No matter how it is financed, government spending, in the long run, crowds out private spending by using up scarce resources which could otherwise have gone to private uses. Depending on how taxes are structured, they can take us even further from 1:1 crowding out by distorting relative prices and redirecting effort from production to tax avoidance. In some cases though, taxes can improve outcomes by internalizing externalities in a way which has lower transactions costs than Coasean bargaining. Externality taxes have some knowledge problems associated with them, but in general, it’s better to tax things which are bad than to tax things which are good, because you get less of the things you tax. If you can’t tax bad things, you should tax things which don’t react much to the tax. Since the quantity of land does not change, taxing it does not impact real economic activity, or so the reasoning goes.
Justifications and Philosophy
Historically, the justifications for Georgism are based on the idea that man is entitled to the sweat of his brow, but natural resources belong equally to all. I have two problems with this line of thinking. First, property rights are based on the ability to coerce others. Belonging equally to everyone implies everyone has equal ability to coerce, which is simply not true. Land is a private good, meaning that everyone can’t share it equally. Secondly, “belonging to everyone” and “belonging to the government” are two vastly different things. The government is one group among many in society, separated from others because of its comparative advantage in violence, not because it “cares about everyone’s welfare” or other such nonsense. Instead of abstract philosophy, I prefer the pragmatic: Georgism is justified if land taxes distort economic activity less than other taxes.
Normally, when a good is taxed, the quantity traded decreased from Qe to Qt (on the graph below), and the price people pay increases from Pe to Pc, whereas the price suppliers receive decreases from Pe to Pp. The deadweight loss of taxation is due to the reduction in quantity of goods traded.
When an inelastic good is taxed, neither the price paid by the consumer nor the quantity supplied changes, hence there is no deadweight loss. The tax imposes a loss on the holder or producer of the good equal to the present value of the stream of tax payments.
People make decisions on the margin, and those decisions take place at a moment in time. Taxation which affects decisions causes deadweight losses, but once a decision has been made, it can be taxed without changing the past, which governments are tempted to exploit. But past taxes affect people’s expectations of future taxes.
Let’s say you are a miner. You look off in the distance and see a mountain which you think might have iron ore. Currently, the mountain is worth $10 million, but if ore were discovered, it would be worth $20 million and the costs of exploring are $1 million. In the absence of taxes, you will buy the mountain and explore it if the odds of finding ore are 10% or higher. Let’s say the government implements a land tax of 10% per year. Assuming an interest rate of 5%, the initial value of the mountain will drop to $3.33 million, and the value of the mountain with ore, post-tax, will be $6.66 million. Now, the miner will only buy the mountain if the expected chance of discovering ore is 1/3. If the tax were levied on the miner after they already did the exploration, the miner would not change their mining efforts at all and the tax would have no effect on production. However, if the tax were levied prior to exploration, or if the miner expected to be taxed in the future, the exploration and development efforts would be dramatically reduced.
Because of the efficient market hypothesis, an expected future tax has a similar effect on asset prices today as an actual tax (discounted by the expected time from now until when the tax will begin). So the secret power of land taxes isn’t anything special about land compared to other assets, it’s that the tax is on economic rents. Any tax on the returns to a past action will have similar inelasticity, so long as people in the past did not expect the tax.
If the government improves the value of land, such as building roads or public parks, they could tax the subsequent increase in land prices without deadweight loss, because they would simply be capturing the positive externalities of their own action. Such a tax would incentivize local governments to more optimally improve amenities. One advantage of a land value tax is that it would incentivize governments to do good policies which increase the value of the lands they govern.
The Caplan Gochenour Critique
There is no such thing as a purely natural resource. Everything which we think of as a natural resource requires human effort to discover and find uses for. Land’s value depends on improvements, network effects, discovery of resources, and other things which require human effort to produce. Georgists attempt to dodge this somewhat by taxing only the “base value” of land and not improvements. Another way to say this is that they won’t tax recent improvements in the land’s value, but on a long enough time schedule, everything is marginal. The fact that downtown Manhattan land costs $1,000+ per square foot has nothing to do with the natural land, but it has everything to do with the investments which people have put into the surrounding area. Truly unimproved land is very cheap. Land taxes target the positive externalities which improvements to one plot of land give to surrounding plots, and taxing positive externalities is even worse for welfare than taxing goods.
In order to affect economic activity minimally, it is important to tax broadly, so that no one good becomes so expensive that people dramatically change their consumption or production of it. At first, it seems like because Georgism advocates a tax on an inelastic resource, it would not affect behavior at all, but in fact, it would put all of the tax burden on one class of good, which can, in fact respond to taxation. For a Georgist tax to draw reasonable revenue, it would have to tax some of the value of the improvement of land, and once it did that, the tax would discourage land improvement and entrepreneurial discovery of new uses for the land. Land taxes may be useful in some situations, such as when the government wishes to capture the rents from past investments, or when financial transactions are hard to observe. But for the most part, a uniform consumption tax is probably superior in terms of welfare. Land taxes, if truly applied to the unimproved base value of land, would be non-distortionary, but to figure out how to calculate that price is a difficult task.
Econlib’s bio of Henry George
Costs of Taxation
Caplan Gochenour critique of Georgism
Steven Landsburg on optimal taxation.
The Ramsey Problem (wonkish)
Rothbard on Georgism
How would Georgism work in Monopoly?