28 October 2008

Land Value Tax - What is Value?

It's time for me to hold my hands up. When I've talked about different methods of valuing land for the purposes of LVT, whether it be self-assessment or taking house sale prices and deducting rebuild costs, I've made an assumption which, on reflection is not reliable. That assumption is that the sale price of land directly correlates to the rental price of land.

Both sale price and rental price are used within the current tax system. For instance, Council Tax is based on the assessed sale value of a house, whereas National Non-Domestic Rates are based on the assessed rental value of a commercial building.

The difference is important for LVT, because the charge is intended to be in proportion to the benefit the landholder is getting from having exclusive rights over a location at a given time, which, by definition, is its rental value. If the sale value does not directly correlate to the rental value then it is, a compromise valuation. I believe that there are examples where it could diverge significantly:

Consider two identical fields, A and B, used for agricultural purposes. I would expect them to attract the same rental price. Now imagine that field A is in an area where it is suspected that the local authority might give planning permission for a housing development, but field B isn't. Field A would have some speculative value, which would probably result in it having a higher sale price than field B, in spite of the fact that they both have the same rental value.

The very nature of sale prices means that they are also much more volatile than rental prices. In effect, the sale price of land is the predicted future rental values rolled up into a single payment. If there is an expectation that future rents will increase, selling prices can rise quickly on the basis of that speculation, while the current rental value remains relatively stable.

That isn't to say that rental price can't be derived from sale price, or that sale prices aren't an adequate proxy in the majority of cases, but I think it is important to acknowledge that they are not the same thing and allow for that when assessing any system of valuation.

I think this shows that one of the biggest problems with LVT is the name. "Land" can be misleading, as it can make people think of "soil" rather than "location" and "value" is ambiguous, as it isn't clear whether the value in question is the purchase price or the rental price. Maybe a term like "site rental value tax" would have been clearer. It would have certainly forced my thinking in the right direction.

10 comments:

Mark Wadsworth said...

Paul, most hard-core LVTers say that it should be based on rental values. To my mind this is stupid:

1. Establishing capital land values is easy, it's total value minus rebuild costs, anybody can guesstimate it for themselves. Rental values of a site is a fairly alien concept and would require lots of 'expert valuers' and too many arguments and make the whole thing highly suspect.

2. Of course capital values and rental values get out of line. That's the whole point of using capital values! If capital values rise too fast, then we have a bubble. A tax on capital values would dampen off the bubble very quickly. A tax on rental values would not dampen a bubble - for example Business Rates are a tax on the current rental value of buildings - they have not prevented a bubble/bust in capital values!!

3. So if capital value is too high, LVT would bring it down, so that LVT is in line with rental value. If capital value is too low, the temporarily depressed LVT bill will make the site attractive to buy. LVT on capital values dampens swings in both directions.

In the light of the house price and credit bubbles, this stabilising influence is surely the Number One Selling Feature of LVT! Don't spoil it now! (and yes I agree that low and stable is better than high and stable, different topic).

4. As to Field A/Field B, again, that's the point of using capital values. If it is announced that a new railway station is to be built, then the capital value and hence LVT from Site A (near the station) will increase immediately, years in advance of the station opening. That generates some of the income to pay for the infrastructure! And yes, it is possible that Field A never gets planning -so what? People gamble on this sort of thing all the time.

5. As to the name, strictly speaking, it's not land value tax at all. Land itself has no intrinsic value, its value derives from (generosity of planning permission x local amenities + scarcity value + bubble element). But that's a bit of a mouthful. I think 'Site Value Rating' is a good alternative.

I have been meaning to do a proper post on this, but those are my thoughts.

Paul Lockett said...

Mark,

1. I agree that in most cases, capital value is more familiar than rental value and easier to arrive at, which, in my opinion, is the main argument in favour of using it. I wouldn't say it's a bad approach, but compared to using rental value, I'd view it as a trade-off.

2. I disagree about a tax on rental values not damping a bubble. Capital values bubble because of an expectation of increased rental values in the
future, as capital values are derived from the present value of future rental values. If people know that a large proportion of those rental values will be taxed, it will damp down the capital value, because they know that any increased future rental values will be accompanied by increased future tax bills, which reduces the gains to be made. At the extreme, consider LVT set at 100% of the rental value; it would reduce selling prices to zero and eliminate speculative bubbles completely.

In essence, saying "we're going to charge away 50% of the total capital value by charging x% of the capital valuation" or "we're going to charge 50% of the total rental value" will achieve much the same result, bubble wise. Both will reduce the selling price to around 50% of the underlying capital value, (ignoring the effect of bringing unused sites back into use). The difference is that with the former, the charge will fluctuate in line with perceived future movements in rental values, whereas with the latter, the charge will track actual rental values, which will generally be much less volatile and therefore easier for the payer to predict.

Neither Council Tax nor Business Rates prevented the bubble, so there doesn't appear to be a guaranteed result from either approach. I think the problems in the current system come down to the charges being set at too low a level to be effective, the valuations being too infrequent and the charges not directly increasing in line with assessed values. I don't think the valuation methodology plays such a major part.

3. LVT on capital values does dampen swings in both directions, but so does LVT on rental values, as per point 2.

4. There are pros and cons to both approaches in this situation, but I view the rental value approach as being preferable, because the landholder won't pay anything more for being close to a railway station until the station is up and running and the benefit has materialised. If you were a landlord, for example, that would mean that the charge wouldn't increase until the rent available from your tenants had increased too.

5. I do prefer SVR as a term, as I think "site" is more easily understood than "land."

Mark Wadsworth said...

We'll have to agree to disagree (he said sadly).

Business Rates (despite what the purists say) is not that different to LVT. It is an annual charge set at about 30% or 40% of rental value of the total building. This did NOT dampen the speculative bubble in commercial buildings. A tax on current rental values does not dampen bubbles, there is no need to debate this.

In fact, rental values generally increase very closely in line with earnings. Long term property values do as well, but subject to 18 year oscillations. So with a tax on rental value (income or corporation tax on rents is a tax on rental values, as is Business Rates) would have much less impact on capital values than you think.

As to your point 4, so what? If you have a choice between buying House A (no new station to be built) and House B (in the vicinity of a likely new station), then House B is worth more.

People who buy House B are happy to pay more to the bank in interest in the meantime, so why wouldn't they be happy to pay rather less to the bank and a bit more in LVT (or SVR if we prefer that term)? That money can be used to pay for the station. And this obviates the need for capital gains tax - the extra tax is paid up front rather than after a subsequent disposal.

Further, from a purely political point of view, other property taxes like IHT, SDLT, CGT are calculated as a % of capital values, so it makes SVR easier to explain and understand. And it is easier to explain to people that every £1 LVT they pay is £1 off their mortgage repayments.

Paul Lockett said...

Mark,

There is no doubt that a tax on rental values does dampen speculative bubbles.

If it were set at 100% of rental value, there would be no capital value and therefore no bubble to occur. It is self-evident, otherwise people renting houses at market rate would be selling off their leases to others!

If LVT is set at less than 100% of rental value, some level of speculation will continue with the remaining capital value, but the same applies to a tax on capital values which leaves a residual value. While there is a residual capital value, there is something to speculate on.

That's why I don't quite understand your claim that Business Rates didn't dampen the speculative bubble in commercial property. How can you be sure that the bubble wouldn't have been even more severe without it?

With regard to point 4, surely it is preferable to make people contribute for the benefit they are gaining now, not pay up front for a gain they may or may not receive in the future.

I agree that a charge on capital values is easier to explain and to calculate and for that reason, it would be the more practical route to go down, but let's not kid ourselves that it is anything other than a compromise which is technically inferior to a charge on rental values.

Mark Wadsworth said...

How can you be sure that the bubble wouldn't have been even more severe without it?

Agreed. The bubble in commercial property was not quite as mad as in residential. But mad nonetheless.

I disagree that LVT on capital values is inferior. Capital values change much more quickly than rental values, therefore a tax on capital values would react much more quickly.

As to point 4, if I buy House B (in the absence of LVT), I am happy to pay the bank interest on the higher capital value (which in turn deries from the fact that a station will be built in future) even before the station is built.

Why is it so terrible for some of that money to go in taxes to fund the station? Why is it better that the uplift in capital value all goes to the previous owner?

Plus as a small government free market liberal type of chap, the idea of 'experts' deciding an arbitrary figure for the rental value gives me the creeps and will not go down well with the voters. As a result, rental values won't be updated and after five or ten years the resistance to an update is so huge that you end up stuck with old values.

Whereas a tax on actual up to date capital values responds to free markets and there would be less suspicion that the results are biased.

Paul Lockett said...

I disagree that LVT on capital values is inferior. Capital values change much more quickly than rental values, therefore a tax on capital values would react much more quickly.

I think this could be both a blessing and a curse. If you tax away, say, 50% of the capital value of the land, it doesn't really matter whether you do it by charging 50% of the rental value or charging a percentage of the capital value, it will suppress a bubble in much the same way.

The problem with using the capital values is the volatility, caused by the fact that capital values oscillate around the generally steadier movements in rental values, which makes a capital value based tax less predictable, both as a source of revenue and as a charge. I'd say that if two charges achieve the same result, the less volatile one would be preferable, all other things being equal.

As to point 4, if I buy House B (in the absence of LVT), I am happy to pay the bank interest on the higher capital value (which in turn deries from the fact that a station will be built in future) even before the station is built.

The flip side is that, if I were renting, I generally wouldn't pay a higher rate until the station is up and running.

Why is it so terrible for some of that money to go in taxes to fund the station? Why is it better that the uplift in capital value all goes to the previous owner?

If the tax were set to capture 50% of the capital value, as above, there wouldn't be any difference in the capital uplift available to the previous owner. The difference would be that the tax with the capital valuation would go up speculatively, whereas the rental valuation would only go up once the station was up and running.

I don't think the use of the capital value is terrible, but it does have some disadvantages. Using rental value restricts the valuation to the current permitted use, whereas the capital valuation can increase in line with speculation that planning permission may change. That could have a negative impact on, say, a farmer who is faced with an increased bill in response to speculation that his land may be given planning permission to be used for housing, even though the rental value with it's present permitted use has not changed.

I'm aware that you would consider exempting farm land from LVT, but I'm sure there are other non-agricultural examples.

Obviously, the situation would be different without a planning framework, but I don't view that as a particularly realistic prospect any time soon.

Plus as a small government free market liberal type of chap, the idea of 'experts' deciding an arbitrary figure for the rental value gives me the creeps and will not go down well with the voters. As a result, rental values won't be updated and after five or ten years the resistance to an update is so huge that you end up stuck with old values.

I tend to agree that the capital value has the advantage of being more transparent (the exception might be agricultral land, where rental value is used frequently), although I don't see any reason that rental value couldn't be updated frequently.

Whereas a tax on actual up to date capital values responds to free markets and there would be less suspicion that the results are biased.

I agree, although I think the self-assessment methodology I outlined previously would be even less susceptible to suspicion and would be superior in that respect, as there would be no assessors involved at all, just individuals creating their own market valuation.

As far as I can see, there are three basic methods for administering LVT:

The valuer assessed rental value, which is economically superior but less transparent with regards to valuation.

The self assessed capital value, which has the most transparent valuation methodology, but is economically not as well directed.

The valuer assessed capital value, which is a compromise sat somewhere between the first two in terms of both transparency of valuation and economic effects.

Personally, I just want a move to LVT, so I'd be happy to have any of the three up and running, but I think it's worth acknowledging that each approach has its strengths and weaknesses.

I think this thread has raised enough issues that it warrants me doing a complete post at some point detailing the pros and cons of each of the three methods.

Mark Wadsworth said...

The problem with using the capital values is the volatility

Agreed. So use a weighted average of the last three years' capital values.

I'm aware that you would consider exempting farm land from LVT

Correct. But it's not so much that I'd exempt 'agricultural land' per se, it's because ag land is worth 1% as much as residential land, it's not worth the hassle taxing it (or nature reserves etc).

As far as I can see, there are three basic methods

You missed off my favourite - being actual capital values based on actual arms' lenght selling prices in each postcode sector (or whatever smaller unit you choose)

Paul Lockett said...

You missed off my favourite - being actual capital values based on actual arms' lenght selling prices in each postcode sector (or whatever smaller unit you choose)

That was what I meant by "valuer assessed capital value," in that the valuation would be done through a central assessment, rather than by the landholder.

I would envisage that would be done on the basis of sale prices in an area, less a valuation of the buildings (such as rebuild costs) involved and an assumption that land value is directly proportional to square footage and is the same per square foot for each plot in the locality.

The one thing I'm not sure how to deal with under that approach is more unusual sites with a less liquid market. How would you value an airport, or Alton Towers, for instance? I think that might be where the self-assessment approach would be best suited.

Mark Wadsworth said...

PL, now we are getting somewhere. You have summed up what I have been saying all long.

Airports is easy - the location value is not as important as user charges on use of air-space.

As to Alton Towers, tricky one, but somehow they manage to agree a figure for Business Rates, so it can't be impossible. The key to these out of town theme parks is that they buy the original land for very little, being ag land, and then need planning permission. It's the planning that makes up 99% of the value.

Paul Lockett said...

Airports is easy - the location value is not as important as user charges on use of air-space.

I agree that the landing/take-off slots are the major value, but even if they were auctioned, there would still be significant value in the location as a whole, especially from car parks.

As to Alton Towers, tricky one, but somehow they manage to agree a figure for Business Rates, so it can't be impossible.

I believe that very few of the buildings are rateable on these sites and for those that are, such as shops, the annual rental value is probably relatively easy to work out. Calculating the value of the whole bare site is probably a good deal more complicated.

The key to these out of town theme parks is that they buy the original land for very little, being ag land, and then need planning permission. It's the planning that makes up 99% of the value.

Agreed. The real difficult is working out how much value that planning permission adds, given that theme parks are few in number and rarely change hands, resulting in little data to work from.

For that reason, while the centrally valued approach is probably preferable for residential property where people would be less keen on constantly self-assessing, this type of unusual site is probably better suited to self-assessment, given the lack of outside data to work from.