|Saturday, December 24, 2016||(Comment)|
The Georgist case for a vacancy tax
If you insist that a pure Georgist system should be implemented from 1st July next year, this essay is of no interest to you. If, on the contrary, you think it necessary to tolerate a transitional period during which Georgist revenue instruments will coexist with non-Georgist ones, this essay is of the highest relevance, because it concerns the kind of Georgist instrument that can best coexist.
Under a pure Georgist system, private entities contribute to the public purse according to the value of land that they own,* regardless of whether they use the land or waste it. By that standard, the problem with the present system is not any lack of taxation on owners who use their land. Transaction taxes, including income tax, payroll tax, company tax, GST, customs and excise, etc., are all taxes on various uses of land; and wherever such taxes are nominally paid by tenants, they are shifted onto landlords by reducing the annual rents that tenants can pay.* Rather, the problem with the present system is a lack of taxation on owners who waste their land.* Such owners avoid transaction taxes by avoiding transactions. They are unaffected by the shifting of taxes into annual rents, because they are not trying to collect annual rents. While the capitalized prices of their land holdings are depressed by transaction taxes, this applies to both acquisition prices and resale prices; and the resulting “capital gains” are taxed at concessional rates.
From a Georgist point of view, a vacancy tax equalizes the tax burden by directly targeting the gap in the present tax net, and only that gap.
A tax on someone else
If the land-wasters pay the vacancy tax, they reduce the amount of revenue required from the rest of us. If, on the contrary, they avoid the vacancy tax by attracting tenants (building or renovating as necessary), they generate taxable activity, expanding the bases of other taxes and allowing their rates to be reduced. Either way, the rest of us pay less.
Moreover, a vacancy tax, by forcing unoccupied properties onto the rental market and reducing rents, makes it easier for employers to pay for business premises, easier for employers to pay workers enough to pay for housing within commuting distance of the premises, and easier for jobseekers to move to where jobs are available. These are multiplier effects. They are also disinflationary and are therefore further multiplied, not offset, by their influence on monetary policy. Through the multipliers, a vacancy tax yields a bigger increase in economic activity, hence a bigger tax cut for the rest of us, than meets the eye.
But as soon as you propose to extend the vacancy tax to all land, and to compensate for it by reducing or eliminating other taxes, it becomes possible to question the adequacy of the compensation. So the land-wasters, together with anyone else who expects to be a loser, will hire “independent” consultants to cook up sets of assumptions and non-sequiturs whereby every man and his widowed grandmother will end up paying more. And their campaign budget will be bigger than yours. And what they say will get more “news” coverage than what you say. And that's if your policy is so clever that the only actual losers are the land-wasters.
Today's politics is about Us and Them. The only popular tax is a tax on Them. The prospective payers of a vacancy tax can be categorized as Them. They are not home owners or family businesses or farmers, or even “mum and dad” property investors who are deemed to “provide” rental housing by turning owner-occupants into renters. They are owners of eyesores — vacant suburban lots, derelict buildings, and downtown “bombsites” — or of high-rise apartments that have never been lived in; or they are “developers” of urban-fringe estates carved into lots of ever-decreasing size and drip-fed to the market so as to keep prices high. They are not Us.
A general tax cut funded by a vacancy tax and its flow-on effects is a tax cut for Us funded by a tax on Them. Politically, that's a near-perfect formula. A further tax cut funded by extending the vacancy tax to all land is a tax cut for Us funded by a tax on Us. Politically, that's a near-perfect target.
A broad-based land tax, precisely because of its breadth, is difficult to apply at a rate high enough to address the inequality between taxation of land-use and taxation of land-waste. A vacancy tax, precisely because of its narrow base, can easily be applied at a high rate — perhaps high enough to take away that inequality in one step.
Naming the tax base
If a vacancy tax is to be an incentive to build, and not an incentive to demolish, it must apply to all unoccupied land, whether there is a building on it or not; if there isn't, the owners must build in order to attract tenants and avoid the tax.
The name “vacancy tax” does not quite capture that requirement, because the word “vacant” has connotations of availability, whereas the tax is mainly intended to induce owners of unavailable land to make it available; the land that is already genuinely available is not the problem, although the tax still helps by increasing the incentive to get tenants.
The name “vacant land tax” is a bit better, because the word “vacant” seems to lose the connotation of availability when applied to bare land, which we obviously need to include in the net. But unfortunately the words “vacant land” might be construed as applying only to bare land, whereas we also need to include unoccupied built-up land in the net. Meanwhile the word “land” by itself probably does no harm, because voters are accustomed to the idea that “land tax” can apply to built-up land.
The name “unoccupied land tax” seems to avoid all these problems. And by adding the word “unoccupied” to a term whose meaning is already well known, it informs “mum and dad” property investors that they won't pay as long as they keep their properties tenanted.
Neutralizing the valuation base
Of course an unoccupied land tax is more likely to be an incentive to build, and less likely to be an incentive to demolish, if it is levied on the site value, and not the capital-improved value (the combined value of the site and buildings). But that feature is far less important in a vacancy tax, which is designed to be avoided, than in a conventional property tax, which is designed to be paid. Under a vacancy tax levied on the capital-improved value, demolishing a usable building may reduce the tax bill somewhat; but it also forfeits the opportunity to avoid the tax immediately and completely by getting tenants — an opportunity that doesn't exist under a conventional property tax.
It is well known that a holding charge on land cannot force the land out of use, provided that it does not take more than the rental value of the land (because a private tenant will pay that much in order to use the land). But a vacancy tax, being solely a tax on non-use, cannot force the land out of use even if it does take more than the rental value; on the contrary, the heavier the tax, the greater the incentive to keep the land in use in order to avoid the tax.
These considerations reduce the importance of the accuracy of valuations and the need for an existing system of valuation, and make it more feasible to use rough-and-ready methods, such as taxing the land on the owner's valuation provided that the owner's valuation constitutes an offer to sell the land for (e.g.) that value plus 10%.
A vacancy tax also neutralizes a more technical issue. By making it uneconomic to hold land vacant for speculative purposes, a vacancy tax must cause some reduction in the sale prices of land. But because the tax is only a contingent and avoidable liability, that reduction is not simply a capitalization of the tax. Hence, if the tax rate is (e.g.) 6%/year and the rental yield of land under that system is also 6%/year, then the tax, if it is actually applied, takes 100% of the rental value; there is no need to account for any further reduction in the taxable value caused by capitalization of the tax. In that sense, a vacancy tax is conceptually simpler than an all-in land-value tax, for which no rate on the current capitalized site value would capture 100% of the rental value (because that would reduce the taxable base to zero).
Neutralizing the widow (and other usual suspects)
A vacancy tax cannot force retirees or other low-income home owners out of their homes: if there's anybody home, the tax isn't payable! Similarly, a vacancy tax cannot hurt tenants or ordinary home owners or family businesses or farmers or religious/charitable/educational/philanthropic users or any other class of occupants: if there's an occupant, the tax isn't payable!
The pretense that a conventional land tax can be passed on to tenants, although false, is plausible enough to be politically fatal, and the budget for promoting that claim always dwarfs the budget for refuting it. Moreover, the claim has the unassailable advantage that it takes fewer words to make than to refute. The corresponding claim concerning a vacancy tax can be refuted twice in a single soundbite: If there's a tenant, no tax is payable; if tax is payable, there's no tenant to whom to pass it on!
Moreover, the vacancy tax is openly designed to be avoided by adding to the effective supply of housing, thereby reducing rents. It is consistently presented in the media as having been invented for that purpose. It is mentioned only in that context. And its critics, as far as I have noticed, have not yet come up with any alleged reason why it wouldn't work as advertised.
Instead, the typical critic will claim that putting pressure on land owners to use their land is a violation of property rights. But if the owner of some other piece of land in the critic's suburb wants to use his/her land in a way that the critic doesn't like, it is a safe bet that the same critic will try to obstruct that owner's plans — not by buying out the land, nor even by taxing the owner's intended use, but by lobbying the government to forbid it outright, without compensation. Being forbidden to use the land optimally is far more injurious to the owner than being compelled to use it (or to sell it to someone who will). It is an even safer bet that the same critic, asked how he/she would prefer to finance government, will defend something that involves a clearer violation of property rights — such as being forbidden to deploy your produced capital without acting as an unpaid tax collector, or being forbidden to work without handing over some of the fruits of your labour. One way or another, the critic will be found to be self-servingly selective in protecting property rights.
Taxing land into use
A conventional land tax induces owners to use the land in order to generate income to cover the tax. A vacancy tax induces owners to use the land in order to avoid the tax. The latter inducement is stronger. And if you want people to believe that you're going to tax land into use, it helps if you're not proposing to levy your new tax on the users!
Marginal tax rate vs. participation tax rate
If you are the secondary earner of a household, your effective marginal tax rate is the fraction of your next dollar of income that the household loses to the government, whereas your effective participation tax rate is the fraction of your total income contribution that the household loses to the government. The marginal tax rate may affect how many hours you work, whereas the participation tax rate may decide whether you work at all.
Land also has a marginal tax rate, meaning the rate on the next dollar of production on that land, and a participation tax rate, meaning the rate on the decision to put the land to its present use rather than an inferior use. If a vacancy tax applies to the inferior use but not the present use, it reduces the participation tax rate by taxing non-participation.
Mason Gaffney and Fred Harrison, if I understand them correctly, have argued that conventional analysis greatly underestimates the deadweight cost of taxation because it allows only for the effect on the level of production on land in its present type of use, but ignores the effect on the type of use — in other words, because it allows for the marginal tax rate but not the participation tax rate. A suitably designed unoccupied land tax addresses the participation tax rate and therefore addresses the major cause of deadweight.
How then should we define “occupation”/“participation”? I submit that it depends on the goal.
If our goal is to obtain the best approximation to a Georgist system and its incentives, then land occupied in accordance with its current zoning should be exempt, while land occupied in accordance with lower zonings should be taxed at higher rates for higher degrees of under-occupation. Such a system would be complex by comparison with an all-in land-value tax, but not by comparison with income tax, or even with the mess that the municipal rates bill has become.
But such a system introduces a disadvantage: we cannot simply say that no tax is payable as long as the land is occupied by... [insert sacred cow here]. In the case of a non-owner-occupied site, we can argue that the owner must attract and retain a tenant in order to avoid a higher tax (and must moderate the rent accordingly), and that the owner can further reduce the tax bill by accommodating more tenants. That is more persuasive than arguing that the owner needs a tenant in order to cover the tax bill, or (worse) wheeling out the supply-demand graphs. But this arrangement does not give an automatic exemption to every owner-occupant who, for political reasons, might be thought to need one.
Alternatively, if our goal is to force a large, but not necessarily optimal, increase in the supply of accommodation at minimal political cost, then we can afford to be generous with the definition of occupation. The definition could be wide enough to grant a tax exemption to any perceived sacred cow, without being so wide as to exempt downtown apartments that have never been occupied (the bête noire du jour), or valuable downtown lots used as open-air car parks, or residential estates on which the developers graze a few token head of cattle while waiting for lot prices to rise.
For example, one could have exemptions for low levels of residential or business occupation, and for sites on which one intends to build one's principal residence, and for part-time residences used for earning income in multiple locations, and for holiday homes in locations where they cannot realistically be let all year round. For fairness, one would need exemptions for sites that are under redevelopment or waiting for building permits, or for which permits are under legal challenge, or for which letting is forbidden by covenants or body-corporate rules in place at the commencement of the tax (although there would also be a strong case for nullifying such covenants and rules as being detrimental to innocent third parties who were not consulted; and if that somewhat reduces the value of your principal residence, well, so do many other causes for which you wouldn't expect any compensation). These exemptions could open up many loopholes. But as long as a large class of sites could not pass through the loopholes, the tax would force a large increase in the supply of accommodation while remaining a tax on Them, not on Us.
What about rolling short-term letting, e.g. via Airbnb? From a purely Georgist viewpoint, such letting constitutes a higher use than long-term letting, because it yields more rent (hence its popularity with owners). And those who would want a vacancy-tax exemption for such letting presumably have more votes than those who would oppose it (i.e. the hotel industry). So it seems that both the Georgist argument and the political argument favour the exemption (although the Georgist argument would require the property to be fully developed). Against the exemption, one could argue that locals need the housing more than visitors do, and that locals need jobs in the hotel industry more than visitors need access to the local housing stock. In favour of the exemption, one could answer that hotels are inefficient because they offer a permanent supply to a seasonal demand. Such equivocation only reinforces the view that the larger constituency wins.
It has been well said that a tax on unoccupied land raises the problem of verifying alleged occupation, whereas an all-in land-value tax raises no such problem. But we have seen that if the land-value tax is required to coexist with other taxes, there are advantages in confining the land-value tax to unoccupied land, for which purpose we need a method of verifying occupation according to the applicable definition.
An obvious part of the solution is to require the owner of the address to prove the claimed degree of occupation. If you have ever submitted an income-tax return — and how many of us haven't? — you know that if you want to claim an exemption or deduction, the onus is on you to substantiate your claim. If you wanted to claim an exemption or concession under an unoccupied land tax, you would expect the rule to be the same. If, in addition, you are an orthodox Georgist and therefore tend to regard any exemption from a land-value tax as an exorbitant privilege, you will think it only reasonable that the party claiming the exemption must provide evidence of activities that contribute to the public purse through alternative mechanisms, indirect as they may be.
Critics, of course, will claim that the “reverse onus of proof” is a violation of fundamental human rights. In criminal matters, indeed it is: I have gone so far as to argue that it is unconstitutional in all jurisdictions — although that has not stopped the legislators from trying it on. Similar objections may arise in tax matters if criminal liability is at stake. But if the only thing at stake is the amount of tax payable, the reverse onus of proof is effectively a tax on the absence of proof. In a system that deliberately taxes production and employment, a tax on the absence of proof looks comparatively enlightened. Certainly it does not raise any human-rights issue that has not already been raised by other taxes.
Nor does it raise any practical issue that has not already been raised by other taxes. If the address is let (continuously or intermittently) for residential or commercial purposes, or owner-occupied for business, the proof of occupation will be a record of taxable transactions (in which case the desire to avoid the new tax will help to enforce the old ones). If the address is a principal residence, or a site on which one intends to build one's principal residence, there will already be mechanisms for establishing this fact for the purposes of land tax and capital-gains tax. If the address is one of several part-time residences used for earning income in multiple locations, there will already be mechanisms for establishing this fact for the purposes of income tax. If the address is occupied by a family member for less than its market rent, the need to prove occupation may be new; but the means of doing so will already have been devised for some other purpose, and assisting the owner to prove that the property is occupied will presumably be a condition of such occupation.
Of course the present tax system imposes participation tax rates not only on parcels of land, but also on other productive units, including people. The effect of an unoccupied land tax on these other participation tax rates would be small, and would come from reductions in other taxes. Large reductions in participation tax rates on factors other than land would require wider reform.
But the unoccupied land tax, by improving the availability of jobs and the affordability of housing and business accommodation, would raise people further from the edge of bankruptcy. I suggest that when people are too close to the edge, they can think of tax reform only in terms of whether they would pay more or less, whereas a retreat from the edge would make it easier to debate more sophisticated concepts such as marginal and participation tax rates, compliance burdens, and their effects on incentives — issues on which Georgists have a natural advantage.
* Disclaimers: In this article, the words own and tax and their derivatives, when used in connection with land, are to be taken only as notational conveniences. For present purposes it is not necessary to debate the breadth of the meaning of land, or whether a “pure Georgist system” collects the whole rental value or only enough to replace the minimum necessary revenue, or whether the shifting of transaction taxes onto land rents is complete or only partial.
1st Draft: Dec.17, 2016. 2nd Draft: Dec.24, 2016. Last modified Mar.5, 2017.
|Tweet||Return to Contents|