I learned something from listening yesterday to Radio New Zealand’s 4-5pm show “The Panel”.  Bernard Hickey was one of the panellists and he was waxing eloquent about the apartment building boom apparently underway in Melbourne.  I’ve long known that Australia was one of the handful of countries in which voting is compulsory in federal elections (in company, apparently, with North Korea, Ecuador and so on, as well as a handful of more respectable stable democracies).

What I hadn’t known, and which Bernard highlighted, is that voting is also compulsory in Australian state and local authority elections.  In fact, it isn’t a universal requirement for local elections, but voting certainly is compulsory in local council/mayoral elections in New South Wales, Victoria,  and Queensland, the three most populous states, with the three largest cities.   Compulsory voting – at least in Australia, North Korea may be different –  doesn’t produce anything like a 100 per cent voter turnout.  But turnout in Australia is far higher than in New Zealand, where voting is voluntary.  At our last local authority elections in 2013, even with postal voting, turnout was only 39 per cent.

This all got me thinking about one explanation sometimes offered for the dysfunctional (“rigged”) housing supply market.  Why, people sometimes wonder, do the existing land use rules persist, even though they seem to put the hurdles in face of starting out as a home owner ever higher?  Our Productivity Commission weighed in on that topic in their land supply report released late last year  The Productivity Commission’s report was very sympathetic to local authorities.  As I noted at the time:

The Commission shows no signs of unease with the concept of urban planning, and indeed seems to treat as wholly legitimate the choices of local councils to pursue particular visions of urban form (especially compact ones). It is simply those pesky voters who stand in the way of councils realising their visions.

The Commission took the view that turnout in local authority elections was one of the problems.  They noted that older age groups were more likely to already own existing houses, and were also more likely to vote in local authority elections.

The significantly higher voter participation of older groups in local government elections, and the markedly higher home ownership rates among older New Zealanders, means that homeowners are likely to be the dominant voters in local government elections.

And

The dominance of homeowners in local government political processes could help to explain a number of the characteristics of land use regulation and the provision of infrastructure discussed in subsequent chapters of this report.

And

Local politicians will find it particularly difficult to resist the preferences of existing homeowners where those owners organise into residents’ associations, where ward voting makes councillors responsive to particular communities, or where community/local boards are formally established to act as a voice for an area.

Before concluding with an official Finding (F3.17 on page 60)

Groups that have high home ownership rates have higher rates of participation in local government elections. The influence of homeowners in local government elections and consultation processes promotes local regulatory and investment decisions that have the effect of reducing land supply for housing.

The young, the renters, the new arrivals disproportionately choose not to vote, and so they get done over in the political process.  House prices stay high as a result.

I’ve never found the story particularly persuasive.  It might be an element in a story for why existing home owners can often limit more intensification in their own neighbourhoods.  But private covenants  –  now pretty pervasive, as the Commission recognises – represents voluntary market instruments to achieve much the same sort of protections for new developments.  But existing home-owners have no strong or permanent interest in preventing the physical expansion of their city, provided that the costs of expansion are appropriately allocated.   And homeowners have children and grandchildren – who will want to buy homes in time – and we might reasonably suppose that they care at least as much about the interests of those generations as local councilors and council bureaucrats do.

I didn’t find it a persuasive story –  although it was a convenient story for council staff and Wellington bureaucrats to tell each other.  It can be hard to find  good voters to back bureaucrats’ preferences….

But the compulsory voting arrangements in Australia provide us with a bit of a natural experiment.  Voting in Australia has been compulsory for a long time, and it has always been voluntary here. I think it is safe to treat those arrangements as prior to the expansion of urban land use restrictions.

If the Productivity Commission’s story was correct –  as any material part of the story – we should be able to look across housing markets in Australia and New Zealand, and see what difference compulsory voting and voter turnout in local elections makes.   Price to income ratios should be lower in Australia than in New Zealand, and lower in compulsory voting states of Australia than in the other states.

The Demographia report is the best collection of price to income ratios.  Here are the New Zealand cites/regions they report data for, from late last year.

There are a lot more cities in Australia (Demographia capture all those with populations above about 100000 –  details are here).  Here are capital city multiples, and the median market multiple for all the Australian cities.

Looking across the New Zealand and Australian cities, there doesn’t seem much prima facie evidence to support the Productivity Commission voter story.   These are snapshot numbers, and the picture might look a little different if one compared markets in a different year.  And there is always other stuff going on in each market –  rating schemes, land taxes, stamp duties, foreign investor restrictions, capital gains taxes, marginal income tax rates and so on –  but there doesn’t seem to be much support for the “middle aged and elderly capture the electoral process and skew housing markets to their own advantage” hypothesis.  In fact, I noticed that in Victoria, voting is compulsory only until age 70 – a concession, apparently, with the effect of (modestly) favouring the interests of the relatively younger groups.

One could generalize the point.  Voting isn’t compulsory in UK national or local body elections and their housing supply markets seem just as dysfunctional.  Voting also isn’t compulsory in the United States, and yet we see hugely different housing supply markets (and housing affordability outcomes) in different growing cities, largely reflecting differences in land use restrictions.  Atlanta is one of the success stories.  When I checked, I found that voting was once compulsory in Georgia –  but it ceased being so in 1787.  It doubt that is making much difference today.

All regulatory provisions are endogenous –  they arise out of political and bureaucratic processes in a variety of often complex and not particularly transparent ways.  In democratic societies, the public can –  ultimately – defeat any law or regulation, if enough of them care enough.  But the Australian experience suggests it is a much more complex story than one which casts local government (and even more central government) as the “good guys”, looking out for the interests of the marginalized, while the middle aged and elderly who choose to turn out to vote, rig the system in their favour.   I’d be putting much more weight on the ideologies and values of councils and their staff, and the entire ‘planning’ industry, (reinforced or enabled by central government officials and ministers).  Make an issue complex enough and sufficiently non-transparent, and it can take a long time for enough people to really realise what has been going on.  And by then the mess can certainly be very –  politically –  difficult to undo.