Andrew Barker's address on housing mobility

CEDA Senior Economist Andrew Barker delivered an address to the Australian Conference of Economists 2023 on housing mobility this week.

My name is Andrew Barker, I’m a Senior Economist with CEDA, the Committee for Economic Development of Australia. I’m based here in Brisbane, so thanks for coming to us, it’s great having the conference in your hometown.  

What I’m going to be talking about today is housing mobility. The background to this work is it came out of the work that we were doing at CEDA for our submission to the Federal Government’s Employment White Paper that followed on from the Jobs and Skills Summit last year. The focus there was very strongly on the need for a more dynamic labour market, so more job mobility, getting people moving to promotions and better-suited jobs.

If you’re familiar with the data, job mobility in Australia has fallen over the past four decades and is now in the bottom third of OECD countries, so a long way below the leading countries for having a dynamic labour market, such as the Nordic countries or the United States.   

One contributing factor is the barriers to moving house and OECD work has estimated that this could hold back productivity in Australia by as much as four per cent through the channel of skills mismatch – the idea that you can’t move to a better-suited job if you can’t afford the housing in the location where the job is.   

Above and beyond this, housing mobility is very important to get people into better-suited housing, which is important for individual wellbeing, and a well-functioning housing market shouldn’t create barriers to housing mobility or to moving house.  

Of course, there are reasons why people might not want to move. Having moved house in the last 12 months, I can attest to how much of a hassle it is. You also might not want to move away from social networks or family, but we shouldn’t have barriers in the housing market that prevent people from moving who would want to otherwise.  

I’ve listed here what I’ll talk about, and I’ll finish up with some analysis of the prevalence of housing mismatch in Australia and some of the factors that may contribute to this. I’ll focus more on policy rather than complicated econometrics or models because that’s what we do as a think tank at CEDA.  

This is some analysis that my colleague was doing while I was at the OECD, and the position of Australia in this chart as an outlier made me stop and take notice that maybe we’ve got something funny going on in our housing market.  

On the X axis, that’s the share of the households that moved house in the past five years, and on the Y axis is the share of households who moved house for work in the past five years. So, there’s a lot of housing mobility in Australia, but there’s a question of whether it’s the right type of housing mobility. Compared to a country like Germany, where people often move house for work reasons, Australians less frequently move to go to a better job, for example.

Drilling down into this data a bit further, private renters are the most likely to move in Australia, even when you adjust for the different characteristics, for example being younger. Renters in Australia are more likely to be forced to move by their landlord than to move for work or study reasons. When we compare this internationally, that’s pretty unusual.  

If you compare that to the United States, which is hardly a renters’ paradise, renters are 10 times more likely in the United States to move for work or study reasons than to be evicted by their landlord. It’s quite unusual this finding in Australia that people are more likely to be evicted.  

This is data that no doubt many of you are well familiar with this – the declining home ownership across cohorts, with younger people less likely to own a house. As the Productivity Commission pointed out in the National Housing and Homelessness Agreement review last year, this is not necessarily a bad thing. Owning a house is not necessarily welfare-enhancing in and of its own, but it does make it increasingly important to have a well-functioning rental market.  

When we compare Australia against other countries, rental tenure securities are very low. There’s an index ranking 31 OECD countries and Australia’s equal lowest for rental tenure security alongside Greece. Australian renters are also relatively constrained in their capacity to make small alterations to the dwellings they’re renting.    

This is quite important because other analysis has shown that security of tenure is protective of mental health, so controlling for other factors, renters have lower mental health on average than owners, but after five to six years of tenure, it catches up and there’s not a statistical difference anymore. What’s going on here is not renters wrecking houses and getting evicted for that reason, clearly, landlords should have the capacity to evict people who are not looking after the house or not paying the rent, but the overwhelming majority of the landlord terminations are for reasons totally separate to the renter, like families moving in or selling the property.  

The first step that we recommend in this work is to constrain the capacity or remove the capacity of landlords to evict tenants without grounds. There have been moves in that direction in Australia across territories, but in several states and territories it’s still possible to evict someone with only 30 days’ notice at the end of a fixed tenancy.  

The second step is to look into using simple metrics to regulate price increases of rents. There’s no point in not allowing people to evict people for no grounds if they can just give them 100 per cent rent increases, so the two need to go hand in hand.  

There’s a wealth of economic research showing that, across the board, rent control is a very bad thing because it dissuades investment in housing and reduces mobility. There’s OECD analysis on this too, showing in countries like Sweden that have rent control, they’ve actually had very poor outcomes in terms of high rents and big waiting lists to get into those housing.  

There’s still a concern that existing tenants can be in a situation of economic hold-up where they don’t want to move because of school or family reasons and so that enables the landlord to put rents up by more than the market amount. One simple solution that was done in Germany with relative success, it was shown not to have deterred investment, was to have a maximum rental increase for sitting tenants that’s linked to a market measure of the local rent increase in that local area. That ties the rent increases back to what’s going on for newly advertised tenancy, so it puts a cap on that, maintaining the market link without putting existing tenants at risk.  

The third step is to level the playing field for institutional investors. One reason perhaps why renters are more likely to move of their own accord in the US rather than to get evicted by their landlords is they have a much stronger build-to-rent sector, and one thing that works against that in Australia is the higher incidence of taxation on housing owned by institutional investors.

Two of the three factors are included here, the other one is for foreign investors there’s a higher rate of withholding tax compared to commercial and student housing investment, but that was actually reduced at the last budget. These two are access to negative gearing, which of course institutional investors don’t have that ability to offset against wage income, and land tax because our land tax is done in a progressive way. If you own a number of different properties then you are more likely to be paying the top rate of land tax, and so this is just an illustrative example using median house and land values in Sydney to show, particularly for housing that has a lower rental yield and a higher relatively higher value of land, how much of a disadvantage that can be to institutional investors.

Whereas when there are less of these barriers, for example with student housing in Australia or in the built-to-rent sector in the US where negative gearing is not possible, we can see more institutional investment coming on, so that playing field should be levelled.

Stamp duty is not exactly economists’ favourite tax, so I'm probably preaching to the converted here, but it's estimated that a 10 per cent increase in stamp duty reduces housing turnover by six per cent, so reducing turnover, as we're concerned about, reducing housing mobility and then the costs for the labour market associated with that.

It's worth recognising that stamp duty revenue, this is adding up across all the states, has soared recently due to both price and turnover effects because there are a lot of sales and that also makes it a bad tax for governments from a fiscal perspective because it's a very hard tax to forecast because of these big changes. But it's been politically difficult to make the shift to more efficient taxes like land tax, and there have been some suggestions recently on how to fill that fiscal gap, through either funding from the federal government, who gets the benefits from increased efficiency, or Bob Breunig had some suggestions around giving credit for those who have recently paid stamp duty but accruing land tax over time.

I'll now come to some of the other potential barriers to moving for people owning houses. This is some analysis of the 2021 Census and it's not really a surprising finding that people have more spare bedrooms as they get older, as dependents move away from home. This real concentration of mismatched housing, of having a lot of extra spare bedrooms among older Australians, motivates the analysis we did for this study, particularly looking at the moving decisions of older Australians who own their house.

Stamp duty is of course the number one barrier to these people moving house because of a significant transaction cost, but another is receipt of the pension because the family home is excluded from the asset test in the pension. That can be a barrier to people downsizing and increasing their non-housing assets and therefore potentially reducing the amount of pension that they include.

To look into this further, particularly for this 65-plus age group, we estimated a very simple model of probability of moving house, controlling for other relevant factors such as age, income, marriage status, employment status and the number of bedrooms in the original house that you're starting in. Causally, we're of course interested in the receipt of the pension affecting your decision to move house rather than vice versa, so we also estimated a model based on whether you were receiving the pension in the year before you chose to move.

In terms of data, the HILDA data set has a very useful question about whether you moved house in the past year, so this enables us to do this analysis on a pooled sample of HILDA data rather than actually tracking people over time. Then we dropped all the observations for people aged under 65, we're focusing on the pension, and multiple observations in the same year from the same household.

For the second specification, we needed two consecutive observations, so only kept those households with two consecutive observations, so we could see if they were receiving the pension in the year before they did or did not move house.

The results here, we found evidence of a correlation that was statistically significant at the five per cent level between the receipt of pension income and therefore a lower probability of moving house. There was a slightly weaker relationship between whether you received the pension in the year before the move. It was only statistically significant at the 10 per cent level. It's in the expected direction, you're less likely to move house if you receive the pension, but it’s not a lot to hang your hat on, only statistically significant at the 10 per cent level.

Summing up, as I was saying at the start, housing mobility is important to enable job mobility, which enables better job matching and thus better productivity. What's new about this work? We show just how unusual that dominance of involuntary mobility is for Australian renters and how unusual that is internationally. We show the potentially significant magnitude of tax disincentives to institutional investment in rental housing, particularly for housing, rather than apartments, given the higher land value and lower yields and therefore more likely to be negatively geared.

What we say about stamp duty certainly isn't new, plenty of people have said that before about it being a barrier to mobility. We show empirically the link between the pension and the lower probability of moving house, so it might be small compared to some of the other reasons for moving house that dominate the decision which is also something that the Productivity Commission showed in their research, but it does contribute to the barriers to downsizing and therefore the housing mismatch that we saw from the 2021 Census data.