep in mind that a large relative wage cut is essentially a subsidy to the employer for that particular category of labour.) Research which has evaluated the effectiveness of those labour market programmes which make use of wage subsidies has shown that only between 10 to 20 per cent of subsidised jobs lead to a net increase in employment.(26) Consequently, the prospects for relative wage cuts leading to large numbers of new jobs is not promising.

On the other hand, a wage cut of around 30 per cent would lead to a major decline in living standards amongst the low paid workforce. This leads to the obvious question: would Australia see the emergence of a large population of 'working poor'? Those who argue that this would not happen make two main points:

  • not all low paid workers live in poor families;
  • a system of tax credits could protect the living standards of low income families.

I examine both these issues in the next section, which deals with the second part of the Five Economists proposals.

An earned tax credits system

How it might work

The Five Economists recognise that the Australian Industrial Relations Commission would be unlikely to agree to a freeze of award wages without some other form of compensation for low-paid workers.(27) This leads them to propose a scheme whereby low wage earners in low income families would receive an earned tax credit as an alternative to a Living Wage increase. In its first year this tax credit would be equivalent to about 2 per cent of the wage of a low paid worker. The idea of a tax credit hinges on a return to the employee of income tax paid. It can take the form of an end of financial year 'bonus', or a reduction in weekly/fortnightly tax rates. This proposal draws on work by Michael Keating and Simon Lambert which suggested reforms to the interface between taxation and social security. Keating and Lambert proposed that the means tests for all of the various types of family assistance should be consolidated into a single test as a way of removing very high effective marginal tax rates (EMTRs). The proposed tax credit would then be incorporated into the family assistance payments and phased down at a common withdrawal rate of 30 per cent.(28) Peter Dawkins and others have also proposed a negative income tax system for dealing with the problem of high effective marginal tax rates. A negative income tax differs from the earned income tax credit in that it does not hinge on tax contributions previously made by the employee/taxpayer(29) The BCA's intervention into the wages and unemployment debate canvassed the negative income tax approach but focussed most fully on the earned tax credit scheme along the lines proposed by Keating and Lambert.(30)

There are basically two goals behind the proposals for reform to the system of income support:

  1. compensation to low income families as an alternative to wage increases for low wage workers, thereby assisting with wage restraint and employment growth

  2. reducing effective marginal tax rates (EMTR) for low paid workers and social security beneficiaries. This aims to eliminate some of the poverty traps which are assumed to act as disincentives in the labour market. It is argued that these poverty traps impede low wage workers taking on more hours of work and dissuade the unemployed taking on jobs.

One of the reasons why compensation is targeted at low income families is that the Five Economists believe that not all low wage workers live in low income households. Using research by Richardson and Harding, they argue that low wage workers live in households that are spread right across the distribution of family incomes. Some commentators, such as Des Moore and Ross Gittins, have taken this research to mean that most low wage workers live in 'well-off' families. Research by Watson and Buchanan, however, shows that this is far from the case. Leaving aside families entirely reliant on social security, Watson and Buchanan argue that low wage workers mostly live in households with other low wage workers and with middle-wage workers. They show that the earnings of low wage workers are not an income supplement for well-off families.(31)

The observation that low wage workers are spread across a family income spectrum leads Peter Dawkins to argue that Living Wage increases are a 'very blunt equity device'.(32) According to this line of thinking wage increases to the lowest paid workers do not make the best contribution to alleviating poverty. Some of the wage increases go to families who do not need the extra income, and for those families who do need it, the pay increase is whittled away by high effective marginal tax rates (EMTR). Commentators have pointed out, for example, that the latest Living Wage increase of $12 for the lowest-paid has delivered only a fraction of that amount into the family budget, due to tax paid on the increase and concomitant reduced income assistance.(33) The effect of EMTRs is further explained below.

By way of contrast, an earned tax credit system would deliver more money to low income families. The Keating proposal-a 2 per cent supplement-would amount to an increase in disposable income of over $10 per week for a low paid worker on two thirds of average male ordinary time earnings. Under the current system, for a worker to get this kind of increase in disposable income, their wage increase would need to be in the order of $70 a week.(34) The main beneficiaries of this proposed scheme are low paid workers in low income families with dependent children. Single persons, and childless couples, earning just above minimum wages, might actually lose under this system.

Australia has a reputation within the OECD for having a a very tightly targeted social security system. Benefits are paid according to very precise means-testing guidelines. Targeting can have desirable consequences in reaching groups in society who might otherwise be overlooked. For example, the Family Allowance Supplement, paid to the low income families of employed workers, is credited with successfully reducing child poverty amongst low income working families. However, a tightly targeted social security system can also create major problems by way of 'poverty traps'. This means that when low income recipients earn additional money, they lose a very large percentage of that extra income by way of tax payments and reduced benefit payments. In some cases, low wage workers receiving some form of social security payment can face effective marginal tax rates (EMTR) of over 85 cents in the dollar if they earn extra income.(35) This kind of penalty is seen as a major disincentive for low wage workers taking on extra work, or for unemployed workers taking up marginal job offers.

It is clear that the compensation issue and the EMTR issue are closely bound together. Keating's proposal deals with both. Not only does the 2 per cent supplement act as compensation, but the phasing out of the tax credit at higher levels of income combined with the tapering of other family benefits means that high EMTRs do not whittle away the benefits of the compensation.

Has it worked elsewhere

What has been the overseas experience of schemes like that proposed by Keating and Lambert? The United States provides a useful example because a similar scheme-the Earned Income Tax Credit (EITC) scheme-has been used in conjunction with low wage jobs to encourage single parents on welfare to enter the workforce. In the early 1990s, two changes took place simultaneously: many states set up welfare-to-work schemes and restricted eligibility for welfare benefits; and the Federal government dramatically expanded the EITC scheme. As a result of these measures large numbers of single parents entered the workforce. Between 1994 and 1997 the welfare caseload fell 22 per cent, about 1.1 million cases. Over the same time period the number of single mothers in work increased by 14 per cent (some 867,000 people).(36)

Before we assess these results it's worth briefly examining how the EITC works. Basically it provides a 40 per cent refundable tax credit to low income families with two or more children up to a maximum credit of US$3556. For a family with one child, the credit was somewhat smaller. The EITC phases out as family incomes rise above US$12,000, tapering away to nothing by the time family income reaches US$29,000. David Ellwood sums it up:

The EITC clearly helps to overcome some of the problems created by the increased inequality of wage earnings occurring in the United States. For the lowest income families, it represents a 40 per cent pay raise.(37)

Researchers who have studied the EITC have mainly focussed on its impact on the labour market; its impact on marriage and its impact on the social safety net. As David Ellwood points out, the system penalises marriage in the long-term because working single mothers face a significant EITC penalty if they marry.(38) While the EITC has not itself undermined the social safety net, it has been tarnished by the draconian climate in which it was expanded. This saw the widespread diffusion of a philosophy which aimed to 'end welfare as we know it' and to severely restrict welfare eligibility, in some cases to a maximum of five years for life. It has meant that those who fall through the social safety net fall a very long way.

It is, however, in the labour market that the EITC has had the most serious impact. On the one hand, it has been successful at encouraging large numbers of single mothers to enter the workforce. It has, moreover, been effective at removing those disincentives which block the movement from welfare to work.(39) On the other hand, it has placed downward pressure on wages at the bottom of the labour market, expanding the size and reach of the low wage sector. Turnover in jobs has been high and the size of the 'working poor' has continued to expand. The EITC has allowed low wage jobs to proliferate because it operates as a public subsidy to low-paying employers.(40) The cost of the scheme has, moreover, become very large. In 1997 it cost the United States government US$28 billion, a figure far greater than the cost of its traditional welfare scheme, Aid to Families with Dependent Children.(41)

The lesson for Australia from the United States experience is that earned tax credits are a successful response to the problem of welfare poverty traps and may warrant further examination. Earned tax credits are not, however, a suitable partner for low wage jobs since, as we have seen from the US experience, they compound many of the problems to be found in the low wage sector. Instead, if earned tax credits can be used as a complement to wage improvements at the bottom of the labour market, then optimum outcomes all round can be expected. In this respect, the Five Economists are mistaken to link their tax credit schemes with an expansion in low paid work. There are too many serious shortcomings attached to low paid work for Australian policy makers to consider expanding the low wage sector here.

The wider context-a discussion

The repercussions of reducing wages

Most of the analysis in the forgoing sections has concentrated on a detailed assessment of the Five Economists' proposals. It is important, however, to address these issues in a wider context. Proposals for cutting wages at the bottom of the labour market belong to a particular strand of economic thinking which believes that deregulated labour markets are good for the economy, and that greater inequality in earnings is both inevitable and desirable.(42) In so far as social costs are involved-poverty, crime, hardship-these can be tackled by government intervention. In terms of the current debate, this government intervention takes the form of income transfers to alleviate family poverty. In summary, advocates of this strand of economic thinking generally believe that Australia should develop a large low wage sector and that the government should restrict itself to ameliorating the social costs which might accompany this.

By way of contrast, I would suggest that the government has a much more decisive part to play in economic management, and particularly in reducing unemployment. In the next section I will offer more details about how this might occur. In this section I will discuss some of the economic consequences which might flow from radically reducing wages at the bottom of the labour market.

There are good reasons for believing that the creation of a large low wage sector in Australia would lead to increased family poverty in Australia.(43) In a recent analysis of low wages and family incomes, Sue Richardson and Ann Harding (from the National Centre for Social and Economic Modelling) modelled the impact of cutting real wages at the bottom of the labour market. They concluded that a reduction in adult wages of 30 per cent (and junior wages of 15 per cent) for low wage workers would leave them about $7 billion dollars worse off. To make this exercise worthwhile, two million new full-time jobs would need to be created, a task which they noted 'requires quite implausibly high elasticity of demand for low wage labour'.(44) In other words, family poverty would increase with cuts of that magnitude because any improvement in unemployment would be insufficient to offset the lost income.

Only increased social security transfers could alleviate that level of family poverty. Yet the social security system would itself come under pressure once a more 'flexible' labour market became entrenched. This would happen because a cut in the order of 30 per cent would have a major impact on the monetary gap between low wage work and unemployment benefits. Unless unemployment and related benefits were also cut by a large amount, significant poverty traps would emerge. Already, in the current climate, commentators have grave concerns about the disincentive effects of the interaction between taxation, social security and low wages. These disincentive effects would be magnified enormously in a climate of lower wages and unchanged levels of social security benefits. Clearly, the pressure to reduce social welfare would be enormous, and this would flow across the board, affecting single parent benefits, sickness benefits and so forth.

The only way to avoid this outcome would be to implement an earned tax credits scheme which left the remainder of the social security benefits system intact. This would allow the gap between unemployment benefits and job earnings (plus tax credits) to remain quite large. As the American experience shows, such a scheme would not be revenue neutral, but would entail large government outlays. The pressure to curtail these outlays over time would be enormous, and restrictions on government expenditure on the social wage would seem to be inevitable. The question needs to be asked: for what gain? In essence, government expenditure on tax credits would constitute a large public subsidy for propping up a low wage sector. Is it really worth having a low wage sector?

We can gain a glimpse into what a large low wage sector would look like by examining the existing characteristics of low wage work. It is reasonable to assume that these characteristics will be reinforced by the emergence of a large low wage sector. The most important characteristics concern poor outcomes in the areas of worker self-esteem, access to training, earnings mobility and career paths, job turnover and productivity.

Irrespective of the final level of family income, the message which very low paid work sends to the worker is that they are not worth much in the eyes of society. If they can only survive by virtue of a government 'top-up', their sense of independence is also compromised. While this notion of self-esteem may be somewhat subjective, there are other, more objective measures which suggest that low paid jobs are also poor quality jobs.

Low paid work is strongly associated with poor access to training. Analysis of Australian Bureau of Statistics training data for those occupations likely to proliferate in an enlarged low wage sector indicate that low paid workers are only about half as likely to get access to training as are better paid workers.(45) Labour mobility figures in Australia show very little upward occupational mobility for low-skilled workers. Research by Burgess and Campbell shows that the large growth in casual jobs during the 1990s did not provide a foothold into secure employment for those casual workers.(46) From this we can infer that earnings mobility at the bottom of the labour market is quite limited in Australia.

Further insights into the nature of low paid jobs are provided by comparing key characteristics of low and high wage firms. Australian Bureau of Statistics survey data shows that low wage firms are far less likely to export their output than comparably-sized firms outside the low wage sector. In addition, they have a much weaker commitment to workplace productivity, evident in their absence of quality control systems.(47)

Probably the most dramatic illustration of what a low wage sector looks like is evident in the United States labour market. As the economy with the largest low wage sector in the OECD, and the country which prides itself on being the land of opportunity, the United States nevertheless performs badly when it comes to earnings mobility. Research by the OECD found that seven out of ten American low-paid workers in 1986 were either still in low paid jobs or were not working full-time five years later. The comparable figure for Denmark was just one-third.(48)

But while there is not much mobility out of low paid work in the United States, there is significant turnover between different forms of low paid employment. Researchers from the Economic Policy Institute in Washington, for example, found a considerable increase in job instability during the 1980s, and large declines in earnings were associated with the most unstable jobs.(49) Research by the Urban Institute on the employment situation in the health care, child care and hospitality industries found dramatic differences in industry mobility between low-paid workers and non-low-paid workers. Whereas about 68 per cent of non-low-paid workers were still in the same industry after 32 months, only 14 per cent of low-paid workers were. A larger proportion of low-paid workers-over 18 per cent-had actually passed through three industries during that time period, all within the low paid sector.(50)

With minimal access to training, high job turnover and negligible prospects for career advancement, it is not surprising that low wage sectors have very low productivity. Again the United States is instructive. The emergence of a large pool of low-wage jobs in the service sector has had a serious and adverse impact on productivity growth. The growth in manufacturing productivity between 1979 and 1990 was 2.9 percent, and between 1990 and 1996 it was 4.2 percent. By comparison, in non-manufacturing it was just 0.3 percent and 0.2 percent respectively. These latter growth rates were a tenth of those prevailing in German non-manufacturing over the same period.(51) As the economic historian Robert Brenner concluded:

The upshot has been a truly vicious circle, in which low wages have made for low labour productivity growth which has in turn rendered 'unrealistic' any significant growth of wages and thereby provided the basis for continued low productivity growth.(52)

An alternative approach for dealing with unemployment

The discussion in the preceding section makes it clear that expanding the low wage sector in Australia runs directly counter to economic strategies which emphasise the 'high road' to the future, the importance of building the 'clever country', and a preference for high productivity workplaces. These approaches emphasise investment in training, and in research and development as the basis for economic progress. Whilst these strategies do not directly benefit the unemployed, particularly the long-term unemployed, they do generate wealth for the country and promise a more secure future. In turn, such wealth and security make possible direct labour market interventions which target assistance to the long-term unemployed. These need to consist of training programs, job subsidies and programs of direct public sector job creation.

Government intervention plays a central role in this scheme of things. Government support for training, and for research and development, is generally favoured by both sides of politics. Direct labour market intervention is more likely under a Labor Government than a Coalition Government, though this tends to be limited in its scope. For example, Working Nation, the centrepiece of the Labor Government's last term, offered little in the way of direct public sector job creation and concentrated on training and jobs subsidies.

Why are training and job subsidies insufficient? Why is direct public sector job creation also necessary? These questions are often posed because of the prevailing ideological climate which 'regards virtually any government spending as socially wasteful'.(53) Public sector job creation is caricatured as 'painting rocks' (or even worse, 'painting barb wire').(54) The notion that the public and private sectors are complementary to the health of the economy-expressed in that classic phrase, the 'mixed economy'-has been missing from public debate since the early 1980s. To some extent, this reflects the demise in the influence of Keynesian economics during the 1970s, a demise which left governments at the mercy of economists who emphasised budgetary restraint and interest rate juggling as the primary purpose of government. This loss of faith in the mixed economy also reflects the influence of vested interests who have much more to gain from emphasising private over public economic activity.

Fortunately, in recent years, some thoughtful economists have sought to re-establish the appropriate balance between public and private economic activity. These include John Quiggin, John Langmore, Bill Mitchell, Martin Watts and Peter Kenyon.(55) Their approach to unemployment flows from a different analysis of the causes of protracted unemployment. Instead of emphasising wage costs as the underlying problem, they emphasise the problems which demand deficiency, compounded by budgetary restraint, create for the economy. John Quiggin, for example, points to the decline in Commonwealth allocations to the states. Between the mid 1970s and the early 1990s, Commonwealth funding in this area fell from 91/2 per cent of gross domestic product (GDP) to under 5 per cent.(56) The results of this at a state level have been evident for over a decade: reduced resources for education, health, public housing and public transport. The impact on employment levels in certain industries has been severe. As Peter Kenyon observes, when public functions have been transferred to the private sector, jobs have not just been transferred, but have been lost:

policies of 're-engineering', 'down-sizing' and 'right-sizing' in the public sector have meant that there are fewer jobs available in toto for those with the characteristics of the people who used to fill these jobs.(57)

Older, blue collar workers, often from a migrant background, have been particularly hard-hit by these kinds of policies.(58)

The proposals put forward by these economists all emphasise a more interventionist role for governments in reducing unemployment. Bill Mitchell proposes a Buffer Stock Employment model whereby the government would act as an employer of last resort, absorbing workers who were displaced from the private sector. Such employment could expand and contract according to the economic cycle.(59) John Quiggin and John Langmore call for increased government spending on community services and infrastructure as a way of simultaneously reducing unemployment and contributing to the 'vitality of the economy', thereby guaranteeing its long term expansion. Peter Kenyon puts forward proposals for greatly expanded public spending on infrastructure. He emphasises the 'growth-enhancing effects for the private sector due to the positive externalities created for the private sector by public infrastructure and public education'.(60) His proposals do not necessarily imply that the government need be the employer in all cases. He sees merit in tendering such jobs to private contractors. This avoids the stigmatising of the unemployed, something which often happens with direct public sector job creation schemes. As he sees it, 'the primary objective is to increase the level of aggregate demand and therefore to increase overall employment opportunities'.(61)

Whatever particular employment strategy is adopted, labour market programs which target the long-term unemployed must play a central role. As noted earlier, the labour market of the 1990s is quite distinctive in its utilisation of labour. There would be little point in increasing aggregate labour demand if all that eventuated was increased hours of work for the existing workforce. It is essential that any increase in the demand for labour should lead to new jobs for those currently unemployed. Traditionally, public sector job creation has been viewed in terms of economic cycles, as the appropriate government response during a recession. Yet in the current climate of strong economic growth Australia still has unacceptably high levels of unemployment. The challenge for policy makers in coming years is to find the right mix of government funding, labour market programs and private sector initiatives which deal with this phenomenon of 'jobless economic growth'.

All of the proposals discussed above deserve to be taken seriously. Unfortunately, the ideological climate which has long marginalised these kinds of strategies still prevails. The most common response to these proposals, even from sympathetic economists, is often: 'who is going to pay'? The answer is not at all elusive. Those in jobs are obliged to pay for the creation of new jobs for the unemployed and the taxation system exists for the very purpose of transferring revenue towards socially desirable goals. The current budget surplus-and the projected surpluses over the coming years-suggest that the taxation increases need only be modest. The current proposals for dealing with unemployment by reducing wages also entail asking the question: who pays? In this case, those least able to afford the cost-award workers on low pay-are being asked to shoulder the burden.


  1. The letter to the Prime Minister was sent to the Australian, October 26, 1998, p. 13. Their individual contributions include: P. Dawkins and J. Freebairn, 'Towards Full Employment', The Australian Economic Review, vol. 30, no. 4, December 1997, pp. 405-17; P. Dawkins, 'Solutions to Unemployment and Avoiding the 'Diabolical Trade-off': A Discussion', in G. Debelle and J. Borland, Unemployment and the Australian Labour Market, Proceedings of a Conference held by RBA and CEPR, ANU, June 1998; P. Dawkins, 'Rebuilding the Safety Net: Reducing Unemployment', Paper to BCA Conference on New Directions: Rebuilding the Safety Net: A Wage/Tax Trade-Off for Australia, April 1999; C. Richardson, 'Unemployment or Tax Reform?', in Debelle and Borland, op. cit.; M. Keating, and S. Lambert, 'Improving Incentives: Changing the Interface of Tax and Social Security', The Australian Economic Review, vol. 31, no. 3, December 1997, pp. 281-9.

  2. As we saw in the 1980s, this 'factor shares' argument had serious shortcomings. A large proportion of the output which was transferred from labour to capital was squandered in speculative investment (such as inflated real estate assets) rather than being channelled into productive investment. Green and Wilson have seen this as a fundamental weakness in incomes policy at a macro level: 'there was no effective mechanism to channel increased profits into investment'. R. Green, and A. Wilson, 'Employment Policy: The Strategic Dilemma of the ALP', Paper for ALP Caucus Economic Policy and Living Standards Committee, May 4 1999, p. 12.

  3. Australian, 26 October 1998, p. 13.

  4. Business Council of Australia (BCA) 1999, Rebuilding the Safety Net, New Directions Discussion Paper 1, March 1999, p. 6.

  5. He proposes turning the AIRC into a 'voluntary advisory/mediation service provider', D. Moore, 'IR Legislation-Further Deregulation?' Address to Australian Industry Group Conference, 5 March 1999.

  6. D. Moore, 'Why Regulation of the Labour Market is Inequitable, Outdated and Inefficient', Occasional Address to Annual Meeting of the HR Nicholls Society, 30 November 1998, p. 8.

  7. G. Debelle, and J. Vickery, 'The Macroeconomics of Australian Unemployment', in Debelle and Borland, op. cit.

  8. Dawkins, in Debelle and Borland, op. cit. and Dawkins and Freebairn, op. cit.

  9. BCA, op. cit, p. 25.

  10. J. Isaac, Paper to Business Council of Australia (BCA) Rebuilding the Safety Net, New Directions Conference, Melbourne, 29 April 1999, p. 5.

  11. See ACIRRT, Australia at Work: Just Managing? Prentice Hall, Sydney, 1999, and A. Moorehead, M. Steele, M. Alexander, K. Stephen, and L. Duffin, Changes at work: the 1995 Australian Workplace Industrial Relations Survey, Longman, South Melbourne, 1997.

  12. See ACIRRT, Australia At Work, op. cit., ch. 6.

  13. Dawkins and Freebairn, op. cit., p. 409. The BCA expects about 40 per cent of new jobs to go to new labour market entrants, BCA, op. cit. p. 35.

  14. For information on the compounding problems of long-term unemployment see P. N. Junankar and C. A. Kapuscinksi, The Incidence of Long Term Unemployment in Australia, Discussion Paper No. 249, Centre for Economic Policy Research, ANU, Canberra, 1991 and T. O'Louglin, and I. Watson, Loyalty is a one way street: NESB Immigrants and Long-Term Unemployment, ACIRRT Monograph, Sydney 1997.

  15. R. G. Gregory, 'What Do We Mean by an Unemployment Solution?', in Debelle and Borland, op. cit., p. 335.

  16. This was one of the main explanations offered by Card and Krueger for their controversial findings that raising the minimum wage caused increases in employment levels, D. Card and A. B. Krueger, Myth and Measurement: The New Economics of the Minimum Wage, Princeton University Press, Princeton, 1995.

  17. ibid., p. 7.

  18. J. Dolado, F. Kramarz, S. Machin, A. Manning, D. Margolis, and C. Teulings, 'The Economic Impact of Minimum Wages in Europe', Economic Policy, October 1996, pp. 327-28.

  19. C. Brown, C. Gilroy, and A. Kohen, 'The effect of the minimum wage on employment and unemployment', Journal of Economic Literature, June 1982, pp. 487-528.

  20. Card and Krueger, op. cit.; D. Card, and A. B. Krueger, 'A reanalysis of the effect of the New Jersey Minimum Wage Increase on the fast-food industry with representative payroll data', NBER Working Paper 6386, January 1998. []; D. Neumark, and W. Wascher, 'Do Minimum Wages Fight Poverty?' NBER Working Paper 6127, August 1997, [].

  21. Dolado et al., op. cit., p. 357.

  22. J. M. Abowd, F. Kramarz, D. N. Margolis 'Minimum Wages and Employment in France and the United States', NBER Working Paper 6996, 1999, [].

  23. OECD, Employment Outlook, Paris, June 1998, p. 31.

  24. D. S. Hamermesh, 'The evidence on the demand for labour', Paper to BCA Conference on New Directions: Rebuilding the Safety Net: A Wage/Tax Trade-Off for Australia, April 1999.

  25. In its own calculations, the BCA assumes wage cuts of the order of 20 per cent or more would be necessary for the most disadvantaged jobs seekers to find employment. They partially base this figure on the level of wage subsidies which DEETYA formerly paid to employers to take on long-term unemployed. BCA, op. cit., p. 35.

  26. Joe Isaac cites research by John Martin which provides a 10 per cent estimate; Stretton and Chapman cite the general consensus in the literature for an estimate of between 15 to 20 per cent. J. Isaac, op. cit., p. 2.; A. Stretton, and B. J. Chapman, An Analysis of Australian Labour Market Programs, Discussion Paper No. 247, Centre for Economic Policy Research, ANU, Canberra, 1990, p. 36.

  27. Dawkins, in Debelle and Borland, op. cit., p. 318.

  28. Keating and Lambert, op. cit., p. 288.

  29. P. Dawkins, G. Beer, A. Harding, D. Johnson, and R. Scutella, 'Towards a Negative Income Tax System for Australia', The Australian Economic Review, vol. 31, no. 3, September 1998, pp. 237-57. Under a negative income tax system low income individuals or families receive payments from the tax system. In that sense, they pay negative tax. In its more radical version it involves a 'flat tax' system with a flat tax rate on private income. One of the main drawbacks with the Dawkins scheme is its high cost. It requires very large budget outlays if it is not to be regressive for low to middle income earners (BCA, op. cit, p. 30).

  30. BCA, op. cit. pp. 34-36.

  31. Watson, and J. Buchanan, 'Beyond Impoverished Visions of the Labour Market', chapter for Academy of the Social Sciences in Australia, Creating Unequal Futures, Canberra, (forthcoming).

  32. Dawkins, op. cit., p. 2.

  33. For example, Herald columnist Ross Gittins. R. Gittins, 'Living Wage' is a poor excuse for a safety net', Sydney Morning Herald, May 1, 1999, p. 58.

  34. Dawkins, op. cit., p. 2.

  35. Keating and Lambert, op. cit., p. 281.

  36. G. Burtless, 'Can the Labor Market Absorb Three Million Welfare Recipients?', Paper prepared for 'The Low Wage Labor Market: Challenges and Opportunities for Economic Self-Sufficiency' research project, 1998, p. 2.

  37. D. Ellwood, 'From Welfare to Work', CEDA Bulletin.

  38. D. Ellwood, 'The Impact of the Earned Income Tax Credit and other Social Policy Changes on Work and Marriage in the United States', Paper to Conference on Income Support, Labour Markets and Behaviour: A Research Agenda, ANU, Canberra, November 1998, p. 26. The likely impact of an EITC on women's labour supply in Australia has been analysed by Patricia Apps. She concludes: 'we would ... expect a significant fall in the participation and hours of work of married women as second earners, but relatively little change in the hours of primary income earners facing lower EMTRs'. 'Reforming the Australian Tax-Transfer System', Paper for Conference on 'Tax and Transfer Reform in Australia and Germany', University of Potsdam 27-29 January 1999, p. 16.

  39. 'The EITC gets money to needy families without offering incentives to curtail work effort'. M. Hout, 'Inequality at the Margins: The Effects of Welfare, the Minimum Wage, and Tax Credits on Low-Wage Labor', Politics and Society, vol. 25, no. 4, December 1997, p. 520.

  40. G. Burtless, 'Rising Wage Disparity and U.S. Poverty: Recent Developments and Policy Responses', Paper for conference on 'Fathers and Sons: The New Forms of Unemployment',
    8-9 May, 1998, Stresa, Italy, p. 31.

  41. T. Eardley, Working But Poor? Low Pay and Poverty in Australia, Social Policy Research Centre Discussion Paper, 1998, p. 33.

  42. The OECD Employment Strategy, Paris, 1994, epitomises this position.

  43. By 'low wage sector' I mean large numbers of jobs which do not pay a living wage. In terms of relative poverty, the low paid workforce is often defined as those workers earnings below two-thirds of median earnings. The failure of such jobs to pay a living wage means that such jobs are only available as a 'supplementary income', and that workers must find their main source of income elsewhere (for example, from another household member or from the government).

  44. S. Richardson, and A. Harding, 'Poor Workers? The link between low wages, low family income and the tax and transfer systems', Paper to the Conference on Fairly Efficient? Equity and Productivity in the Australian Labour Market, ANU, Canberra, June 1998, p. 30.

  45. Watson and Buchanan, op. cit.

  46. J. Burgess, and I. Campbell, 'Casual Employment in Australia: Growth, Characteristics, A Bridge or a Trap?', The Economic and Labour Relations Review, June, vol. 9, no. 1, 1998, pp. 31-54.

  47. Data reproduced in ACTU, Living Wage 1997/98: ACTU Written Submission, 29 January 1998.

  48. OECD, Labour Market Policies: New Challenges-Policies for Low-Paid Workers and Unskilled Jobs Seekers, Paris, 1997, p. 10.

  49. L. Mishel, J. Bernstein, and J. Schmitt, The State of Working America 1998-99, Economic Policy Institute, Cornell University Press, Ithaca, 1999, pp. 228-29.

  50. N. Pindus, D. Dyer, C. Ratcliffe, J. Trutko, and K. Isbell, Industry and Cross-Industry Worker Mobility: Experiences, Trends, and Opportunities for Low-Wage Workers in Health Care, Hospital, and Child Care, The Urban Institute,1997, table 7A.

  51. R. Brenner, 'The Economics of Global Turbulence: A Special Report on the World Economy, 1950-98', New Left Review, no. 229, whole issue, p. 241.

  52. ibid., pp. 206-07.

  53. P. Kenyon, 'Infrastructure Spending and Unemployment: Government Responsibility for Growth and Jobs', The Australian Economic Review, vol. 30, no. 4, 1997, p. 429.

  54. Judith Sloan's disparaging description. 'Some policy responses to long-term unemployment', The Australian Economic Review, no. 2, April June 1993.

  55. Kenyon, op. cit.; J. Langmore and J. Quiggin, Work for All: Full Employment in the Nineties, Melbourne University Press, Melbourne, 1994; W. F. Mitchell, 'Full employment abandoned-the macroeconomic story', in M. Carman, and I. Rogers, eds, Out of the Rut: Making Labor a Genuine Alternative, Allen and Unwin, St Leonards, 1999; W. F. Mitchell, and M. J. Watts, 'The Path to Full Employment', The Australian Economic Review, vol. 30, no. 4, 1997, pp. 436-44.

  56. Langmore and Quiggin, op. cit., p. 3.

  57. Kenyon, op. cit., p. 422.

  58. See O'Louglin and Watson, op. cit.

  59. Mitchell and Watts, op. cit., p. 441-42.

  60. Kenyon, op. cit., p. 427.

  61. Kenyon, op. cit., p. 428.



Five economists have proposed a plan to cut the jobless rate. In this letter to John Howard they argue the case.

DEAR Prime Minister,

This letter offers some ideas for policies that could complement your tax reform package, to help in substantially reducing unemployment over the next few years. We believe that the continued steadiness of fiscal and monetary policy of recent times and the continuation of microeconomic reform, of which tax reform forms part, are both important elements in a set of policies to reduce unemployment. Our specific additional proposal is to link industrial relations reform with reform of the tax and welfare system. This would complement your present policies and significantly enhance the chances of achieving large reductions in unemployment.

The growth in jobs necessary to substantially reduce unemployment will require higher economic growth and lower growth in wage costs. Tax reform will lead to a more robust economy and improve the chances of realising higher economic growth. Evidence suggests, however, that enhanced economic growth alone will not be sufficient to reduce unemployment much below 7 per cent in the long run. Economic growth will need to be complemented by policies directed at improving the relationship between growth and unemployment, involving a linking of your industrial relations and tax and welfare reforms. The linkage would include an attempt to achieve a lower growth in real labour costs, alongside an integrated approach to reforming the social security system and labour market programs. The proposed reforms to family payments in your tax package are a solid start in the area of social security reform.

Recent research at the Reserve Bank of Australia, by Debelle and Vickery, suggests that a 2 per cent fall in the growth of real wages, for one year, could lead to a permanent reduction of unemployment of about one percentage point. Lower growth in wage costs leads directly to a higher demand for labour. It also results in lower inflationary pressure and enables the Reserve Bank to conduct a more expansionary monetary policy.

We suggest that the Commonwealth Government should mount an argument in the Industrial Relations Commission for the replacement of Living Wage adjustments, at least for the time being, with compensation for those on low wages in low income families by providing tax credits to boost their incomes. This would be both more equitable and better for labour market efficiency than the established approach of relying on award wage interventions to secure equity. Unlike the Accord, there would be no expectation that unions or employees would be restrained from negotiating over-award increases.

While the number of employees who depend on award variations for wage increases has decreased substantially, it is still around 30 to 40 per cent, and many enterprise agreements are also affected by awards. If the "award wages safety net" were frozen for four years, this would reduce the growth in real award wages by about 10 to 12 per cent, due to inflation. Allowing for the coverage of awards and for the possibility of over-award increases, this should reduce the average growth in real wages by about 3 to 4 per cent. Based on Debelle and Vickery's estimates, this should reduce the level of unemployment by 1.5 to 2 percentage points below the level that would otherwise be attainable, to about 5 to 5.5 per cent. Earnings would be suppressed more for low-skilled than for high-skilled workers. It is likely that lower growth in wages for the low skilled will be more helpful than lower growth in wages of the high skilled, as the unemployed are disproportionately low skilled.

Why would this approach also be better for equity? First there is the reduction of unemployment. Second, the use of the tax-transfer system to improve the income for low-income families which include wage earners, would be better than Living Wage adjustments. Recent research by Sue Richardson and Ann Harding has highlighted how those people receiving low wage rates are well spread throughout the distribution of family incomes. This makes Living Wage adjustments a very blunt equity device.

Indeed, the proposed tax credits would be more valuable to many such low-income families than "Living Wage adjustments". For many low-wage earners in low-income families, especially those with children, an increase in their wage has little effect on their net income after the effect on tax and social security entitlements. This proposal would improve the position of low-income families on low wages, in the spirit of the famous Harvester decision. This should prove to be a powerful argument in the Industrial Relations Commission.

Recent research at the OECD shows that no single policy can be expected to solve the unemployment problem. A range of policies will be required, acting on both the demand for and the supply of labour. As well as tax reform and the award rates freeze, this should include the upgrading of the education and training systems and a systematic approach to labour market programs. Instruments such as re-employment bonuses as well as the tax credits would increase the incentive for people to move from welfare to work. The tax credits should form part of a rationalisation of our complicated tax and social security system and a move towards a negative income tax system. Michael Keating has recently shown how much tax credits and your proposed rationalisation of family payments could be consolidated to achieve even lower effective marginal tax rates.

The Government will need to commit funds to the "welfare to work" policies, including tax credits and labour market programs. We believe, however, that the costs would be quite modest, especially if compared with the medium to long-run benefits. In the long run, the result of reaching around 5 to 5.5 per cent unemployment would mean substantial savings in unemployment benefits. There would be an added tax revenue dividend from these policies helping to maintain solid ongoing economic growth, and very importantly an equity dividend from the lower unemployment and enhanced assistance to low-income families. We hope that you find these suggestions helpful and would be very pleased to meet with you or any members of your Government to discuss proposals in more detail.

Yours sincerely

PETER DAWKINS Professor and Director, Melbourne Institute of Applied Economic and Social Research, The University of Melbourne

JOHN FREEBAIRN Professor of Economics, Department of Economics, The University of Melbourne

ROSS GARNAUT Professor of Economics, Australian National University

MICHAEL KEATING Visiting Fellow, Economics Program, Research School of Social Sciences, Australian National University

CHRIS RICHARDSON Director, Access Economics.

(c) Nationwide News Proprietary Ltd, 1998. THE AUSTRALIAN 26/10/1998 P13