Killer Arguments Against LVT, Not (221)

From Chapter 16 of the Institute for Fiscal Studies/Mirrlees reveiew, which is very much in favour of shifting from taxing earned income to taxing the rental value of land but gets all prissy when it comes to valuations:

We have, or can obtain without excessive cost, a measure of the land area occupied by each property. The Land Registry holds information on the boundaries of each property and this could be converted into area measurements. No doubt there would be some initial disputes over the size of assigned areas, but this should be a one-off transitional problem. Once the measurement of each property was agreed, it would become part of the description of that property. If we are then able to supply a land value per acre (or hectare, or square metre), it is possible to combine this with the area of the property to compute the implied land value.(1)

The biggest practical obstacle to the implementation of a land value tax, though, is that it would require the valuation of land separate from any structure erected on it.(2) If there were a competitive market for land, with a high number of transactions, then the value of land would be directly observable. But in most areas and sectors, the number of transactions in land (separate from any buildings thereon) is low. In the absence of a sizeable market, it is difficult to determine what the market price would be.(3)

It is worth noting that since we are looking at taxing a rent, the figure for land value does not have to be exact—or even approximate—for the LVT to be efficient. The value of each plot of land falls by the present value of the tax imposed on it; in principle, each plot could be taxed at an arbitrarily different rate without compromising the efficiency of the tax.(4) However, to the extent that valuations are not accurate, inequities will be created between taxpayers—just as they can be created by inaccurate valuations under the current property tax regime, but the inequities will be worse if the valuation is less accurate.(5)

1) Correct.

2) Easy. The land value is the value of the land minus the value of the building, or more to the point, the rental value of the site is the rental value of the building plus any existing taxes thereon (council tax or business rates) minus the rental value of the bricks and mortar, which in turn is a smallish percentage of the rebuild cost/replacement value plus the costs normally borne by the landlord (repairs, insurance).

We already have a system for establishing the rental value of commercial land and buildings for Business Rates; residential is even easier because about half of homes in this country are generic semi-detached houses, which are physically almost identical, so any difference in rental values between the cheapest and most expensive is entirely down to the location, if we just taxed the difference, we'd be most of the way there. Farmland is a red herring in all this, the most important thing is to scrap the negative LVT applied to farmland (i.e. agricultural land subsidies). Farmhouses are liable to Council Tax and substantial farm buildings are liable to Business Rates, and we could apply LVT to those and leave the bare land alone.

3) No, it's easy, see (2).

4) Correct. The Blue Socialists hate LVT in principle, they're going to hate it however it's calculated. All that really matters is relative values, so the main things is that a house in a nicer area doesn't pay less than an identical house in a less-nice area. And if we end up with a situation where a house in a slightly nicer area pays exactly the same as a house in a slightly-less-nice area, then no harm done, for example, if the rental value of Plot A is £9,000 and the rental value of Plot B is £7,000, and the tax on each is £6,000, then the effective rate on Plot A is 67% and on Plot B is 86%. It would be 'nicer' if they both paid exactly 75%, but so what if they don't?

Further, however the tax is calculated, it's just a number expressed in £. The owner of Plot A is indifferent whether that £6,000 is calculated as 7% of the selling price of the house at some fixed point in the recent past; as 12% of the notional selling price of the land at some fixed point in the recent past; as 300 sq yards @ £20 per square yard; as 67% of the current site only rental value or a lower % of the total rental value of the whole building. It's the way that people respond to the £6,000 - and the fact that their income tax bills have gone down by £6,000 - that is important.

And as we know, any tax on land and buildings always falls most heavily on the land element and only affects the bricks and mortar (i.e. the decisions the owner makes as to the fabric of the building itself or what to do with it/in it) if it exceeds 100% of the site only rental value. If it is more than 100%, then it just acts like income tax on the earned element of the rental income, thus discouraging new construction in that area (which is what the NIMBYs and Greenies want) but not putting a complete kybosh on it.

5) Yes, LVT would work best if it could be shown that the effective % rate on each plot is exactly the same (in which regular revaluations will help), based on actual selling, rebuild and rental values etc, but with the best will in the world, no two valuers would arrive at exactly the same results or exactly the same tax bill for each of nearly thirty million separate plots in the UK.

Again, so what? At present, most taxes are collected from earned incomes. The actual tax rate payable by two individuals or two businesses working equally hard and adding a similar amount of value can be wildly different, depending on whether the business is VAT-able or not; whether it's a company or a partnership; whether the profits are paid out as salary, interest, rent, dividends or pension contributions; whether the individuals are counted as employees or self-employed; how much the companies or the individuals earn; whether the individuals are in receipt of means-tested benefits etc.

As a matter of fact, effective tax rates on income are anywhere between negative and over one hundred per cent, with an average of about fifty per cent, but the economy still limps on. It would be a doddle to set up a system where the effective LVT rate is somewhere between 60% and 100% of all site only rental values, there'll be winners and losers from this, but there'll be winners and losers from shifting taxes from earned income to the rental value of land anyway, so this is a minor issue.