## Monday, August 31, 2015

### So, about the inefficiency of stamp duties…

Australia’s economic commentariat is now almost unanimously on board with the idea that stamp duties on property transactions are immensely inefficient and should be abolished in favour of land taxes.

I’ve long held the view that land taxes are the best form of taxation. But the idea that stamp duties are exceptionally bad is not clear cut. Key reference points for this belief are the modelling exercises of economists estimating the welfare losses from these taxes.

The main problem, however, is that there are no transactions in the equilibrium economic models they use, so there is no way to model a transaction tax such as stamp duty. Equilibrium models are ‘pre-solved’ by the Walrasian auctioneer to determine the distribution of goods to a single representative agent.

Here’s what the Australian Treasury had to say when they tried to model the welfare effects of stamp duties.
It is inherently difficult to capture this type of capital transaction tax in a model with a single representative agent. The approach adopted here treats real estate services as an investment good which improves the productivity of the firms, including the housing sector. One way of thinking about this is that real estate agents play a valuable role in finding producers that value the capital the most. Therefore a potential owner will be willing to pay a real estate fee equal to the profit they will enjoy over the previous owner. Within this setting the conveyance duty is treated as a tax on the value of investment and subsequent productivity gains facilitated by the transfer of land and structures.
Translated it reads “our model can’t capture transaction taxes so we’ll just assume the tax is something else to fit it into the model we do have.”

The best micro-level analysis comes from Davidoff and Leigh, who find that the main impact of higher stamp duties is to reduce the frequency of home sales, and of those home sales, some will be from people relocating. In addition, stamp duties are fully incident on the landowner, meaning that they cannot be considered a tax on investment as they are in the Treasury model, since a higher stamp duty lowers property prices by an equal amount. It doesn't add anything at all to the cost of property.

It is not clear that the welfare effect of reduced home sales is negative if some (or many) of those sales are merely fuelling speculation in the housing market. The basic result of all transaction taxes in asset markets hold-- if some of the transactions are simply speculative churn then there can be large positive welfare effects from reducing turnover through transaction taxes.

So I urge caution about calls to cut stamp duties, even if those calls are accompanied by the proviso that such a change must be accompanied by higher land taxes, and especially if those provisos could likely to be ignored.

2. What kind of stamp duties are you talking about? When I bought my first house in the Boston area my attorney explained that I had to spend $250 on a stamp for the transaction. He noted that while the American revolutionaries were willing to rebel over a colonial stamp tax, the first act of the free Massachusetts legislature was to enact the very stamp tax I was paying as part of my closing costs. It was an amusing bit of history, like waiting for jury duty in the same court room that Daniel Webster used to plead his cases, though I gather not his famous one against the devil. I think whole transaction cost maybe$80K and I was on the hook for a 12.25% mortgage which I considered a fairly good deal at the time. $250 was a drop of blood in the hemorrhage. I wanted to buy a house, I needed a place to live and available rental properties were sucky, to be polite. I would have paid$2500 without a lot of argument.