An article in the August 9 issue of The Economist, “The Time May Be Right for Land-Value Taxes,” suggests to reconsider the land-value tax advocated by American economist Henry George in his 1879 book Progress and Poverty.
George proposed to tax away the rent on the unimproved value of land, and to replace all other taxes by that single one. The improved value, represented for example by a house or an industrial building, would not be taxed. Presumably, the implicit rent a homeowner earns on the land on which his house stands would be taxed. The receipts from the tax would finance public goods. These goods boost land rents–think about a police station or even a park or a road–and benefit the landowners more than the landless, who have paid wage taxes and other taxes to finance the public goods. In other words, the rent tax would be redistributed to the landless in free public goods.
By the way, the Economist article contains many good illustrations of economic theory. For example, a land tax is capitalized in land values, which decrease by the present value of future tax payments. A Danish government study apparently provided an empirical confirmation (a useful citation would have been appreciated).
A convincing argument exists for taxes on unimproved land, from the double perspective of economics and ethics.
The economic argument is that land being in fixed supply, a tax on unimproved land does not change the allocation of resources. It will be totally capitalized—negatively—in the land prices. In traditional public finance theory, a tax on unimproved land is an efficient tax. In 1978, Milton Friedman said that “the least bad tax is the property tax on the unimproved value of land, the Henry George argument of many, many years ago” (quoted by Fred Foldvary of San Jose State University, a proponent of the Georgist tax).
The basis of the ethical or distributive argument for Georgist taxes is well summarized by The Economist:
The most famous attempt to justify land ownership was made by John Locke in the 17th century. He argued that because people own their own labour, toiling on the land confers ownership rights over the resulting product (a farm, say). Yet even Locke said this only works as long as there was “enough, and as good, left in common for others”. This proviso may be met in a wilderness but not in booming cities. In any case, such a combination seems a shaky justification for acquisition. Robert Nozick, a 20th-century libertarian philosopher, doubted whether pouring his can of tomato juice into the sea, combining the two, meant that he could then claim ownership over the ocean. Without a good basis for land ownership, how can it be fair for landlords to get rich from rent?
As attractive as a Georgist tax or other similar land-value taxes appear, there are many issues to explore. Let me mention a few. (The more I talked to Fred, the fewer issues I was left with! Perhaps I am, like it has been said of Donald Trump, overly influenced by the last person I talked to?)
First, even taxing the whole rent would not produce enough government revenues to replace all existing taxes. Total (federal, state, and local) government expenditures in the US amount to about $6 trillion a year. Economist William Larson estimated the value of all land in the US at $23 trillion in 2009; let’s increase the estimate to, say, $50 trillion in order to account for increasing land prices since then. Assuming an annual rent of 5%, taxing all of it away would yield “only” $2.5 trillion a year.
This problem, however, would be more an advantage than a disadvantage if it could force a reduction of government expenditures by some 40% ($2.5 trillion is 42% of $6 trillion). Such a change could be gradually realized over several years, perhaps a generation. Of course, this scenario assumes that governments can make and keep such a commitment. It is even more difficult to imagine in a federal system.
A second issue is that a new land tax expropriates, at least partly, the current owners of land, who paid for the capitalized value of future non-taxed (or lightly taxed) rents when they purchased their land. This problem could be solved by reducing the landowners’ income taxes or other taxes by an amount equivalent to their capital loss. As all governments in America levy about $5 trillion in taxes per year, a long transition would again be necessary.
A third issue–or perhaps a complex of issues–is whether land-value taxes should tax away the whole rent (or more) as George wanted, or only part of it as Friedman probably envisaged. A related issue is whether confiscation of the whole rent by government means the same as the nationalization of land. And once the government has this whole field of taxation open to it, will it not stealthily nationalize all land by charging land taxes higher than the rent and high enough to make the private possession of land uneconomical?
It seems to me that there are three possible lines of argument against taxing all land rents, two of a moral nature and one more economic:
-
Reformulate the Lockean argument in a way to make it defensible.
-
Argue, like Anthony de Jasay, for a presumption of liberty and a prima facie case that property simply is, and may only be contested if a precise tort can be proven. (See my Econlib article “An Unavoidable Theory of the State,” and the links therein.)
-
Make a public-choice argument to the effect that a government with a monopoly of land would a very dangerous Leviathan. As argued by Geoffrey Brennan and James Buchanan in their 1980 book Power to Tax, the unrestrained Leviathan will charge in taxes what the market will bear.
We seem to always come back to the question of the justification of the state and, if such a justification is found, to the problem of chaining Leviathan.
READER COMMENTS
Bill Workman
Aug 17 2018 at 11:09am
Your observation, “A second issue is that a new land tax expropriates, at least partly, the current owners of land, who paid for the capitalized value of future non-taxed (or lightly taxed) rents when they purchased their land” brought to mind Tullock’s “Transitional Gains Trap” paper:
https://www.jstor.org/stable/3003249?seq=1#page_scan_tab_contents
Fred E Foldvary
Aug 17 2018 at 11:17am
The phrase “a government with a monopoly of land” is unclear. It should be changed to ” a government with a monopoly of land rent.” In Georgist policy, possession – including use and transfer – is individual. As to government collecting taxes higher than the rent, in a proper system, land holders are able to appeal the assessment to a board and ultimately to a jury. If the county assessors are generally over-assessing, they would be sued for fraud. As to public choice, the taxation of income and sales are much more centralizing than a land value tax, which facilitates local government.
robc
Aug 17 2018 at 2:59pm
I have written a couple of articles on the SLT at glibertarians.com. Here is the most recent. It is kind of goofy, but I thought some of you might enjoy it.
https://glibertarians.com/2018/04/funding-libertopia-a-gedankenexperiment/?highlight=gedankenexperiment
James
Aug 17 2018 at 8:02pm
How does a land tax not influence hebavior?
The unimproveed value of land is not some objective physical characteristic of land. If Alice owns some undeveloped land and Bob buy some adjacent property and improves it, the market value of Alice’s land will increase and so will her taxes even if she leaves her property unimproved. Alice will be incented to motivate Bob’s development decisions.
Pierre Lemieux
Aug 17 2018 at 9:21pm
@James: It does not influence behavior in the sense that, ceteris paribus, it does not change the supply of (unimproved) land and, therefore, does not change the quantity demanded either. It only affects the market value of land.
benjamin weenen
Aug 18 2018 at 5:37pm
If Alice’s land increases in value, that is because it’s productivity has too. There is an opportunity cost to Alice of not making best use of that productive potential, LVT or no LVT.
The LVT is set as the opportunity cost of leaving a site unused, so merely turns a notional cost into a real one. In theory this shouldn’t change incentives, but in practice it might as there is often a mismatch between the two. So the LVT corrects market inefficiencies, aiding resource allocation.
benjamin weenen
Aug 18 2018 at 7:45am
Like the payment of wages, or for goods and services provided, a LVT (land fees) is compensation for loss of opportunity. That it is collected, spent or redistributed by the state is a separate issue.
Failure to do so bakes in excessive inequalities and resource misallocation. This has necessitated the socialisation of incomes and capital in order to partially mitigate the ill effects of excessive inequalities, while adding further economic inefficiencies.
As all taxes on output are incident to some degree or another on land values, their reduction and replacement would see a rise in aggregated land rents. As would and increase in GDP from the alleviation of deadweight losses.
While it is unlikely that a land rents alone could cover total government spending, in conjunction with other user fees, royalties and Pigou taxes it could.
It would be best if these were all collected by a non-governmental agency and re-distributed as an equal share ie a Citizens Dividend.
Government spending on the services it provides then need only be financed from a Poll Tax/Head Tax.
This would reduce inequality (eliminate excessive inequalities) and optimise the economy for growth.
Furthermore, it would mean all non-citizens are tax free, though of course they wouldn’t be entitled to the dividend until they gained citizenship.
Those citizens who object to paying tax on the “all taxes are theft principle” need only give up their citizenship and dividend to go tax free.
Jon Murphy
Aug 18 2018 at 9:29am
Can we say that the supply of land is fixed? I’m not sure that really holds true. To wit: Boston’s Back Bay project where they turned part of Boston Harbor into a landmass. Or the artificial islands in Dubai.
benjamin weenen
Aug 18 2018 at 5:23pm
While volcanoes and subduction change the amount of dry land, and the amount of fish in the sea or oil under the ground changes, because Land is defined as everything not supplied by human effort, it is perfectly inelastic in supply.
Reclaimed land is thus an improvement i.e capital. If said piece of reclaimed land, like any other form of immovable capital, were to occupy a relatively productive, thus valuable location, it would be subject to a LVT if such a tax were in use.
Matthias Goergens
Aug 19 2018 at 9:34am
The author way to effectively gain more productive land is changes in zoning and other regulations.
But a LVT aligns interests nicely here, especially if the level of government that has zoning authority also collects the LVT in question.
(In practice deciding the level of government to collect which portion of LVT is probably an important question. Another important question would be: If you want to couple an LVT with a basic income over which area do you want to average the payments. Eg the whole country, a state or a county or a city?)
benjamin weenen
Aug 19 2018 at 11:40pm
Small plot owners cause a tragedy of the commons by trying to maximize development, thus incomes from that plot. This is why we need zoning/planning in order to quota our shared environment.
Large landowners with many adjacent plots don’t do this. In order to maximise their returns from all plots, they must balance development with the need to preserve/enhance the location their property occupies.
In essence, the LVT turns the state into the largest landowner. In order to maximise revenues, it would have to balance development needed for growth while providing locational amenity that people would prefer to spend their incomes on rather than alternative goods/services. From this, resources are optimally allocated in an economy, maximising society’s wealth and welfare.
So yes, the LVT aligns the incentives of everyone, including those between the state and it’s citizens.
100% of land’s rental value should be collected 100% of the time. Only possible in theory, but like all forms of justice, we should strive to get as close as possible. This is both optimally fair, thus efficient.
LVT is compensation for loss of opportunity of we should all be entitled to an equal share. Ideally this would be everyone in the world. However, we don’t yet live in a world where that is possible, but it can be done within a nation. That can be done as a UBI, or as a central government grant to the nation’s regions set on an equal per capita basis.
LVT should not be a local tax, as this doesn’t solve much. The wider it is collected and redistributed the better.
Matthias Goergens
Aug 19 2018 at 9:29am
Reducing other taxes would increase land rents. (Land prices on both sides of the Swiss/German border are a good illustration of that.)
I am not sure whether each one dollar reduction in other taxes would increase land rents by at least a dollar.
Compare the Henry George theorem.
Land Value Taxes are levied on the market value of the unimproved land. As the Economist notes the tax gets capitalized back into the value of the land. Let’s compare the income available to an investor when owning a perpetual bonds with land ownership:
Income (from bonds) = (bond) price * interest rate
Income (from land) = land rent – (land) price * land value tax rate
An arbitrage argument suggests that equal yearly income should be had for the same price:
price * interest rate = land rent – price * LVT rate
Solve for price:
price = land rent / (interest rate + LVT rate)
Given this knowledge we can now calculate how a given land value tax rate translates into an effective land rent tax rate:
effective land rent tax rate = (price * LVT rate) / land rent = LVT rate / (interest rate + LVT rate)
Assuming a positive interest rate, that means that no matter how high the LVT rate goes, the effective land tax rate approaches but never reaches 100%.
Expressed alternatively: even for an LVT of something absurd like 250% per year, there would be a low enough land price to make private ownership of land profitable. And, increasing LVT above a certain minimum wouldn’t raise the tax take much—but also wouldn’t do any damage to the economy.
Yes, that’s the whole point of an LVT. Asymptotically it gives the whole land rent of the country to the government. Nothing more and nothing less. It’s not a drag on any economic activity, because that economic activity would pay the land rent anyway.
The main practical problem with government monopoly is that the government is inefficient and has bad incentives. But that doesn’t apply in the land case, as the government effectively just auctions off the land to the highest bidders.
(Similar to how Singapore’s government auctions off the right to operate cars, yet no one thinks that they have a monopoly on cars.)
Another interesting way to attack this problem would be to pay out a large chunk of the land value tax take as basic income equally to all residents.
Thanks to economic inequality, the median person owns significantly less than the average amount of land. If memory serves right, that includes most owner occupiers.
Transaction taxes like stamp duty should be abolished, and past payments could be claimed as a credit against LVT. (Count them on the same basis as-if LVT had been in effect. Eg, if you paid 10% stamp duty a year and a half ago, and LVT is set at 5% per year, than you could claim a tax credit of 10% – 5%*1.5 = 2.5% LVT rebate in the first year.)
Pierre Lemieux
Aug 21 2018 at 6:52pm
Thanks, @Matthias, good explanation.
Matthias Goergens
Aug 19 2018 at 9:55am
Hmm, I typed a long comment, but it seems to have gotten eaten by the spam filter?
The gist is that:
If LVT is levied on the market price of the land, there’s always an equilibrium price of land that makes land ownership profitable, and that keeps the tax for any parcel of land at slightly less than the land rent.
Reducing other taxes increases land rent. Thus increasing the LVT tax take. Current land prices (and implied land rents) are only a lower bound on the possible LVT tax take.
Government ownership itself is not a problem. Eg Singapore auctions off the right to operate cars, yet doesn’t directly control cars.
Thanks to economic inequality, the median person owns much less than the average amount of land. Thus there will be lots of extra revenue available to pay off even the owner-occupier middle class with a universal income that’s greater than their land tax bill.
Gene Laber
Aug 20 2018 at 8:12am
Paragraph four: “A convincing argument exists……….improved land”
Don’t you mean unimproved land?
Pierre Lemieux
Aug 21 2018 at 10:56am
Indeed, Gene. It was a very bad typo. I just made the correction. Thanks!
Adam
Aug 20 2018 at 2:03pm
Henry George’s tax didn’t consider the link between location and land rent. Muth’s central city model is instructive. People work in a central business district (CBD) and live the area surrounding the CBD. Travel time varies with distant from the CBD, so people trade off the (a) cost of locating closer to the CBD and (b) the cost of travel time. Land rents get bid up closer to the CBD and fall as the travel increases. An efficient or nth best land tax accounts for time values, congestion in the CBD and congestion in travel. The Henry George 100% tax is inefficient and, IMHO, foolish. Here’s one analysis of property tax in a Muthian model : https://editorialexpress.com/cgi-bin/conference/download.cgi?db_name=UEA2013&paper_id=17
Fred Foldvary
Aug 21 2018 at 12:36pm
The Henry George 100% tax is on economic rent, not the total rental paid by a tenant. The economic rent is the portion of rental not needed to be kept by the title holder in order to put that land to its most productive use. Therefore, by definition, a tax on economic rent is not inefficient; it has no deadweight loss.
Pierre Lemieux
Aug 22 2018 at 6:12pm
Interesting point, Adam.
Jeff
Aug 21 2018 at 1:50pm
If you’re taxing either the value or the rent of unimproved land, but the land actually is improved, with buildings, water, sewer lines, etc., how do you value the improvements vs the unimproved land? Do you just assess the improvements at their original cost? Replacement cost? What if nobody actually knows how much the sewer line is worth, since it was put in decades ago?
If you value improvements at cost, what do you do about worthless (i.e., market price is zero) land parcels in Detroit that have houses on them? Clearly, those houses cost something to build or replace, so the implication then is that the land has negative value. So should we be paying negative taxes to the owners?
I know my county assesses my house and land separately, but I’ve never found their breakdown especially convincing.
Jeremy Edwards
Aug 22 2018 at 11:56am
The author is clearly not familiar with LVT in any great detail.
The first objection – that a LVT would not be sufficient to cover all government expenditure – is completely bungled. You cannot use $23t as an estimate for land values in the U.S. because land values are currently suppressed by taxation on labour and capital. This is the principle of ATCOR (all taxes come out of rents) that Professor Mason Gaffney has written about in great detail.
It is hard to get an accurate estimate of how much money a land value tax would initially raise, especially if the tax is phased in over a decade or so to avoid a market crash or credit crunch, but ultimately, yes, a land value tax would be sufficient for all government expenditure.
Moreover, plenty of neo-Georgists favour using a LVT to fund a UBI while keeping progressive taxes on capital and income, at least poverty is alleviated and government expenditure can be reduced to the natural cap set by land values.
Comments are closed.