This is a post I have been meaning to write for some time, but the current spur is the Treasury briefing for the incoming minister. It is mostly very good, but included in it is the canard about over-investment in housing. This is a refrain we hear repeatedly, usually in phrases like “New Zealanders love affair with housing”, but I think the theoretical and empirical basis for the assertion of overinvestment is weak at best.
The main issue is one of tax-induced distortions. With owner-occupied housing, the purchase of a house is an investment, which generates an inputed rent that the owner pays to himself. The payment of this imputed rent is not subject to GST, nor is the rental income subject to income tax. There is also the issue of whether the absence of a capital gains tax is a further distortion, but I have discussed that issue here, here, here, and here. In this post, I want to explain why I don’t find it obvious that there is a distortion due to the lack of tax on inputed payments of a homeowner to himself. In part, this is a response to comments by Phil Meguire in the first of the capital gains posts.
To start with, let’s get an obvious point out of the way. Housing is a good thing, and nicer housing is better than worse housing, holding all else the same: If non-neutral taxes induce more investment in housing and less in other forms of investment than a neutral system, there is a cost, but it is the difference between the value of the flow of goods that would have been produced and the value of the flow of services produced by the housing, a difference which is not necessarily significant. I suspect that some of those decrying our love affair with housing are making the mistake of thinking that just because the purchase of housing services is a non-market transaction it doesn’t contribute to economic well-being. Similarly, it is irrelevant if the alternative investment would increase labour productivity, increase exports, etc. productivity and exports are just means to increasing the value of the goods the economy can consume; so is building houses. (Surprisingly, Greg Mankiw appears to make this mistake here, but to be fair we can attribute this to the need for simplification in an op-ed article, and his basic point—that the U.S. should get rid of the crazy mortgage-interest deduction—is sound.)
The next point to note is that the optimal tax rate on owner-occupied housing has to be considered in the context of an existing distortion: With an income tax, saving is subject to double taxation. If I choose to work today in order to buy consumption goods today, there is a difference between the value of what I produce and the value of what I consume, because of an income tax. But if I choose to work today in order to buy consumption goods in the future, there is an even bigger distortion between the value of what I produce and the value of what I consume, because the value of my work is taxed twice, once on my labour income and again when the interest income is taxed. We then have a classic second-best Ramsey taxation problem. The theoretical optimal tax on the return from investing in owner occupied saving will be somewhere between having no taxation on the saving (so as not to distort the decision between consumption today and investment in housing), and the full double-taxation rate applying to other forms of saving (so as not to distort the decision between investment in housing and other investment). So if normal investment returns are taxed at, say, 33%, the optimal housing tax will be positive but less than 33%.
And finally, here is the point that is overlooked in all the commentary on housing investment that I have seen. The flow of services from owner-occupied housing is currently subject to a tax. In other words, this graph, taken from the Treasury’s briefing to the incoming-minister, is wrong.
To see why, consider how a value-added tax like the GST works. Businesses charge GST on their sales, but deduct from their net tax liability the GST they have paid on purchases of goods and services. Expenditure on investment therefore reduces their tax liability, which is what makes the GST ultimately a tax on consumption and not investment income. When consumers buys a new house, they are making an investment, the return for which is the flow of housing services they will receive over time. The inputed payments they make to themselves are not subject to the GST, but unlike other business investment, the initial house purchase is, which has the same effect. (And, as I noted in an earlier post, the existence of a GST pushes up the prices of second-hand goods by the rate of the tax, so the argument is no different for owners who purchase an existing rather than newly built house.)
This is another example of the beauty of pure value-added taxes. By shifting in part from an income tax to a value-added tax, the rate of double taxation on delayed consumption is reduced. At the same time, the effective rate of taxation on the consumption of owner-occupied housing is increased from zero. We end up with a system where the tax rate on owner occupied housing is somewhere between 0 and the rate applying to other investment, just as Ramsey tax theory would stipulate. Whether it is higher or lower than the optimal rate is a difficult empirical question to which I don’t know the answer. But I have not seen any of “New Zealander’s love-affair with housing” commentators seek to address it. Furthermore, given that there are no simple methods of increasing the tax rate on owner-occupied housing that don’t bring their own problems (we have no mortgage-interest deduction to get rid of, for instance), I find it hard to believe that we have an over-investment problem that is worth solving.