Last week, the Australian Bureau of Statistics released its biennial trade union membership statistics. The results were dire for unions, but they also have implications for further reform of the Fair Work Act.
The ABS estimated that only 12.5% of employees belong to a union in their first job. Once allowance is made for how the ABS treats people whose housemates answer “don’t know”, and the fact that some people belong to a union in their second job, the true rate of union membership is probably closer to 14%. Still dire, just not quite as much.
The continuing drop in union density (the share of employees who are union members) has major implications for government economic policy.
Numerous studies show unions raise wages, and that unionists have higher wages than non-members. When union density is higher in a workplace, wages in that workplace are higher — especially when a collective agreement is negotiated.
So it is that union-based enterprise agreements in Australia have consistently had higher wage increases than non-union agreements — on average, by about 0.2% per year since 2017. The gap would be larger if unions’ bargaining power was not dragged down by the high number of non-members covered by negotiations for union collective agreements.
Studies have had trouble consistently estimating by just how much unions raise wages, in part because unions have ‘spillover’ and ‘threat’ effects on non-union wages. If unions are strong, when unions raise wages then unionised employers raise non-union wages to avoid resentment (and legal action), while non-union employers raise wages to prevent a labour shortage (the spillover effects) and to reduce the incentive facing their employees to join a union (the threat effect). This makes a simple comparison of union and non-union wages, even after controlling for other things like age, industry, occupation and so on, misleading, and possibly understate the union impact on wages.
When union power declines, these effects diminish, as employers become less concerned about losing staff to unionised firms and less worried about the need to offer higher wages to discourage unionisation. So the gap might increase. On the other hand, declining union power can directly reduce the union wage and hence the gap.
So overall, whether the measured gap between union and non-union wages rises or falls depends on the net effect of these forces.
Even if the benefits of belonging to a union might look like they’re growing, it doesn’t mean they are. They might actually be shrinking, as unions lose their power.
So it was that, when I started studying unions in the 1980s, the country with a very large union/non-union wage gap — probably the largest in the developed world — was the United States, also the country with the one of the lowest union densities and the weakest unions.
Wages matter for the Australian government, because it was elected on a promise of doing something about low wages. Its recent Secure Work Act, amending the Fair Work Act, was directed at this. Failure to raise wages could not only have political consequence, it could affect tax revenues, consumer demand, economic growth and budget deficits. And with labour markets showing monopsonistic tendencies, there’s no guarantee there would be offsetting employment gains.
The latest union membership show just how difficult raising wages is becoming. The issue is not just the fall in density.
In 2022, the median union member received 26.1% more per hour than the median non-member.
That’s quite a difference. In particular, it’s a lot larger than the 20.6% gap in 2014, or the gap in any year since then.
And it’s huge compared to the gap of just 14.9% in 2006. (I focus on median hourly, rather than weekly, earnings because the latter are distorted by the number of hours per week that people work.)
Those estimates do not control for anything else other than hours worked, but that gap has become large.
The gap between union and non-union wages has widened most for part-time employees.
There are probably a couple of factors at work here. One is the changing composition of the unionised workforce, picking up a greater proportion of people in higher-paying work. That accounts for part, but not all, of the widening union/non-union pay gap.
That aside, it is definitely getting harder for unions to generalise the wage increases they are earning for their members. Awards no longer do it for them. The ‘spillover’ and ‘threat’ benefits from unions to non-union members decline as employers see unions as being less powerful.
True, employers still have a financial incentive to increase profits through discouraging union membership. And as the gap between union and non-union wages rises, that incentive on employers grows. But it’s no longer an incentive to discourage unionisation by raising wages of their non-union staff. Instead, it’s an incentive to focus more on some of the union-busting techniques, including possibly illegal ones, that gained prominence in the USA in the 1980s.
This situation increases the strength of arguments in favour of multi-employer bargaining, — but not the sort that the Secure Work Act enabled, which is mostly about encouraging multi-employer bargaining across non-union workplaces. The non-union wage won’t be raised to the union wage by this, because unions will find it too hard to sign up so many non-union workplaces.
Instead, multi-employer bargaining needs to include unionised workplaces with collective agreements. These are presently excluded from what will be the most important new form of multi-employer bargaining. In other countries, multi-employer agreements cover both union and non-union firms. The same should apply in Australia.
The Fair Work Act should also be amended to remove the remaining, substantial, restrictions on union activity (for example, in the logistics of secret ballots and legal industrial action). Otherwise employers will have growing incentives and a continuing ability to keep unions out. The results will be lower wages for all employees, and economic and political problems for the government.