Andrew Leigh

‘EXPLAINING THE RISE OF AUSTRALIAN INEQUALITY’

JUST IDEAS TALK #2

PER CAPITA’S REFORM AGENDA SERIES

MONDAY, 5 DECEMBER 2016

There are many forms of inequality, but perhaps the starkest is the difference between those who own no assets and earn their living by selling their labour – and those who earn vast assets, and can live off the proceeds.

Between these two extremes lies home ownership. It’s not a perfect marker, but if you don’t own a home, it’s likely you live by the sweat of your brow. Conversely, if you’re living off your investments, it’s a pretty good bet you own your home.

At the end of World War II, Australia was a nation where just 53 percent of households owned their homes.[1] In the major cities, the figure was just 46 percent.[2] Most city-dwellers rented. And most homes were made of wood or fibro cement.

Then in the post-war years, something remarkable happened. The Australian home ownership rate surged. By 1954, it was up to 63 percent. By 1961, it was 70 percent. In just over a decade, the distribution of Australian housing wealth became significantly more equal.

It wasn’t just homes. Shared prosperity in the post-war decades meant cars became cheaper. By the 1960s, most Australian homes had a vacuum cleaner, a washing machine, a television and a fridge – items that in the pre-war era were only owned by the most affluent.[3] Even access to university was shared. For someone like my grandfather Keith Leigh, attending Melbourne University would have been impossible on a modest clergyman’s wage. Only a post-war veteran’s scholarship made it feasible.

The intellectual seeds for these changes were sown in John Curtin’s white paper on full employment, and his clearly professed view that ‘there will have to be a fairer distribution of wealth’.

But the surprising thing is what happened next. 

Under Robert Menzies, the Coalition did not choose to mount a stirring defence of property investors. They did not rail against attempts to increase home ownership as ‘socialism by stealth’. Instead, they presided over a steady rebalancing of the economy, and even went so far as to incorporate home ownership into their story of what the Liberal Party stood for. With the home ownership rate now the lowest it has been in six decades, this is worth remembering. The Liberal Party has not always been the party of landlords.

When it came to wages, Menzies presided over average wage growth that was five times faster than it had been in the first four decades of the twentieth century.[4] With today’s conservatives fearful of a ‘wages breakout’, this is worth remembering. Australia once had a Liberal Party that worked to deliver strong real wage growth.

Equality is a Labor value. But it isn’t just a Labor value – it’s also an Australian value. Robert Menzies knew this. But do his heirs?

This is the second of three talks on inequality in a series I’m calling ‘Just Ideas’. In the first talk, I focused on the question of why we should care about inequality. Today, I will be looking at the drivers of inequality. The final talk will look at what we might do to reduce inequality.

Unions

Let’s start with the role of union membership. At the time of Australian Federation in 1901, union membership was relatively rare, with about one in ten workers in a union. But in the next dozen or so years, the share of union members quadrupled, reaching four in ten by 1914. Not coincidentally, this was also the period of most rapid growth for the political party representing organized labour, the Australian Labor Party. For Labor and labour, the early years of the twentieth century were heady days indeed.

From 1920 to 1980, the union membership rate in Australia averaged 49 percent, which is where it was in 1982. Then it began to plunge. Forty percent in 1992. Thirty percent in 1997. Twenty percent in 2006. According to the latest figures from the Australian Bureau of Statistics, 15 percent of the workforce are now in a union.[5] In the United States, the figure is 11 percent.

At this point, it’s worth noting that the Australian Bureau of Statistics’ figures are for 2014, and more recent data from the Australian Council of Trade Unions suggests a higher membership rate.[6] But with that caveat in mind, let’s stick with the Australian Bureau of Statistics figures, since these allow us to look at different sub-groups. Their headline figure suggests that union membership is now back where it was in 1906 – 110 years ago. Among private sector workers, the official membership estimate is 11 percent. For overseas-born workers, it’s 12 percent. For workers aged under 25, it’s 7 percent. Among the industries where union membership has now fallen below one in ten are accommodation, agriculture, defence, printing and performing arts.[7] If the current rate of decline continues, the overall union membership rate will reach 10 percent by 2022.

Over the years, unions have brought about lasting gains in the workplace. Sick leave in the 1920s. Annual leave in the 1930s. The eight hour day in the 1940s. Unfair dismissal protection in the 1970s. Banning asbestos in the 1980s. The weekend. Careful economic research finds that unions have a causal impact on making workplaces safer.[8] Today, unions are making the case for family and domestic violence leave.

Unions have often found themselves on the right side of history. Maritime unions refused to load ‘pig iron’ onto Japanese ships in the late-1930s because they foresaw the risk that it would come back in bombs. When 200 Gurindji people walked off the Wave Hill cattle station in 1966, it was the trade union movement that supported the right of Indigenous people to be fairly paid. If you’ve ever enjoyed Centennial Park and the Sydney Botanic Gardens, then you should thank the union members who stopped them being destroyed in the 1970s.

Unsurprisingly, unions also increase wages. One recent study finds that unions increase wages by 5-10 percent.[9] As an aside, given that union dues are generally 1 percent or less, this is a pretty good rate of return.[10] But for the purposes of thinking about inequality, it’s important to note that the impact of unions is strongest at the bottom of the distribution. Unions often campaign for dollar pay increases rather than percentage increases, for pay equity across workplaces, and for pay equity across industries.

The 2012 Social and Community Sector Equal Pay Case is a good example of this. In bringing the case, the five unions noted that around 80 percent of workers in the social and community sector were women. They argued that because they did ‘caring work’, they had been systematically underpaid, compared with those in other occupations who had comparable skills and worked in similar conditions. Fair Work Australia agreed, and laid down an eight-year transition to better pay for these workers.

But as union membership declines, it will become steadily more difficult for these kinds of cases to be brought. One analysis by Jeff Borland estimates that in the 1980s and 1990s, declining union membership was responsible for up to one-third of the increase in Australian wage inequality.[11] This parallels what has been found in Britain and the United States.[12] 

To illustrate the point, take the gender pay gap. Using the Melbourne Institute’s Household, Income and Labour Dynamics in Australia survey (HILDA), I calculated the gender pay gap in hourly earnings across nearly 8000 workers. Among non-union workers, women earned 13 percent less than men. But among unionised workers, the gap is 7 percent, approximately half as large. This suggests that unions play a role in narrowing the gender pay gap.

The same holds true for Indigenous workers. My analysis suggests that if they’re not in a union, Indigenous people earn 18 percent lower hourly wages. But among unionised workers, the wage gap between Indigenous and non-Indigenous workers closes to 5 percent. As with the gender gap, unions don’t close the racial pay gap entirely, but they make a big difference.

Every time you see someone proposing policies that make it harder for unions to organise, remember what this means for our economy. A larger gender pay gap. Indigenous Australians being left further behind. More economic inequality.

Technology and Globalisation

Another significant driver of inequality is the twin forces of technology and globalisation. These are best discussed together because they interact. For example, most technologies Australia uses weren’t invented here. But they also impact our economy in similar ways. For example, suppose we created a machine that turned iron ore into televisions.[13] Dump a shipload of iron ore into the machine, and it would churn out 2000 televisions. That’s how trade works. You can also think of the change in the terms of trade during the mining boom as being like a big technological improvement. By the peak of the boom, the ‘trade machine’ turned a shipload of iron ore into 20,000 televisions.

In general, both globalisation and technology tend to increase the wages of those further up the distribution. Machines are particularly good at taking the place of workers doing routine jobs, which is why we have seen a precipitous decline in occupations such as typists and bookkeepers. If you line up Australian occupations in order of hourly wage, and graph the change in numbers of positions over the past few decades, you get a U-shape. In Australia, as in many developed nations, the labour market has become increasingly polarised.[14]

As these middle-paid occupations disappeared, the people who did them have cascaded down into lower-paid occupations. There are plenty of security guards and baristas – but because these occupations don’t require a great deal of training, the increase in demand for their services hasn’t translated into an increase in their wages.

In the case of globalisation, we know that open markets have brought significant benefits to Australia. For example, an analysis by the Centre for International Economics found that the trade liberalisation over recent decades benefited the average Australian household by around $4000 annually.[15]

However, it’s important to also bear in mind the distributional impacts of open markets. Since the work of Wolfgang Stolper and Paul Samuelson, economists have recognised that while trade raises aggregate living standards, the benefits are not equally shared across the community. From Peter Cook to Craig Emerson, Penny Wong to Jason Clare, Labor’s trade representatives have consistently held the view that market liberalisation must be accompanied by a strong safety net. Free trade and social democracy go together.  

Recent United States studies have reinforced the importance of this approach. In the 1990s, the US imported twice as much from developed as from developing studies. Among economists, the general consensus was that trade had only a small impact on inequality.[16]  But today, with America’s developing country imports outstripping developed country imports, studies by David Autor and co-authors suggests that trade may have a larger impact on inequality.[17]

While I have not seen a similar analysis for Australia, it is worth noting that the composition of our imports has also shifted significantly over this period. In 1990, the average wage among our top five import suppliers was 95 percent of the Australian average. Today, the average wage in our top five import suppliers is 56 percent of the Australian average.[18] This is not an argument against open markets, but a reminder of the importance of active labour market programs and a strong social safety net. 

Taxes and Transfers

Another significant driver of inequality is the taxation system. Analysing top income shares in five English-speaking countries, Tony Atkinson and I estimate that about one-third of the increase in top income inequality in the past generation was driven by reductions in top marginal tax rates.[19] Indeed, over the course of the past century, it’s astonishing how closely top income shares track the after-tax share at the top (ie. 100 percent minus the top marginal tax rate).

There were good reasons why Australia reduced the top marginal tax rate from 60 percent in the early-1980s to 47 percent in the 1990s, and to 45 percent in recent years.[20] But it’s also important to recognise that the benefits of further rate reductions at the top go overwhelmingly to the most affluent. For example, on 1 July 2017, the Coalition’s 2 percent ‘Temporary Budget Repair Levy’ will cease. This will effectively provide a 2 percent tax cut to those in the top tax bracket, which starts at $180,000. By my calculations, 94 percent of the benefits of this tax change will accrue to the top 1 percent of adults. Given that the top 1 percent has doubled its share of national income in the past generation, do they really need another tax cut?

And then there’s the question of the breadth of the tax base. During the Hawke-Keating years, Australia saw the introduction of capital gains taxation and fringe benefit taxation – changes which closed off significant tax loopholes used disproportionately by those at the top. By contrast, the Howard years saw a halving of the capital gains tax rate and the creation of new superannuation tax breaks – to the significant benefit of those at the top.

A similar debate emerges when we look at the transfer system. When it comes to reducing inequality, the Australia social safety net has traditionally been the best-targeted in the world. According to analysis by Peter Whiteford, the average advanced country gives twice as much welfare to the bottom fifth of the population as to the top fifth, while Australia gives twelve times as much to the poor as to the rich.

As with tax, there is nothing automatic about this outcome. Under the Howard Government, the redistributive effect of government payments declined by one-quarter. Labor took the opposite approach when we were in government by increasing the use of means testing to ensure those who received transfer payments were those who really needed them.

Any time we tweak the transfer system, it’s worth remembering that we’re dealing with the most efficient inequality-reducing welfare system in the world. Bolt on a regressive paid parental leave plan (as was proposed by the Abbott Government), and you make it less efficient. Increase unemployment benefit waiting periods for young people, and you increase inequality. Change pension indexation so that it’s linked to prices rather than wages, and you boost inequality. Or, as I mentioned in my last talk, government can raise the pension by one-tenth, and move a million people out of poverty.

Market Concentration

The final driver of inequality that I want to discuss today is market concentration. Now, I’ll be the first to admit that this is an area where my views have evolved markedly. When I wrote Battlers and Billionaires in 2013, I barely touched on market structure as a potential factor affecting inequality. If you’d asked me my view of markets at that stage, I probably would have said that I thought our market concentration wasn’t too serious a problem, and that the Hilmer competition reforms of the early-1990s had done most of the necessary work.

My view began to shift when I read a terrific book by my friend and co-author Tony Atkinson, titled Inequality: What Can be Done? which argues that excessive concentration can be a key driver of inequality. At around the same time, the Economist magazine published an analysis of competition in the United States, which showed that across 893 industries, markets were heavily concentrated.[21] The analysis also showed that since the 1990s, two-thirds of US industries had become more concentrated. This prompted Australian National University researcher Adam Triggs and I to conduct a similar study on Australia, which is published this month in the Australian Economic Review.[22]

Unlike the United States, where a government agency collects data on market share, our Australian analysis relied on figures collected by IBIS World, a private firm that carries out research on a variety of industries. To be precise, we went through 481 industry reports, and crunched the figures on the industry share of the largest four firms.

A common rule of thumb is that a market is regarded as concentrated when the biggest four firms have more than 30 percent. By this benchmark, most Australian industries are concentrated markets, with the average market share of the big four coming in at 36 percent.

Triggs and I find that Australian markets are even more concentrated than in the United States, where the average market share of the big four is 33 percent. To take a few examples, their big four liquor retailers have 10 percent of the market, while ours have 78 percent. Their big four health insurers have 35 percent market share; ours have 93 percent. Their big four cardboard box manufacturers have 36 percent of the market; ours have 88 percent.

There are a plethora of Australian markets where the big four have more than an 80 percent market share: ranging from baby food to beer. This is partly due to more mergers. The number of mergers that took place last year was three times as many as in the early 1990s. The value of those merger deals was nearly ten times as high.

Another indication of rising concentration is a drop-off in the number of firms in the economy. There are around 20,000 fewer Australian businesses in operation now than in 2012. For all the talk about agile innovators, new firms are significantly less likely to get off the ground today than they were a few years ago. The start-ups are stopping.

Increased market concentration hurts consumers directly, through higher prices and less variety. Disadvantaged consumers will bear the brunt of this. The poorer you are, the more likely it is that you will spend your entire pay check. Low-income consumers are less likely to avoid excessive markups by driving to the next suburb, or hopping online to get a better deal.

In a highly unionised economy, some of the excess profits from higher market concentration might have flowed through to wages. But with union membership at a 110-year low, it is more likely to show up in executive salaries or profits. And as it happens, CEO pay has massively outpaced average worker salaries over recent decades, while the profit share has risen.[23] By the way, market concentration may also be responsible for the slump in business investment we’ve seen lately.

* * *

So if anyone asks you what today’s talk was about, here’s what you might say. The big drivers of inequality are union membership, technology, globalisation, taxes, transfers and market concentration. Put another way, the strongest friends of egalitarianism include a fair transfer system and good unions, while some of the biggest challenges are robots and monopolies.

There’s one other big driver of inequality that I haven’t mentioned, and that’s education. I’ve held off talking about it today because I want to focus on it properly in my third and final talk, on what can be done about rising inequality. Watch this space.

But because we’re heading into the Christmas season, I did want to say something timely about one simple policy tool to reduce inequality: weekend loadings, commonly known as penalty rates.

Now, as every economist knows, weekends and public holidays exist for a simple reason: to solve the leisure coordination problem. Suppose two people want to meet up in a park one morning to kick a football around. In a world without weekends, both would have to coordinate their schedules with their employers. But in a world with weekends, they can simply get together on a Saturday or Sunday. Without weekends, there would be less sport, music and social activities in the world.

Weekend loadings discourage employers from unnecessarily rostering people on for weekends. In some cases, they are passed directly on to customers, as with cafes and restaurants that add a menu surcharge on weekends and public holidays.

But weekend loadings also have an impact on inequality. To see this, I took another look at that Melbourne Institute HILDA data to see who works on weekends. It turns out that weekend workers tend to be lower-paid than average. Those who work on Saturdays or Sundays earn 80 percent of those who only work from Monday to Friday. Weekend loadings go to people who tend to earn less.

Often, inequality can be complicated and statistical. But sometimes, the role of public policy economists is to cut through the dross, and say it like it is. Cut weekend and holiday loadings, and you’ll make the nation less equal. If you love Australian egalitarianism, one simple way to defend it is to support lower paid workers getting a bit extra for working when everyone else is on holiday.