1. What is Georgism?

Georgists, who take their name from Henry George, a nineteenth century physiocrat, believe that the ownership or use of land should be taxed. Most of them reject all other forms of taxation, so they are also known as Single Taxers. They do not believe that buildings or other improvements added by the landholder should be taxed — only the "unimproved" land.

There are many versions of Georgism, ranging from rhetorical formulations barely distinguishable from communism, to voluntary market-based arrangements sometimes described as geo-libertarian. It is difficult to get a clear picture of Georgist doctrine, because its apologists tend to slide illegitimately from one version to another logically incompatible one, according to the exigencies of the argument. We may charitably assume that this is because they often lack a sufficiently precise or coherent understanding of their own beliefs and proposals. Readers who seek to debate with Georgists should beware of this propensity (a very human failing which Georgists are by no means alone in falling prey to).

I shall now attempt to elucidate the various forms of Georgism:

Type I (Natural Georgist):
Type Ia: All land is initially owned in common by all mankind and landholders must pay its natural value into a common pool.
Type Ib: All land is initially owned in common by all mankind and landholders must pay its site value into a common pool.
Type Ic: All land is initially owned in common by the community and landholders must pay its natural value into a common pool.
Type Id: All land is initially owned in common by the community and landholders must pay its site value into a common pool.

Type II (Distributive Georgist):
Type IIa: All land is and must ever be owned in common by all mankind and landholders must pay its natural rental into a common pool.
Type IIb: All land is and must ever be owned in common by all mankind and landholders must pay its site rental into a common pool.
Type IIc: All land is and must ever be owned in common by the community and landholders must pay its natural rental into a common pool.
Type IId: All land is and must ever be owned in common by the community and landholders must pay its site rental into a common pool.

Type III (Market Georgist):
Type IIIa: Land may be owned in common by all mankind and landholders must then pay its natural rental or natural value into a common pool.
Type IIIb: Land may be owned in common by all mankind and landholders must then pay its site rental or site value into a common pool.
Type IIIc: Land may be owned in common by the community and landholders must then pay its natural rental or natural value into a common pool.
Type IIId: Land may be owned in common by the community and landholders must then pay its site rental or site value into a common pool.

Type IV (Geo-libertarian):
Type IVa: Externalities in land use are unjust and must be internalised.
Type IVb: Externalities in land use are economically inefficient and must be internalised.
Type IVc: Externalities in land use are economically inefficient and should be internalised where practicable.

Type V (Physiocratic modifier — apply to any of the above types):
Type Va: Real estate (fixed land or location) is no different from other types of land (natural resources), so the "single" tax must apply to all forms of land.
Type Vb: Real estate (fixed land or location) is different from all other types of land (natural resources), so the "single" tax need not apply to all forms of land.

Type VI (Pragmatic modifier — apply to any of the above types):
Type VIa: The full land tax or rental must be paid.
Type VIb: Less than the full land tax or rental may be paid.

The above list should cover most of the variants you are likely to come across, though I would not claim that it is absolutely exhaustive (for example, Georgists almost invariably hold the egalitarian position of common ownership, but market Georgists could also admit of joint ownership with unequal shares).

Most traditional Georgists are essentially of Type II (Vb, VIa). However, many modern geoists, as they often prefer to call themselves, are doctrinally Type IVa (Vb, VIa) geo-libertarians, though their proposals have usually not caught up with their principles and remain cast in the Type II mould. Since Georgists often "pick and mix" among the variants, considerable caution in labelling individual Georgists and Georgist web-sites is warranted; most such mixtures must be characterised as illegitimate, though some overlap (between Types III and IV for example) may be acceptable.

I shall add one more type of "Georgism", derivable from the egalitarian principles and arguments underlying Types I and II.

Type VII (Communist "Georgism"):
Type VIIa: All land is and must ever be owned in common by all mankind and all men must have equal and unhindered access to all land.
Type VIIc: All land is and must ever be owned in common by the community and all members must have equal and unhindered access to all community land.

I place "Georgism" in quotes here because this is more communist than Georgist. All true Georgists must believe that landholders may legitimately exclude others from the held land, so long as the appropriate single tax is paid. This is an important point, because many Georgists or would-be Georgists have not understood this implication of their doctrine. Georgism per se does not guarantee access to land for every person; it merely requires landholders to pay for the privilege. In theory, if every current landholder refused to allow you onto his land you would have nowhere left to stand — whether under Georgism or conventional land ownership "in fee simple". In both cases your quandary is resolved if some of the land remains unheld or as commons where all have an equal right to be — as is actually the case at present. Argument from an assumed "right to stand" can thus legitimately lead only to a demand for at least some land with free public access, not to the Georgist single tax.

2. Georgist Principles

It is not easy to find any coherent explanation of the ethical principles supposedly underlying the various forms of Georgism. I here attempt to elucidate them.

P1 (Common ownership):
P1a All land is initially owned in common by all mankind because Nature says so.
P1b All land is initially owned in common by all mankind because God says so.
P1c All land must always be owned in common by all mankind because Nature says so.
P1d All land must always be owned in common by all mankind because God says so.

P2 (Community ownership):
P2a All land is initially owned in common by the community because Nature says so.
P2b All land is initially owned in common by the community because God says so.
P2c All land must always be owned in common by the community because Nature says so.
P2d All land must always be owned in common by the community because God says so.

P3 (Justice in ownership):
P3a All must for the sake of justice receive the full value of that which they own.
P3b All should for the sake of efficiency receive the full value of that which they own.

P4 (Justice in creation):
P4a All must for the sake of justice receive the full value of that which they create.
P4b All should for the sake of efficiency receive the full value of that which they create.

The "Nature says so" tags stand for a range of possible arguments based upon the fundamental nature of man and the physical world or a priori logic. The only principled argument in this class that I have seen spelled out by Georgists in any detail is the "right to stand" argument: human beings need land for their very existence, even more than they need food and drink; without a place to stand (a physical location for one's body) no one could exist even momentarily, let alone survive; no one has the right to deny your right to life by excluding you from land; thus land is intrinsically owned equally by all. This conclusion, however, is a non sequitur, since it is perfectly possible to exist, and even to survive indefinitely, on land that is owned by someone else. Nor has any basis been adduced for the supposed positive right to life, which if accepted would logically force us to apply similar arguments to all resources necessary for our survival, even those made available through the labour of others. This is the communist doctrine of "to each according to his need" — not where Georgism wants to go. At most, considerations of this sort might plausibly suggest a moral obligation to leave a modicum of land open to free public access, or otherwise ensure that no one is altogether lacking a refuge, a conclusion which I showed in the previous section is insufficient to generate Georgism.

Another kind of "Nature says so" argument contends that common ownership, and the Georgist single tax, is more efficient than each of the alternatives — more conducive to the common good. Although most Georgists do appear to hold this opinion, they do so without explicit admission that this is their primary justification; indeed, their emotional rejection of all other forms of land ownership as unjust or offensive seems inconsistent with a merely pragmatic belief in the supposed empirical superiority of the single tax.

The "God says so" tags stand for the possibility that God, the original creator and owner of the world, deliberately placed the world into the common ownership of mankind (or regions of the world into the common ownership of the inhabitants of those regions). There seems no reason to doubt that God could have chosen so to do — and in that event we should presumably be under a moral obligation to respect His wishes. This would lead to either Type I or Type II Georgism, dependent upon the conditions God placed upon His gift. However, there seems no evidence, either scriptural or teleological, that God actually did intend such a general system of common ownership; if anything the boot is on the other foot, with both scripture and the natural world providing many instances of private ownership. Although some of the early Georgists held views of this sort, modern Georgists seldom rely on "God says so" arguments. In passing, observe that "the creator says so" could also apply quite legitimately to super-intelligent aliens or "gods" who create a planet and cause it to be inhabited, or to human engineers who construct a space colony and specify the terms upon which immigration shall be allowed.

Type Ic & Id and Type IIc & IId Georgism seem to require respectively P2a or P2b and P2c or P2b. However, Georgists have to the best of my knowledge presented no convincing rationale for restricting common ownership to the local community instead of to all mankind, beyond the practical consideration that it might be politically less contentious or more readily achievable.

Type I Georgism goes with P1a, P1b, P2a and P2b, whereas Type II Georgism requires P1c, P1d, P2c or P2d. However, it is not clear why, if it is legitimate for landholders to compensate the common owners by means of rental payments, it is not equally legitimate to compensate them by means of a once-and-for-all payment equal to the capitalised value of the rental, thereby purchasing the land outright. A rationale for rejecting Type I Georgism (more convincing than the "right to stand" argument) is lacking (though most Georgists do reject it nonetheless).

Type III Georgism seems to be based (at least in part) on principles like P3, wherein not only private owners but also common owners are entitled to reap the benefits of their ownership — and in which it is assumed that common ownership will at least sometimes be the most efficient form of ownership. Libertarians will for the most part find market Georgism ethically uncontentious, though they may doubt its practicability and consider P3 unnecessarily strict.

Type IV Georgism is based in part on principles like P4, which are already familiar to economists as the rule that for maximum economic efficiency all factors should receive their marginal products, though the P4a insistence that this is a point of justice, rather than merely a generally good thing, is perhaps novel and possibly excessive. These principles, unlike the egalitarian and rather ad hoc P1 and P2, are solidly grounded in standard economic theory, and good utilitarian justifications for them can be put forward. Consequently I find the geo-libertarian forms of Georgism the most interesting and worthy of study.

Most of the justifications for Georgism boil down to the contention that things would be better under Georgism. In other words, they are not so much moral justifications as pragmatic or utilitarian ones. I have no complaint with this, so long as Georgists are open about it and avoid taking a holier-than-thou attitude to those who disagree with them. Unfortunately, although — not unlike the supporters of every other political programme in the world — Georgists of all flavours continually contend that following their prescriptions would do wonders, they seldom attempt to back up their sound bites with hard analysis or define the supposed benefits of the single tax objectively. I would like to be able to give references to coherent economic analyses of Georgist policies by Georgist authors. I regret that I have not been able to find any; instead I have had to piece together a semblance of Georgist theory from various usenet, email and forum discussions, and from miscellaneous snippets, faqs and ideas on their web-sites, and then to apply my own analyses to their often vague proposals. Of the web-sites, the least unsatisfactory seems to be the English site, which contains the only really definite proposal I've seen:
http://www.landvaluetax.org
http://www.landvaluetax.org/1939bill.htm
http://www.landvaluetax.org/f39bill.htm

3. Definitions

To debate Georgism we need a clear understanding of the terms employed. Georgists themselves are often unclear as to their meaning, or slide illegitimately from one possible meaning to another, so that it is impossible to know exactly what is being said. Here are some definitions:

Georgism:
The belief that the ownership or use of land should be taxed.
Land:
1) All physical resources.
2) Real estate or location.
Natural Land:
1) Physical resources in the absence of any human activity.
2) Real estate in the absence of any human activity.
Affected Land:
1) Physical resources affected directly or indirectly by human activity.
2) Real estate affected directly or indirectly by human activity.
Improved Land:
1) Physical resources modified by human activity.
2) Real estate modified by on-site human activity.
Unimproved Land:
1) Physical resources unmodified by human activity.
2) Real estate unmodified by on-site human activity.
Improvement: Any aspect of human activity that changes the value of land.
Property: A piece of land and its on-site improvements.
Externality: Any result of human activity on one property that changes the value of other properties.
Property: A piece of land and its on-site improvements.
Natural Value: The value of land in its natural state unaffected by any human activity.
Unimproved Value: The value of land in its natural state unmodified by on-site activity.
Improved Value: The value of land inclusive of on-site improvements.
Site Value: The value of land net of on-site improvements.
Property Value: The value of a property inclusive of on- and off-site improvements.
Value: The capital value of property or land; the price one would pay for permanent ownership.
Rental: The current value of landholding; the price one would pay per unit time for an impermanent lease.

For the most part, when Georgists talk about land, they mean real estate or location, either because they hold a physiocratic position that "land" is different from all other natural resources, or because they recognise, perhaps subconsciously, that to apply a collectable rental to the physical substance of all raw materials, commodities, manufactured articles or agricultural produce would be absurd or unworkable, and would logically entail a tax on the human body itself. Because they lack a sufficiently clear rationale for why real estate is different, there is some ambiguity as to whether taxable land includes the soil and mineral wealth or merely the physical location. In what follows, unless otherwise indicated, definition (2), land equals real estate, should be presumed.

Any kind of human activity can affect land values — including simply passing through or being in the neighbourhood, as well as exploring, mapping, occupying, homesteading, farming, building, trading, speculating or developing. In particular, activities on one property will affect the value of other properties, either positively or negatively. In general, all activities will to some degree affect the values of all properties in the vicinity — and in principle "vicinity" means the entire region of mutual interaction, and thus the whole world. These externalities are what Georgism commonly attempts to capture and redistribute, sometimes in the mistaken view that these are identical to natural value.

Improvements can in general be either positive or negative (an example of a negative improvement is a derelict building it will cost money to demolish). Moreover, a given "improvement" may create both positive and negative externalities for other properties, varying in magnitude and sign from property to property (a rubbish dump is a negative externality for adjacent properties, but a positive one for properties further away who gain a place to dump their rubbish); this is the root of the NIMBY ("not in my back yard") syndrome.

It is important to distinguish between on-site improvements (the work or responsibility of the landholder) and off-site improvements (the work or responsibility of other people); failure to do so will lead to confusion between site value and natural value, or between property value and site value. It is unfortunate that Georgists have chosen to label bare land, that is, land without on-site improvements, as "unimproved land", since it can lead them to forget that the off-site improvements that increase its value are themselves the product of specific actions of particular individuals, and do not merely arise by some mystical and mysterious process out of society at large. Another point it is important to remember is that land taxes will in general cause market prices to fall below the full land values — because one is purchasing only partial not full ownership.

Finally, we need to distinguish between land values and land rentals. Value is the price one would be willing to pay for permanent ownership; rental is the price one would be willing to pay for an annual lease. An annual tax is thus a rental paid to the community. Georgists often talk about taxes on land values when they should be talking about rentals. Values and rentals are related by V = Integral from 0 to infinity{<r> exp(- Integral from 0 to t {<i>} dt) } dt, where <r> is the expected net rental and <i> the expected interest rate at time t. If the expected rates are constant (an unlikely eventuality) this reduces to V = <r>/<i>. Note that the market expectations may subsequently prove in error; hedging and speculation on the uncertainty in those expectations is therefore possible, and indeed, probable; and unless Georgism can eliminate those real uncertainties, speculation in land and property will remain.

4. What is the Natural Value of Land?

The natural value of land is identically zero.

This is so important I shall say it again. The natural value of land is zero. Even the most fertile soil or the most staggering mountain vista is utterly worthless if nobody has ever seen it and nobody knows it's there.

Remember, natural land is land that is unimproved by any human activity whatsoever. And once the value of land has been enhanced by human activities such as discovery and exploration, it is no longer natural land, but affected land. Nature has not created the added value; human beings have; and it is to human beings that the benefits should flow.

To state the same thing another way. Value is the market price human beings will pay for ownership of land. But by definition natural land cannot be brought to market, since bringing to market is a human activity. Thus natural land has no value.

What if one were to purchase rights in a newly-discovered but unexplored continent as a speculation? The ownership of the tallest or most spectacular waterfalls, for example, even though no waterfalls are yet known to exist there. Or the highest peak, though no land more than six inches above sea level has yet been seen. Is that not to place a value on natural land? In fact it is not, because that speculation is itself a human activity based upon knowledge derived from previous human activity — the efforts of geologists and geographers to understand what a new continent is likely to be like, the efforts of historians and economists and businessmen to understand how new discoveries and new continents affect markets, and so on and so on. The speculative value of new land has thus been created, not by nature, but by human beings.

Site value is not natural value, for the site value includes the externalities created by human activities on other land in the vicinity. Nor, in general, is the site value created equally or in common by all and therefore could not justifiably be held to be owned in common in the way that natural value (were it not non-existent) conceivably could.

5. What is the Site Value?

The site value is the value of land net of on-site improvements; in other words, the price of an empty lot. More precisely, the purchase price of an empty lot free of taxation. It is this site value (or the corresponding site rental) that Georgists normally propose to capture for the community by means of the single tax. The site value is that component of the value of a property which is created, not by the landholder or previous landholders, but by causes off-site, and thus, it is alleged, cannot justly be considered the landholder's property. It includes the natural value of the land (which however is identically zero), the contribution of the municipality, and the contribution of other landholders and residents. Since all human activity takes place within one or more properties (inclusive of properties owned by the municipality or held in common by and for the community) the site value of a property is simply the sum of the externalities directed to that property from all other properties.

It should thus be apparent why Georgists wish to capture site value for the community; it is a windfall to property owners resulting from the activities of other people. However, that windfall has not been created by the common activities of the community, but specifically by the other properties in the vicinity. Each property contributes externalities to the other properties as well as receiving externalities from them. Each property contributes to the site value of all other properties. On average, properties contribute just as much as they receive. Georgists ignore this side of the equation.

Let the contribution of property i to the value of property j be E_ij. Let the full value of property i be V_i and the site value S_i. Let its net windfall be W_i. Then:

V_i = Sum{E_ij, all j}
S_i = Sum{E_ij, all j not equal to i} = V_i - E_ii
W_i = S_i - Sum{E_ji, all j not equal to i} = Sum{(E_ij - E_ji), all j not equal to i} = Sum{(E_ij - E_ji), all j}
Sum{W_i, all i} = Sum{Sum{(E_ij - E_ji), all i} , all j} = 0

Although site value is a clear and comprehensible measure of externalities received, there is, so far as I can see, no comparable measure of externalities contributed, which perhaps is why they are apt to be forgotten.

Site values can be estimated by considering the market prices of nearby empty lots. We assume an urban environment, subdivided into many lots, a sufficient fraction of which are at any time empty, or due for demolition, and a sufficiently liquid market in those empty lots. In these circumstances it is possible to create a reasonably reliable map of site value, which we can expect to be low in the surrounding countryside and increase progressively towards the city centre. Discontinuous contributory factors, such as the extra value of corner lots, scenic views or site aspect, can be extracted from the statistics and used to adjust the estimates for affected properties. Such assessments will not be perfect, but are likely to prove sufficiently accurate for purposes of taxation.

Typically, site values will tend to contribute around one third of the total value of the property, buildings and other on-site improvements accounting for the remaining two thirds. Other things being equal, the more that has to be paid for the lot, the more expensive and valuable a building it makes sense to construct on it.

Rural site values are more problematic, because it is difficult to find such a thing as an empty lot, devoid of on-site improvements, or even identify the land's natural state. Any parcel of land that has been left wild is probably not characteristic of neighbouring land — or else it would have been similarly utilised. Rural property values are also strongly dependent upon natural features unique to each parcel of land — whether the ground is level or undulating, high or low, chalky or loamy, or slopes north or south, etc. — but the contribution of these natural features is not a natural value. It is a variation in the profitability of on-site and off-site improvements (the source of both improved and unimproved values), leading to divergent investments on-site and divergent impacts of externalities.

Much of the value of rural land is due to clearance, drainage, irrigation, hedges, fencing, farm and forest tracks, etc. — things that have little consequence for urban lots. Since no market comparable to the market in urban empty lots exists, it is hard to do much better than simply assume that rural site values are, say, one third of total values, as in urban areas. Even if we were to estimate the current cost of bringing land to its present standard from the natural state, we could not know whether these improvements would be appropriate to today's market; without an independent measure of site values, their present value can only be guessed; and without a market value for the improvements, the site value cannot be extracted. To the best of my knowledge, Georgists have failed to propose any non-arbitrary mechanism for separating improved and unimproved values in the rural situation.

6. What is the Single Tax?

The Georgist single tax is a tax upon the site value or site rental. Although Georgists of Types Ia, Ic, IIa, IIc, IIIa, IIIc nominally call for a tax upon natural value (which however is actually zero), in practice they wrongly conflate natural value and site value. Essentially all Georgists of all Types propose taxing site value.

For Type I Georgism, a one-off tax equal to the site value is levied upon the landholder. This is trickier than you might think. Our hypothetical Georgist municipality would have to calculate site values, as described above, before announcing its policies or allowing news of them to leak out. Once landholders suspected that they were about to be single-taxed, the market price of empty lots would fall, notionally to one half of the full price (at which price purchasers, after subsequent payment of the single tax, would just have paid out the full price of the lot). However, holders of more than one property might deliberately sell their empty lots at very low prices in order to minimise local site values (this is practicable because the single tax will fall on all their properties whereas the empty lot sales that define site values are relatively rare). Afterwards, once the one-off single tax had been collected, market prices would revert to full value.

Except in a violent revolutionary setting, thoroughgoing Type I Georgism seems politically infeasible. It is too blatantly unjust to property owners, most of whom bought their property at the full market price (paying the full site value as well as the value of the buildings), and have moreover paid taxes on it ever since. A limited Type I Georgism may however be feasible in the homesteading of previously unoccupied land, or the conversion of farmland to building land. However, we can assume that most Georgists would reject such a once-and-for-all tax, since it would leave property holders subsequently in full and undeniable ownership of their land, free to engage in speculation and enjoy whatever windfalls came their way.

Type II Georgism, that is, capture of the site rental, is easier to implement and less obviously unjust (though appearances are deceptive). Based on historic data, the annual single tax for each property is calculated to equal the site rental, which may be approximated by the site value multiplied by the current interest rate. If the full site rental is successfully captured by the single tax the market price of empty lots will fall to zero.

The tax authority must then adjust the tax rate annually to maintain close to a zero price, the corrections being typically equal to the current price (which could be negative) multiplied by the current interest rate. How can a market persist with zero or negative prices? Simple. Each landholder is liable for the single tax on his property. If he no longer wants the land he is not permitted merely to abandon it, but must transfer liability to another person; the price paid is the positive value of the land plus the negative value of the tax liability. If the expected liability exceeds the perceived value, the landholder will have to pay to get someone to take it off his hands. If the landholder dies without net assets or skips out the authority may need to take possession of the land and either use it itself or place it back on the market.

Is the authority then free to cheat landholders by raising tax assessments at will? Not really. Excessive assessments could be appealed to the courts; evidence of negative prices on empty lots would be hard to refute. Some limited room for manoeuvre and small-scale gerrymandering would remain, due to the statistical nature of the assessments and the inescapable presence of market uncertainties, but gross fraud would be impracticable (unless the courts themselves were unashamedly corrupt, which seems unlikely). All in all, we could expect Type II Georgism to be much less prone to political manipulation and excessive imposts than our existing systems of local taxation.

How much would the single tax amount to? On the order of 5% of GDP, similar in magnitude to present day local taxation, but a factor of ten less than central government expenditure, and much less significant than some Georgist rhetoric might lead one to believe. (A rough calculation follows: In the UK, residential site values are currently around £40K, yielding an imputed rental of £1600 per annum at an interest rate of 4% pa. Other uses, including commercial, industrial and public services, account for around 20% of total site value, equivalent to another £400 per household, to a total of £2000 pa. Another estimate starts with the value of suburban land, £100/m², at a population density of 100m² per person, to a total value of about 2 x £10,000 per person, allowing for the contributions of city centre land, up to £10,000/m², and rural land, down to £1/m². At 2.5 persons per household that's £50,000 or at 4% pa, £2000 pa, in agreement with the first estimate. With 20 million households, that's £40 billion a year, or about 5% of GDP. I would not expect results for other countries to differ greatly, though details of the calculation will of course vary.)

7. The Effect of the Single Tax

Given a single tax of the sort described above, what are its consequences? Georgists are disposed to argue that not only will it not distort the market, it will eliminate property speculation and increase the supply of affordable housing. They are seldom able to perceive the contradiction inherent in such claims. They also seldom specify any baseline for comparison. It should be obvious that if the single tax had no effect on the market it would be completely pointless. It should also be obvious that there is no a priori reason to suppose that investment decisions will be identical in a free market or under the single tax, nor that land use will not be radically altered. Structural transfers alter the dynamics of any economy; it is a matter for analysis whether they do so in a beneficial, catastrophic or relatively neutral way.

The initial effect of the single tax on the market is slight. Because the existing capital investment in property is effectively trapped in its present use and location, land use and gross rents will initially stay more-or-less the same. The net effect is a systematic transfer of wealth from landholders to the tax authority. However, we may assume that the authority now returns these revenues to the community, either by direct dividend or by the reduction of other forms of taxation. The members of the community, in their capacity as tenants (and even householders are in a sense tenants in their own properties), have more money to spend, so rents will rise as the landholders seek to pass on the tax. If the distribution of the tax is egalitarian in nature, as Georgists often assume, low rents will rise more than high rents; however, if the "single" tax merely replaces or reduces an existing tax, this will not occur, and all rents will tend to rise by approximately the same factor. Commercial rents will also tend to rise by a similar percentage, because the additional spending money will cause prices generally to rise, leaving firms with more money to pay higher rents. The whole rigmarole ends up with landholders passing on the new tax to the beneficiaries of the tax in the form of higher rents.

Georgists are apt to deny that landholders will be able to pass on the tax by increasing rents, or to attempt to make it so by legislative fiat, but in this they are inconsistent. One of their main arguments against the present system is that landlords and landowners are able to capture the benefits of any tax reductions or wage increases in the form of higher rents, without doing anything to deserve it, by exactly the above mechanism. And, in the short term, and with some provisos, they are correct. Where they go wrong is in forgetting, or denying, that in the longer term the supply of rentable properties is elastic (as is the demand, since the number of persons per household is not fixed and rent can be diverted to other goods); if rents rise it becomes profitable to build more houses or apartments, thereby stemming the rise. Supply and demand works with property just as it does with other goods. More slowly, perhaps, than the market for widgets, but still effectively.

So the single tax has raised rents, encouraging investment in new construction. However, since all benefits of location are taxed away from the developer, there is no longer any incentive to build preferentially in the most economically beneficial locations (that is, where the positive externalities are large). One might just as well build anywhere. In fact, it's a little worse than that; the expected benefits are taxed away, but the uncertainty remains, so building in higher-tax areas is riskier, for zero expected gain. Risk aversion suggests that one should then choose only the cheap locations. In a perfect market urban site values will fall sufficiently to compensate for this effect; but if the tax authority does not immediately reduce tax rates to match, average urban site market prices will stay persistently slightly negative, creating a slight net disincentive to build there. In principle, if the tax authority could restrain its natural short-term greed, and assess site values with total honesty, this disincentive could be eliminated; but in the real world political bodies are not so trustworthy.

A further point to note about the Georgist incentive structure is that site value, and hence the single tax, is a somewhat arbitrary function of the subdivision into separate properties. Let there be two properties, combined into a single property. Then, assuming positive externalities:

S_1&2 = S₁ + S₂ - E₁₂ - E₂₁ < S₁ + S₂

Thus combining plots has reduced the net site value (some externalities have been internalised) and should thus reduce the single tax. A similar effect can be obtained by moving activities from one property onto another. Thus Georgism provides an incentive to internalise externalities, favouring malls over single shops, apartment buildings over single dwellings, complexes over simplexes.

The form of land-use towards which the single tax pushes us is one in which the countryside is randomly dotted with perpendicular towers (tapering wastes land), 200 metre or so on a side, 2000 metre or so high, each tower a complete small town of 50,000 or so, inclusive of apartments, shops, offices, services and factories, but paying no more tax than a single suburban house. The internal economy of those towers will have some similarities to Type III market Georgism, but with the crucial distinction that no one who wishes to cease renting a unit when the current lease expires has any further liability; finding a new tenant is the responsibility of the tower owner. Most of the surrounding countryside will be abandoned or common land and thus effectively exempt from taxation.

That doesn't mean that existing towns would just disappear — there is too much already invested in them — though dereliction would tend to set in over much of the urban area over the course of a century or so as abandoned properties reverted to common ownership. This trend would be offset by the sharp reductions in site value and tax rates entailed, allowing continued economic occupation of neighbouring properties. An interesting effect arises where the single tax is imposed upon green properties (parks, gardens, playing fields and the like) within an area of high site values. Although such properties may create considerable positive externalities for the area, there is no way they can collect rent commensurate with the imposed tax. The landholders will thus find it necessary to maximise their rent base by building over the green properties, as densely as possible, ideally in the form of a single immense tower, and hope that they will thereby cause neighbourhood site values to fall markedly and reduce their own taxes. It is indeed a general rule under Georgism that landholders will wish to reduce the positive externalities their properties emit and increase the negative externalities as much as they can get away with, because by doing so they reduce their own exposure to the single tax.

We can thus see that over the long run a Georgist single tax would drastically change the patterns of land-use, by imposing an artificial and one-sided incentive structure. As described here, the results would not be economically disastrous, but could nevertheless be expected to be far from optimally efficient. From a Georgist's perspective many of the consequences would be perverse (a society in which most people are tenants of a few big landlords, whose property is hugely more valuable than the small parcels of land on which it is situated). It is therefore all too likely that various ad hoc regulations would be introduced in an muddled attempt to overcome the laws of economics and force the evolution of society down what may be perceived as politically more palatable paths (as in today's interventionist welfare states); it is hard to predict what forms this distorted Georgism might take, nor how much damage they might cause.

8. How might the Single Tax be Modified?

An unfortunate feature of the single tax as described above is that it imposes a permanent liability upon landholders. Land is a hot potato. An alternative approach might be to permit any landholder to abandon possession at any time. Legal possession would revert to the communal tax authority and collection of the single tax would cease. Persons wishing to make temporary use of the land would then not be deterred from doing so by the danger of being stuck with an unserviceable liability.

The difficulty with this approach is that there would no longer be room for a liquid market in empty lots at a price fluctuating around an average of zero. Negative prices would not occur (since the cost of abandonment is zero) so the price would have a floor at zero. But the tax authority and the courts need those below-zero prices to tell them when site value assessments are too high. It is true that abandonment is evidence of what would otherwise be a negative price and could perhaps be used as a proxy, but combining rates of abandonment with above-zero prices in a statistically sound fashion, in order to update the site value map, would not be easy, and would at best introduce further uncertainty and lack of transparency into the proceedings.

The tax authority could attempt to ameliorate this problem by entertaining a policy whereby abandoned land is always brought back onto the market (with the possible exception of land set aside for public use), by progressively reducing the assessed site value (and hence tax rate) until a willing holder is found.

The ability lawfully to abandon possession of land would tend to hasten the land-use changes discussed above, but the end result would be similar, except that instead of mainly abandoned and common land, there would be widespread private holding of low-value land for non-non-portable-capital-intensive activities (yes, that word does make sense — try starting at the end and working backwards!), that is, activities that don't tie you financially to that piece of land.

An alternative to the single tax is a site-rental auction. All the land is owned in common by the community; anyone who wishes to use it must rent it; and each lot is rented out to the highest bidder. Calculations of site values and the maintenance of a sufficiently liquid market in empty lots thus become redundant.

The problem here is that although buildings and other on-site improvements are supposed to remain purely private property, there is no easy way a landholder can remove his property without destroying it if he is outbid in the annual auction. A competing bidder might therefore risk paying considerably over the odds in the reasonable expectation of getting free use of the buildings until the following year, when the original owner may end up paying the market rental on the full property value, not merely the site value, in order to guarantee access to his own property. A sufficiently cheeky tenant could even demand rent from the property owner for allowing the property to remain on the tenant's rented land!

In effect, the community is expropriating almost the full value of these properties, including the value of on-site improvements by the erstwhile owners. This is clearly inconsistent with Georgist ideals, although it would presumably be possible to compensate the owners for this loss or buy them out prior to the implementation of the auction scheme.

In the longer term, an annual rent auction would make investment in permanent buildings economically infeasible. Portable, disassemblable or disposable buildings would predominate. Then the losing bidder could simply fold his tents like the Arab and silently steal away. Although mobile homes and offices are certainly possible, and meeting halls can be replaced by tents, I cannot see such a drastic shift in land-use as being other than economically damaging — perhaps disastrous. Utilities would be especially problematic under such a regime, and heavy industry would be in scarcely better shape.

The solution is to allow bids for more than a single year ahead. If you can outbid all other bidders for each of a sequence of future years, then you will have guaranteed control over the land for that full period (though you will have to pay the full amount of all your bids, that is, the full rent for the whole period, straight away). The longer the lease, the less evanescent your investment in the land need be in order to remain profitable.

The trouble with this, from the Georgist perspective, is the likelihood of speculative bidding on most lots out to hundreds or thousands of years in the future. The ownership of all or part of the lease would subsequently be linked to the sale or rent of the property; the special status of the land and its unimproved site value would then be lost. To all intents and purposes, existing non-Georgist arrangements, with all their supposedly objectionable non-Georgist features, would simply reappear. There would be undeserved windfalls, greedy landlords, idle rentiers, wanton speculation, and corporations ravishing the environment for filthy lucre. Georgists could not without inconsistency support an open-ended auction, whilst compromise schemes, such as auctioning ten-year leases instead of annual leases, would seem both morally arbitrary and economically unsatisfactory.

Finally, Type IV Georgists might wish to modify the single tax to include only the net windfall W_i, that is, the excess of externalities received over externalities contributed. Some properties, those that receive more than they contribute, would pay a net tax; other properties, those that contribute more than they receive, would get a net payoff. If this could be achieved, and assuming that the transaction costs were sufficiently low, there would be net economic benefit from this internalisation of the externalities of land use, and the Type IV Georgist programme would have merit.

Unfortunately, no procedures for calculating this net windfall have been presented. The gross windfall is of course the site value, but the externalities created by the property are spread in an unspecified fashion over all the other properties, and inextricably tangled up with all the externalities created by the other properties too. Beyond the obvious fact that the externalities created are on average equal to the externalities received, we have no basis for assigning a net windfall figure to specific properties. In extreme cases we might be able to make a stab at it, but for most properties our best guess is simply a net windfall of zero — and no net tax. Without a better procedure than this we cannot proceed with Type IV Georgism.

9. Internalisation by Indifference Votes

As noted above, there is a genuine economic case for internalising property externalities, though not a very strong one; the total externality is a small proportion of GDP (around 5%); the total net externality is even smaller (say 2%); and the potential economic gain is only a second order function of the net externality (and could hardly exceed about 1% of GDP). The total externality is small precisely because the market has already found ways of internalising most positive externalities; there is a strong incentive for market players not to give away more than they have to. Positive externalities are self-limiting. Negative externalities, by contrast, have to be restrained by law, because in the absence of restraint there is a strong incentive for market players to off-load their costs onto others and maximise their profits at the others' expense. There are actually two sorts of negative externality; those that violate rights and those that do not. The former we call torts — and penalise. The latter are competitively self-limiting (if I open a business in competition with an existing firm I reduce the value of their business — impose a negative externality — but by the same act create a reciprocal negative externality directed from them to me — so in general it is in my interest to avoid competition rather than court it).

This is not quite the whole story. There are many public goods and public bads with strong externalities that are currently managed through political decision-making, many of them involving land-use. Shall we build this motorway or not? A motorway is not something that some people can buy and others not, like a widget; it's an either-or decision in which members of the public have a strong, and legitimate, interest. The market — at least as it is normally understood — is not very good at handling this sort of decision, because of the large number of interested people and the strong positive and negative externalities. Unfortunately, political bodies aren't very good at handling them either, being too easily swayed by special pleading, bureaucratic interests, emotive populism and political expediency; and even honest administrators generally lack the sort of distributed knowledge and price information the market provides. Furthermore, political decision-making is apt to create large and unjust net externalities for the losing side, even when the issues are finely balanced.

A generic solution to such problems is the indifference vote. Let's go back to the motorway question. The project's proposers (a motorway construction company, perhaps) state how much it will cost (how much they are willing to build it for). Interested parties then make bids either for or against the proposal. If the sum of the bids in favour minus the sum of the bids against exceeds the cost, then the motion is carried. The winners pay off the losers in the amounts of their respective bids; the net bid left over after compensating the losers pays for the project. If the bidding takes the form of an auction, parties will progressively increase their bids until they are indifferent whether they win or lose (winning means you get what you wanted, but have to pay for it; losing means you get paid for not getting what you wanted). Bidding ceases when all bidders on the losing side have reached their indifference points, and at least one bidder on the winning side is still short of his indifference point and can thus make the winning bid. There is no limit on who can enter the bidding or the number of bidders.

Thus anyone who believes the proposed motorway would either benefit or harm him can put his money where his mouth is and enter the bidding on the indifference vote. All expected externalities of the project are thus internalised, and the full range of available information utilised. This is not to claim that actual externalities will always match expectations; no procedure is capable of eliminating all uncertainties.

Indifference voting is clearly a viable option for decision-making concerning public works and public activities (such as closing off a public highway for a weekend of nude bicycle races), private construction previously subject to planning consent, the appearance of private houses and gardens and indeed, almost anything of interest to more than one person. Subtleties arise with multiple or repeated propositions, and when the rights baseline is not symmetrical — what if the motorway has previously been rejected, but another company comes up with a new proposal? However, there is reason to believe that a just form of indifference vote can be devised to cover all such eventualities, though as yet I do not have a completely general solution.

Georgists, or Type IV geo-libertarians at least, should certainly be able to support indifference voting for public works. However, if they are motivated less by the desire for economic efficiency than by envy, the fact that there will still be windfalls for the fortunate may be too much for them to accept, as may the fact that it is possible to "pervert" indifference voting into a form of gambling (bidding above your indifference point in a game of "chicken"), even though gambling does not prevent the decision-making procedure from functioning, merely introduces an inefficiency that the gamblers pay for (non-gamblers don't lose out, because they simply bid up to their indifference points, then stop).

10. Initial Appropriation

Debates with Georgists are apt to devolve into arguments about first appropriation of land. Let us admit up front that this is a contentious topic and that none of the standard treatments is entirely sound. There is a reason for this. Rights can be managed according to the principles of justice only once their boundaries have been adequately defined. A dispute as to where an ill-defined boundary ought to have been placed depends for its resolution either on the threat positions of the antagonists or on some rather arbitrary idea of fairness. Yet, by its very nature, unappropriated land lacks boundaries.

Ideally, one would like to be able to define all such boundaries, whether geographical or behavioural, in advance; but even if this were feasible, it would involve considerable costs that would probably not be economically justifiable at this early time; boundary-making is a form of capital investment, but the resources available for such investment are always scarce. That is why we do not attempt to create infinitely detailed contracts for every transaction we enter into (a contract is a map of behavioural boundaries). Inevitably, we often find ourselves having to draw new boundaries in the middle of a dispute, preferably in such as way as to avoid dashing the legitimate expectations of those claiming prior or present possession; and to avoid making anyone worse off than they previously were; and to promote future economic efficiency; and to approximate a market solution in the fair distribution of net benefits. In these muddled circumstances perfect justice will not always be achievable.

Methods of allocating title to previously unappropriated land should be viewed less as matter of principle and more as practical solutions to the economic problem of defining boundaries and bundles of rights. This is why there is so much talk of staking claims, setting markers, erecting fences, occupying and using the land, and excluding others. These solutions don't have to be flawless or universal — just good enough to work. More important than initial allocation is the subsequent protection of property under the rule of law, and the market freedom to transfer rights and redraw boundaries in the ongoing economic process of Pareto optimisation.

From a geo-libertarian perspective we might want to claim that he whose actions creates new economic land, or makes land valuable, should own that land outright, or, where there is more than a single wealth-creator, should own an equity in that land or receive monetary compensation in lieu. Homesteading is not a bad approximation to this principle, especially if there are markets on which explorers and pioneers, or their backers, can capture the just rewards of their contributions. As the land fills up, externalities come increasingly into play, but it is not a matter of absolute justice that all externalities must be internalised, any more than that all boundaries must be precisely specified. It's a question of trade-offs. Knowing that externalities are six of one and half a dozen of the other (typically we gain roughly as much as we lose), and that capturing them is expensive, we usually prefer to let most of them go. This is choice, not necessity: by investing extra effort and initially purchasing all the surrounding land — not merely that which the pioneers intend to use themselves — then selling on appropriate terms to latecomers, early settlers can if they wish internalise most of the benefits of clustering — but if they are greedy risk driving potential settlers elsewhere. More often than not, they decide it's not worth it.

It is not morally reprehensible either to allow externalities to float free or to speculate to capture them; it is simply a prudential choice. Regrettably, and inconsistently, Georgists are apt to condemn both options. They share with socialists the illusion that the present and future value of things can be known with certainty and laid down unerringly, even in the absence of a working market.

It would be a mistake to pretend that the appropriation of land has always been carried out without injustice. It would be a much more serious mistake to seek to overturn existing patterns of land ownership on account of those alleged or actual ancient wrongs. Both the wrongdoers and their victims are long since dead. Landholders today obtained their properties in good faith, on the basis of well-defined and legally recognised rights and prescriptions fundamental to our society; to reject those long-standing rights in favour of some Georgist alternative, without adequately compensating the owners for their loss, would be grossly unjust. That said, so long as all current right-holders are bought out honestly, with resources honestly obtained (not taxes), Georgism, based on common or communal ownership, need not be rejected as intrinsically inequitable, though we may still doubt whether it would be efficient or, in the long run, politically viable.

11. Concluding Remarks

I have attempted to categorise the varieties of Georgism commonly encountered under descriptive headings and tie them in to the principles and policies to which they relate. I have attempted to elucidate both their differences and their similarities, and to show how they all come together in the demand for the Georgist single tax. Although some of the categories define positions held by very few Georgists today, I felt it necessary for the sake of consistency and completeness to address the logic of those positions, as well as those more frequently held, especially in view of the fact that many Georgists habitually combine arguments from a wide range of Georgist persuasions, even those in which they do not actually believe. However, I have tried to avoid arguing simply from strawmen.

I have attempted to explain how a single tax could be assessed and to analyse its likely effects, both in the short term and the long term. I have considered modifications that might be introduced in order to obviate some of the more obvious difficulties. I have been obliged to sharpen up the Georgists' generally vague proposals in order to have sufficiently coherent policies to criticise; I believe I have been reasonably successful in this, and that Georgists ought to be able to accept my account as a fair portrayal of Georgist ideas — though since they will dislike my conclusions they probably won't! My conclusions are that the single tax would not be beneficial, but would lead to economically damaging changes in land use and loss of capital investment. The Georgists' goals would not be met.

Nevertheless, procedures in some respects similar to those of Georgism may have limited applicability. The indifference vote, whereby the expected externalities of public works and other activities may be compensated and rational decisions taken, is one example. Allocation of the radio spectrum by means of annual auctions, in effect allowing broadcasters to pay site rental for the use of a common property, is another. Payment of ground rent for the maintenance of manmade space habitats or terraformed planets is a third. It would be as much of a mistake to insist a priori that everything must be privately owned as to insist that everything must be owned in common. All such forms of ownership can coexist within a free market under the rule of law; and although the optimal mix can be expected to change with time, these developments should be achievable without further injustice, for the most part through voluntary transactions on the open market, or on the pseudo-markets of indifference voting or public auctions. Violence to existing rights is not required and should be rejected. Georgism itself seems at best morally arbitrary and sadly prone to the espousal of morally objectionable demands; in its theory it is incoherent and in its policies economically detrimental. I conclude therefore that it is an erroneous doctrine and should be dismissed.

© Paul Birch, 29th Aug. 2002.

Home