Monday, 30 December 2013

The 'shadow tax' on housing: drawing the incorrect conclusion from a small sub-set of correct facts.

Something which has been bugging us is the Neo-Classical concept of the 'regulatory tax' or 'shadow tax' on housing, which seems to say that the main, if not only, reason for high house prices is restrictive planning laws, see for example this from the LSE:

In the US and elsewhere, zoning policies and other land use regulations are now widespread. Christian Hilber and Frédéric Robert-Nicoud look at the reasons behind these policies, finding that, driven by lobbying from developers and property owners, places that are more developed tend to adopt tighter land use regulations.

With land regulations operating as a form of ‘shadow tax’, of over 50 per cent of housing value in some cities, land regulations may now have become too much of a barrier to development in urban areas.


Correct facts

They then list lots of interesting statistics showing the correlation between build/population density, location values, and how restrictive planning laws are. To their credit, they explain that more restrictive planning regulations seem to be caused by higher build densities.

Which is of course blindingly obvious.

Imagine a little commuter or farming village/hamlet out in the countryside. If somebody wants to build one additional home, he can plonk it anywhere he likes and there is no need for any restrictions (local NIMBYs will oppose it anyway, separate story). The additional one or two cars make no difference to how crowded roads are, he can built his own sceptic tank, compost his own kitchen waste in his garden, maybe obtain fresh water from a well etc.

But in a large conurbation, people have to look at the bigger picture - they have to decide which areas to retain as parks, where the roads and parking spaces will be, where to route the utilities and drainage, how to get rid of the resulting household waste and so on. And once some maximum density has been achieved, it requires a colossal step-change before the next level can be reached. In London, for example, this means Cross Rail or the new Thames super-sewer, which are multi-billion pound investments which are of course vehemently opposed by existing NIMBYs (who are actually collectively cutting off their noses to spite their faces, as we will see).

Incorrect conclusion

While they correctly identify location values as a 'shadow tax' (or privately collected tax*) Their incorrect conclusion is that in the absence of planning regulations, location values and the price of a single unit of housing in those areas would fall.

We can tell straight away that this is nonsense. If you go back in time long enough (a century or three), places where the great cities like Los Angeles, San Fransisco, New York, London etc now stand were very sparsely populated and location values were negligible. There was no discernible difference between the value of these locations and anywhere else inhabitable on the US or European land masses at that time…

The full facts

… and where has most of the new construction taken place? In those great cities. Where are location values highest? In those great cities. It is a multi-factor feedback loop:

If an area is ever so slightly preferable to another (natural infrastructure such as a harbour, coastline, river, flat dry land for buildings and roads etc) then more people move there. It only requires one little spark to ignite things.

More people => more specialisation, more efficient usage of natural or man-made infrastructure
More specialisation etc => higher wages, profits, trade
Higher wages etc => More people want to live or set up their business in the area
More people wanting to live and work there => higher location values (the amount which people are prepared to pay to live there)
Higher location values => more investment in buildings, higher population densities
Higher population densities => more man-made infrastructure (container ports, wharfs, mains water and sewerage, railways, motorways etc)

And so on and so forth. We could summarise these in a sort of flow chart, but each element feeds into and is fed by every other element, and there are plenty of other self-reinforcing elements I have not yet mentioned, so it would get very confusing. And at the centre of this whirlpool is of course the Land Monopoly Black Hole, that is where all the extra value disappears.

Something else they wilfully ignore

Further, location values within an urban area follow the same general pattern, i.e. they are proportional to population density/total population. The average value of land per acre in a large city is ten times as much as in a small town etc.

And the gradient within a large city is much steeper than in a small town. While the value of land at the very edge of a small town might not be much different to the value at the outer edge of a large city, the value increases as you head towards the centre.

So in a large city the value/acre in the very centre are ten times as high as in inner urban areas, which in turn are ten times as high as in outer-urban areas at the outer edge = a ratio of one-hundred-to-one between centre and outer edge. But in a small town, the value in the very centre is only ten times as high as at the outer edge. And so on.

You can observe this very easily by remembering that location values, build density and population density are three different aspects of the same thing (see feedback loop above).

So the usual supply/demand rule does not really apply to land.

If all European motor manufacturers decided (or were somehow forced) to produce and sell twenty per cent more cars every year, you would expect the price of new and thus second hand cars to go down. But they would have to pay more for their inputs: steel, rubber, car workers' wages etc, and they would probably end up all going bankrupt (which is why they produce the number of cars they do - that is their profit maximising level of production). Further, the price of a new car is decided as between customers and manufacturers. There is no competition between customers.

Conversely, relaxing planning laws means that the value of the most important input i.e. land will always go up to soak up the difference between what the highest bidding customer is prepared to pay and the build costs. The price you pay for a finished house in that location is fixed and decided by whichever individual customer bids the most. The land value cannot be competed away (or else, show me evidence that it can) and land owners will never go bankrupt, even if you abandoned all planning laws.

Loosening planning restrictions only has a measurable downward impact on land values at the existing outer margin but an upward impact on the 'new' area

Residential land on one side of 'the fence' (which demarcates the Hallowed Green belt) is worth £500,000 per acre; farmland on the other side is only worth £5,000 per acre. But if you shift 'the fence' a hundred yards out, the value of the existing residential land might fall slightly to £450,000/acre (no longer has direct view over the HGB) and the value of the new residential land rockets from £5,000/acre to £400,000/acre. The total location value of land in that conurbation will always be slightly higher afterwards than beforehand.

(The only counter-force here is that most people actually like having a back garden to themselves and to be near fields and forests. So half of new graduates in the UK move to London but not all of them. The best of both worlds is having a big plot with a beautiful view over nature within a few minutes of the amenities of a town centre).

Loosening planning restrictions has a measurable upward impact on land values in the centre

So what happens to the average rather than marginal value of land in a conurbation if we move 'the fence' a few hundred yards further out? More houses get built and people fill them up, so the values in the centre and inner-urban areas go up even more, because there is now a larger pool of customers, workers, entrepreneurs, more specialisation etc.

And more subtly, the convenience value of being at the centre, rather than having to slog your way in from the outer edge depends on the distance from the outer edge to the centre (in the same way as the value of a place in a queue is dictated by how many people are behind you more than by how many people are in front of you) is now greater, pushing up location values in the centre even further.

And what happens if we allow people in the inner-urban areas to build more densely (smaller gardens or higher buildings)? The value of that land goes up even more, obviously. The developer knows that the basic rental value of a residential unit is £10,000 a year, and people are only prepared to pay a small premium to have their own back garden (a luxury rather than a necessity). So if he can build fifty flats instead of twenty terraced houses or ten semi-detached houses, he'll go for fifty flats and make three or four times as much money as if he built ten detached houses.

The same goes for city centres, the most valuable bit is at pavement level, ideal for retail, pubs, restaurants etc, and above that is offices. While most people don't like living high-rise, people probably couldn't care less whether their office is in a five-storey or twenty-storey office block.

Here endeth

So while I have no sympathies with NIMBYs whatsoever, both they and the Neo-Classicals are completely missing the point. The point is that location values arise quite naturally from the way that people behave, and whatever you do with planning, location values are the naturally arising, minimum and irreducible level of "tax" in any organised society.

Your only decision is whether to allow this tax to be collected privately (for the benefit of a few individuals only) or to pool these values and spend it on stuff which benefits everybody (which of course includes cutting taxes on wages, output, profits etc).
--------------------------
* A publicly collected tax is when private individuals are forced to pay money to the government, which the government then distributes to other private individuals (including their own friends and family). A privately collected tax is when private individuals are forced to pay money to other private individuals without the official government actually stepping in.

4 comments:

Sanjay Mittal said...

Elevated land values in cities are caused (as you suggest) by what economists call “economies of agglomeration” (e.g. it benefits banks and stockbrokers to crowd into the City of London). But I don’t think that proves that planning restrictions have no effect at all.

The cost of land accounts for about a quarter of the cost of the average house according to a Policy Exchange work called “Making Housing Affordable”. So if planning restrictions were abolished, that would knock a fair bit off the cost of the average house.

Possibly the main effect of abolishing planning restrictions would occur in the countryside. But if so, that would draw people out of cities which would cut the price of houses in cities.

Mark Wadsworth said...

SM, yes. "agglomeration" is one of the many things which I didn't include in the feedback loop. It was long and complicated enough.

PE are talking shite anyway - the value of the land/location is on average at least half the value of the whole home in the UK.

As to the countryside, it has been the case that people have been moving to the cities to live and work ever since the economy became "industrial" rather than "agricultural".

Admittedly, a lot of people like living in commuter suburbs, giving the illusion of country living with the good wages from working in a town, but there would be no big drift to the countryside.

For example - look at the "tiger estates" in Ireland which are all in the middle of nowhere and standing empty, despite the houses are very cheap. And rents and so on in Dublin are largely unchanged.

Ben Jamin' said...
This comment has been removed by the author.
Ben Jamin' said...

@Sanjay

We've correlated the relation between property tax and affordability between 150 of the largest cities in the USA.

It comes out at 0.56.

Given the the US is a huge continent, there are plenty of other factors other than tax to take into consideration.

1) Household income
2) Density
3) Topographical supply constraints
4) Desirability.

Difficult to achieve over the whole of the US, but if we narrow it down to one state, we get a tax/affordability correlation of 0.9

Basically, what we can say with utter confidence, is that given all the factors that go to make up "House Prices", planning has very little if f**k all to do with it.

Policy Exchange are just spinning propaganda.