The End of Industrial Relations Reform?
Andrew Norton
Click here for PDF version

WorkChoices polling revealed the political dangers of industrial relations reform, explains Andrew Norton

From the earliest polls on WorkChoices, the Howard government’s industrial relations reforms were in political trouble. Opponents outnumbered supporters by nearly three to one, a persistent ratio through successive surveys. Along with the negative direct issue polling, there was also the problem that voters claimed WorkChoices made them less likely to cast a Coalition vote.(1) If the Liberals conclude that WorkChoices contributed to their November 2007 election defeat, they will take it off the political agenda for a long time to come.

But why was WorkChoices so unpopular in the first place? The government’s previous industrial relations reforms, despite being controversial in the mid-1990s, achieved acceptance.(2) Extensive polling on various aspects of WorkChoices lets us identify public concerns about exposing vulnerable workers to the market, and see how the public responded to the government’s argument that WorkChoices would be good for the jobs and the economy.

Overall attitudes
Legislatively, WorkChoices became possible when the government acquired its Senate majority on 1 July 2005. Politically, WorkChoices was dead on arrival. That month, surveys by ACNielsen and Roy Morgan Research confirmed what an ACTU-commissioned poll had found in May: that most voters with an opinion opposed the reforms.(3) Though Morgan found fewer people against them than ACNielsen, 47% compared to 60%, that was because more of its respondents declined to offer a view. An earlier Morgan inquiry as to whether voters had heard of the reforms prompted reticence on the for/against question. Of those expressing an opinion, the proportion against was near identical for both pollsters, 73.4% and 74.1%.(4)

What is striking about these results is how little they changed. The unions and the government spent millions of dollars on advertising, but to no effect, or at least no net effect. In three Morgan polls between July 2005 and April 2006, a period covering parliamentary debate on the legislation and its coming into effect on 27 March 2006, attitudes shifted by only 0.8%, excluding those without an opinion. In the ACNielsen series of polls, published in the Sydney Morning Herald and The Age, recorded in figure 1, opposition ranged between 57% in July 2005 and 60% in March 2007, a year into the new system. Support went from a low of 21% to a high of 24%. Restricting the sample to those with a view, opposition varied between 69% and 74%.

Such stable opinion over time is rare, and especially unusual for intensely-debated issues. Whether out of self-interest or principle, it suggests that voters already had views on the underlying issues raised by WorkChoices. If so, Labor and the ACTU won the WorkChoices debate by successfully linking their stance to already existing public beliefs.

Figure 1: Support for WorkChoices

Question: Do you support or oppose the government’s changes
to industrial relations?
Source: ACNielsen/The Age

Self-interest?
Workplace law affects the material self-interest of many Australians. Nearly two-thirds of Australians aged over fifteen are in the workforce. Many others depend financially on employed family members, further extending the workplace’s reach. Despite this, the polling strongly suggests that self-interest alone cannot explain why people opposed WorkChoices. Three pollsters have asked their respondents about the personal effects of workplace relations reform.(5) None find that the proportion of people perceiving negative effects matches, or comes anywhere near matching, the proportion who oppose WorkChoices. They are against it even though they do not expect it to cause them any personal detriment. Nor can self-interest explain support for WorkChoices. All three pollsters find fewer people saying that they will be better off under WorkChoices than they find in favour of the reform.

For the personal impact questions, unlike the more general support/oppose questions, opinion shifts over time. Two of the three pollsters find a declining proportion of people thinking that they would be worse off as a result of WorkChoices. Figure 2 reports the results of a personal impact question asked as part of consumer research carried out by Sensis. Because it is not primarily aimed at collecting political opinion, it interviews a wider range of people than ACNielsen and Newspoll, the other two pollsters with similar questions. Sensis surveys people aged fourteen and above, and not just those who can vote. Apart from an upward move in February 2007, its surveys record a downward trend in ‘negative impact’ answers. The most recent survey, in August 2007, found 29% of respondents perceiving a negative WorkChoices impact, compared to 45% when the question was first asked, in August 2005. The last two Sensis surveys occurred after the government announced a fairness test, which may have influenced the results. Apart from a drop in August 2006, the percentage of people believing that WorkChoices will have a positive impact is stable, at between 10% and 12%. In all surveys except the first, in August 2005, ‘no real impact’ is the most popular answer.

Figure 2: Personal impact of WorkChoices (Sensis)

Question: Do you believe the government’s changes will have a positive impact on you, a negative impact, or no real impact either way?
Source: Sensis Consumer Report

The ACNielsen polls, as seen in figure 3, find the same trend as Sensis. In the period from October 2005 to March 2007—from the debate surrounding the WorkChoices legislation to a year after it came into effect, but before the new fairness test—the proportion of people thinking they would be worse off dropped from 31% to 21%. As in the Sensis survey, the proportion of respondents believing that they would be better off shows no clear trend. Because ACNielsen also asked a general support/oppose question, we can compare those answers with replies to the personal effects questions. In the most recent survey, 2.8 times as many people said that they opposed WorkChoices as said that they would be worse off, and 4.8 times as many people supported it as said they would be better off. For reasons that are not clear, however, a much larger proportion of respondents tell ACNielsen that WorkChoices will make no difference than tell Sensis it will have no real impact.

Figure 3: Personal effects of WorkChoices (ACNielsen)

Question: From what you have heard about the industrial relations reforms,
do you think you personally will be…
Source: ACNielsen/The Age

Newspoll finds more of its respondents than the other two pollsters perceive some personal effect of WorkChoices (figure 4). Both ‘better off’ and ‘worse off’ responses are higher, particularly compared to the ACNielsen survey. There are nearly three times as many Newspoll respondents as ACNielsen respondents believing that they are or will be better off, and half as many again reporting to Newspoll that they are worse off as say the same to ACNielsen. Newspoll only asks this question of people in work, and the answers may reflect more detailed knowledge of WorkChoices. Question wording may also explain discrepancies between pollsters. Newpoll’s question encourages respondents to choose ‘better’ or ‘worse’ in their reply, rather than ‘no difference.’ Despite this, as with Sensis and ACNielsen, Newspoll finds that more people detect no personal WorkChoices impact than perceive either positive or negative effects.

Though finding the same overall pattern of opinion as Sensis and ACNielsen, Newspoll finds no downward trend in people perceiving negative WorkChoices effects. Looking at results from the other two pollsters, it appears as if initially concerned respondents became less worried as their understanding and experience grew. The stability evident in Newspoll’s numbers casts some doubt on that conclusion, at least insofar as it applies to people who are employees. Perhaps the shift showing in the Sensis and ACNielsen surveys reflects answers from people without jobs. It is not clear, however, why they would not typically have replied ‘neither’ or ‘no difference’ all along.

Figure 4: Personal effects of WorkChoices (Newspoll)

Question: How you think the changes to industrial relations will affect you personally? Do you think you will be better, or worse, off?
Source: Newspoll/The Australian

Concern about others
That general opposition to WorkChoices significantly exceeds perceived personal disadvantage suggests, as was said during the workplace relations debate, that some people are concerned about the reforms’ effect on others rather than on themselves. Indeed, I drew this conclusion myself, in an analysis of pre-WorkChoices industrial relations public opinion.(6) Australians believe that an imbalance of power exists in employment relationships, at least for some workers, and that labour market institutions should protect those workers’ interests.

The Australian Survey of Social Attitudes (AuSSA) has twice, in 2003 and 2005, asked its respondents to agree or disagree with the proposition that ‘employees will never protect their working conditions and wages without strong unions.’ The answers were similar in each survey, with the 2005 AuSSA finding 46% agreement and 28% disagreement. Both AuSSA surveys asked whether ‘individual contracts favour the employer over the employee,’ and around half of respondents in each year agreed. The 2005 AuSSA also asked whether ‘trade unions are very important for the job security of employees.’ Slightly more than half of its respondents, 53%, agreed. In all cases, a plurality of people surveyed assumed that employers were stronger than employees, and in two of the three questions pluralities saw unions as having a role in correcting the power imbalance.

Other AuSSA questions produce less-easily-interpreted responses. In 2005, majorities agreed both that ‘employers and employees should be able to negotiate pay and conditions directly’ (57%) and that ‘award wages are the best way of paying workers and setting conditions’ (69%). AuSSA is not the first survey to find results like this.(7) Though confusion may be a factor, the initial propositions require choices that are not logically necessary. Affirmative responses to apparently mutually exclusive propositions make sense if the public believes wage-setting methods should vary with the situation of workers.(8)

In this view of how the public sees workplace relations, employees capable of negotiating individually should be allowed to do so. Managers and professionals, who are more likely to be employed on individual contracts than anyone else, are the most highly-paid people in the workforce.(9) They can look after themselves. By contrast, low-income workers on awards need unions and/or tribunals to protect their interests through setting award wages and conditions. Consistent with this, when a pollster asked which system its respondents would personally rather be under, it found a wide range of responses (table 1).

The problem with WorkChoices, on this reading of public opinion, is not that all or even most workers will be worse off. A majority tell pollsters that WorkChoices makes no difference or leaves them better off, reflecting their strong labour market position. The difficulty, according to public opinion, is that a subsection of the workforce, which is already relatively poorly paid, is being pushed into employment arrangements that could further disadvantage them at work. On an intuitive level, it seems unfair.

Table 1: Employee wage-setting preferences 2001–2004

Question: Under which of these do you think you would be better off?
Source: Saulwick AgePoll

Fairness
Given the prominent fairness theme in anti-WorkChoices advertising, it is a pity the pollsters did not explore opinions about the reforms’ fairness more. The only direct fairness question investigated an otherwise largely overlooked aspect of the workplace relations debate, effects on employers. In May 2007, Newspoll asked whether the Coalition’s new WorkChoices fairness test, which prevented employers from lessening the overall package of workers earning under $75,000 a year, and Labor’s alternative industrial relations policy, were fair on employers. 54% thought the Coalition’s policy was fair; while 41% described Labor’s policy that way. With 25% of Labor supporters improbably declaring the Coalition changes unfair on employers, partisan loyalties appear to have influenced these results.

Inevitably, pollsters’ inquiries on unfair dismissal laws require at least partly leading questions. Framing the issue as fair-on-the employee versus unfair-on-the-employee, without raising other considerations, produces predictably low levels of support for change. When the 2005 AuSSA asked if there should be a law protecting all Australian workers from unfair dismissal, only 11% disagreed. Two Morgan polls, in July and October 2005, specified that WorkChoices only exempted employers with fewer than a hundred staff. They found about a quarter agreeing with the then proposed change. The ACTU poll in May also found about a quarter in favour. It seems support for unfair dismissal law reform doubles when it is restricted to smaller employers. Business lobby groups and the Howard government argued that unfair dismissal laws harm small business, and perhaps this message persuaded some voters.

A barely challenged strength of the Labor/ACTU anti-WorkChoices campaign was the assumption that the status quo—whatever it happened to be—was the base test of fairness. For example, a study of post-WorkChoices agreements in the retail and hospitality industries found that 76% removed Saturday penalty rates.(10) Saturday penalty rates in these industries reflect an era when most employees worked Monday to Friday, little was open on weekends, and a wife at home did the shopping during the week. Those conditions no longer prevail. Many families can only easily shop outside normal business hours, and large numbers of would-be employees such as students can only conveniently work at weekends. Saturday penalty rates look like anachronisms, not benchmarks
for ‘fairness.’

In taking up the terminology of a ‘fairness’ test, the government gave away the debate about whether the status quo constitutes the base for ‘fairness,’ and implicitly conceded that its previous policy had been unfair. Perhaps, given very poor polling for the government in general and WorkChoices in particular, they felt they had no choice. Arguably, it played some role in the declining proportion of people the Sensis survey found perceiving a negative personal impact. Yet it further complicated the argument for WorkChoices by undermining claims that it could increase employment and productivity.

Jobs and the economy
With the detail of WorkChoices clearly unpopular, its political success depended on the government convincing voters that it had other benefits, for the economy overall and for jobs in particular. The phenomenon of people putting aside their own interests for the benefits of others can work in favour of reforms as well as against them. Unfortunately for the government, here, too, it struck major obstacles.

The issue cycle
Arguments based on the economy and jobs are most likely to strike a chord in public opinion when these are the electorate’s top issues. That was the case when most economic reforms occurred. Nearly every year between 1982 and 2004, Roy Morgan Research asked its respondents to nominate the three most important issues the federal government should be doing something about. Unlike other important issues polls, Morgan does not prompt its respondents by reading out possible answers. Instead, it codes into categories what people spontaneously say. In the 1980s and 1990s, economic problems converted into frequent mentions of economic issues. As figure 5 shows, through much of the 1980s and early 1990s more than half of all issues mentioned were economic. The public wanted political action on economic issues, and the political elites responded to this demand. Broad bipartisan consensus on the reform agenda helped secure the electorate’s acquiescence in specific policies they did not, according to polling at the time, support.

Figure 5: Most important issues 1982–2004

Question: Thinking about Australia as a whole, what are the three most important things the federal government should be doing something about?
Economic cluster, 1982–1992:
Unemployment, business/industry/rural growth, stabilise/improve economy, lower tax, overseas trade, decrease deficit, reduce imports from overseas, reduce size and cost of government, reduce cost of living, interest rates
Economic cluster, 1994–2004:
unemployment, taxation/lower taxes, tax reform, taxation/no GST, other taxation, economy and finance, industry and business, petrol prices, interest rates
Source: Roy Morgan Research

By the time of the last survey, in May 2004, economic issues barely rated a mention. Social services such as health and education had taken over as the most important issues. The decline of economic concerns is not surprising. The unemployment rate had been trending down for a decade; in that month it was less than half of its early 1990s recession peak. Only 9% of Morgan’s respondents gave unemployment as one of their three most important issues, down from 74% in 1992. Statistics on real wages and consumer confidence told the same story: most Australians were no longer experiencing the economic problems of the 1980s and early 1990s. Only 6% of Morgan’s respondents gave answers coded as ‘economy and finance’: a fifth of the 1995 figure. The mid-2000s were not a good time to introduce significant economic reforms. WorkChoices was designed to solve difficulties that most voters either no longer saw as problems or did not see as important problems.

WorkChoices and jobs
Compounding the government’s difficult situation, most voters did not accept its argument that WorkChoices would help create jobs. While more people thought that WorkChoices would be good for creating jobs than supported the package overall, they were still outnumbered by voters who thought that WorkChoices would be bad for jobs. Indeed, as figure 6 shows, the proportion of Newspoll respondents believing that WorkChoices would be bad for jobs rose over time. The last time this question was asked, in March 2007, 48% of people said WorkChoices would be bad for jobs, and 33% said it would be good. A similar pattern of answers was seen in another Newspoll question, on whether WorkChoices would be good or bad for the economy. In March 2007, 51% of people thought it would be bad, and 32% believed it would be good.

The government’s main case for WorkChoices creating jobs rested on its claim that unfair dismissal law deterred employers, especially in small business, from taking on new staff. Though support for abolishing unfair dismissal claims rose when smaller employers were mentioned, it was still found in only a quarter of those polled—about the level of overall support for WorkChoices. The counterintuitive idea that making it easier to fire also makes it easier to hire did not persuade. A belief that many workers would be ‘unfairly’ sacked, though never tested in the polls, may explain why so many people thought WorkChoices would be bad for jobs.

A second argument for WorkChoices as a job creator, made more by economists than the government, was that employment would increase if workers were not priced out of the labour market. The public understood that under WorkChoices wages could go down as well as up. A Galaxy poll in July 2005 found that 63% of respondents were worried employees would be forced to accept lower wages and conditions. Though the idea that lower prices encourage higher consumption is familiar from everyday experience, it seems that this argument was also lost.(11)

But why, over time, did more people come to see WorkChoices as bad for jobs? Between October 2005 and March 2007, the proportion increased by nine percentage points. The short answer is that underlying opinion, as in the overall support/oppose questions, was stable. What changed was the proportion of respondents offering an opinion. As the percentage of uncommitted respondents shrank, more people answered ‘bad for jobs.’

Figure 6: WorkChoices and job creation

Question: Overall, do you think these changes will be good or bad for creating jobs?
Source: Newspoll/The Australian

The stability of opinion on WorkChoices as bad for jobs is surprising. After WorkChoices, employment increased strongly and the unemploy-ment rate fell. With every monthly release of employment statistics, the idea that WorkChoices was bad for jobs became less tenable. On this point, the public’s unsophisticated understanding of economics, under which coincidence and cause blur, ought to have worked in the government’s favour. Yet it did not, and the government went backwards in public opinion on WorkChoices and jobs.

More than any other WorkChoices survey finding, the lost argument on jobs displays WorkChoices’ political failure. Though WorkChoices critics explained away post-implementation job gains as being due to other factors, few asserted that WorkChoices would be bad for job growth overall (even if some individuals would be unfairly sacked). The government lost a point its opponents did not seriously contest. With this loss, hope evaporates that voters might accept a trade-off of weaker conditions for more jobs.

The future of workplace relations reform
Whether workplace relations reform has a parliamentary voice depends on how the Liberals interpret their election defeat. The May 2007 fairness test backdown, along with ruling out further reform during the campaign, suggests that senior Liberals see WorkChoices as an electoral liability. Its consistently very negative polling will count heavily against future Liberal leaders revisiting the issue in Opposition.
Can public opinion be improved? Some public beliefs are open to challenge with empirical evidence. Employment statistics, for example, contradict the idea that WorkChoices is bad for jobs overall. Contrary to fears about diminished job security, legal job protections have little impact on retrenchment rates or employee fears of dismissal.(12) Yet if WorkChoices is seen as bad for jobs despite widely reported good employment figures, a counterintuitive case against job security laws seems unlikely to convince.

Most employees with market-set wages, as table 1 suggests, perceive no troublesome power imbalance in their own employment relationships. This, however, is unlikely to dissuade them of the view, seen in polls on wage-setting methods and unfair dismissal, that other workers would be paid too little or sacked too easily without labour-market protection. ‘Fairness’ has been part of Australia’s industrial relations ideology for generations, a point the Coalition conceded by rebranding its ‘no disadvantage’ test as a ‘fairness’ test.

While think tanks, academics, and business lobbies will continue to argue about how we should see labour markets, expectations of change should be modest. WorkChoices had behind it twenty years of debate on the issues, the Liberal Party support base, and millions of dollars in advertising. Yet all this wasn’t enough to get it past 25% support.

Unfortunately, the more likely way that industrial relations reform will return to the political agenda is via something nobody wants: economic conditions sufficiently bad that the issue cycle turns back to where it was in the 1980s and 1990s. That was a remarkable period in which economic policies could be implemented in defiance of public opinion but without causing electoral defeat. Voters did not support tariff reduction, opposed most privatisation of government-owned firms, and at the time supported the industrial relations status quo.(13) Yet until the Coalition’s defeat in November 2007, government changed only once in nearly a quarter of a century, compared to three changes in the preceding twenty-five years.(14) It suggests that the right circumstances, combined with good political management, can minimise the electoral impact of unpopular reforms.

Unfortunately for the Coalition and Work-Choices, the times were against them. The success of previous economic reforms removed the political conditions needed for further reforms. At this point in the issue cycle, workplace relations reform is a solution that voters think is worse than the problems it is designed to fix. While prosperity lasts, we have reached the end of industrial relations reform.

The polling data: Much of the polling research on which this article is based is available on the web. See Newspoll, http://www.newspoll.com.au; Roy Morgan Research, http://www.roymorgan.com.au; and for the Australian Survey of Social Attitudes and the Saulwick Polls, http://assda nesstar.anu.edu.au/webview/ index.jsp. ACNeilsen research is reported in the Sydney Morning Herald and The Age. A search of these newspapers’ websites will show reports of many of the polls reported here. Galaxy polls appear in News Ltd tabloids.