Suppose you’re renting a long-term reserved parking space from the government, and they raise the price. You can now decide to vacate it if you don’t think it’s worth it. That’s a “user fee”.
But suppose instead you pre-pay for the spot for an entire year at a rate of 10$/day, and then the government doubles the price. Your bill has now immediately increased by 3,650$, and there’s nothing you can do about it. If you sell/sublease the parking space, it’s only worth the original amount you paid for it, so that’s no help. This has the same economic impact as if you just picked a random citizen and slapped him with a 3,650$ tax for no particular reason.
Flawed Policy #1: Increasing the price after the sale
Next imagine that the government also allows no option to buy the space outright, no matter the price. An economist points out that the net present value of the space is 91,250$ (assuming a 4% risk-free interest rate). He offers to pay the government 200,000$—more than double the net present value—for the space. The government says no. This is terrible for society because that money could have been invested and easily paid out greater expected dividends than the parking space offers.
Flawed Policy #2: Insisting on holding a particular asset type regardless of value relative to other asset types
These two “flawed policies” are precisely the problem with a Land Value Tax (LVT) as commonly proposed.
Rectifying the LVT
For a Land Value Tax to make any sense, it has to be instated at the time of initial acquisition from the government, just like the ongoing parking fee we discussed above. If the LVT goes into effect after the land acquisition, then it’s like the second parking example; the landowner double pays.
Second, landowners must have the option to purchase land outright in lieu of the LVT, otherwise society is worse off. If you have even a shadow of a doubt about this, realize that government pension funds are already invested in countless assets other than land. If you insist the government prefer land over any other asset type, then you’re effectively arguing that these pension funds should convert all their non-land holdings to land. Any fund manager would call you crazy if you suggested that.
Jousting at Windmills
Now imagine you “rectify the LVT” as I just described. Some people will rent land from the government. Others will buy it outright. Those who buy it outright are just like the landowners we have right now. A rationally implemented LVT is essentially no different than the system we already have.
As I explain in another post,
There are precisely three reasons for any tax or subsidy. They are:
1. To directly affect the welfare of the recipient. (Money has a decreasing marginal utility, so you generally want an extremely progressive wealth tax, with redistribution to poorer citizens.)
3. Because the income was unearned, thus taxing it away won’t distort incomes. (Incredibly minor in significance compared to the first two.)
The LVT supposedly serves #3, because it taxes away positive externalities that cannot affect the behavior of the landowner, thus cannot disincentivize productivity like a wealth tax can. However this is very minor concern compared to the marginal utility of a tax as a function of wealth.
For example, suppose you can tax 1M$ away from someone worth exactly 1M$, but who won it all by mere chance, vs. taxing that same amount away from a billionaire who’s wealth was entirely a function of his hard work. The billionaire’s utility is barely budged by the tax, whereas the proverbial “lottery winner” is wiped out by the tax, down to a zero net value. That drastically decreases his utility.
While a wealth tax may technically be a disincentive on productivity, that’s a pretty absurd argument given that every rich person who continues to work is by definition not affected much by incentives, given that the marginal utility of a rich person’s dollars fast approaches nil.
And the act of working to increase the external value of one’s property is obviously not a negative externality that should be discouraged. An LVT is not Pigovian.
Missing Our Target
Even if we ignore the differential marginal utility of wealth, and buy the argument that we want to tax “unearned” wealth, an LVT effectively becomes a random tax on a name picked out of the phone book. For instance, suppose Alice buys land at a value of X. It appreciates to X+Y. Then she sells it to Bob for X+Y. Then the LVT goes into effect. Bob suddenly loses the assessed net present value of all future LVT payments, i.e. Y. If he sells it to Eve, she pays X, not X+Y. So in effect, 100% of the LVT is paid by Bob. So an LVT is effectively just a one-time tax on the poor sap who happens to own the land when it goes into effect. You might as well throw darts at a map and tax the targeted household by some arbitrary amount of money.
Now technically we would want to say that Bob loses Y. But actually, an LVT is (insanely) discussed as being some arbitrary percentage tax on X+Y rather than a tax purely on 100% of the value of Y. So actually, Bob doesn’t lose Y, but instead loses the net present value of e.g. a 4% tax on X+Y. And if Y should happen to be a negative number (i.e. the land depreciates), then does Bob get a check back from the government? I haven’t heard any Georgists touch on this one.
Nevermind the insanity of trying to properly calculate what percentage of a property’s value is “unimproved”.
A common pro-LVT argument is that it incentivizes land owners to put their land to the most productive use. E.g. this article from The Economist says that LVT
may even stimulate economic activity, by penalising those who hoard land and keep it idle (a big plus in desolate post-industrial cities where much land is vacant)
There are two critical flaws in this argument:
- It’s wrong. The LVT is not dependent on the “idle-ness” of the land. As we just saw, it is purely based on external factors. It is a function of the unimproved value. As the first sentence of the wiki page on incentive says, “An incentive is a contingent motivator.”. But the LVT isn’t contingent on anything the owner does, therefore it can’t be a motivator. If I tell you you’re going to get a spanking whether you’re good or bad, that’s not an “incentive” to be good.
- Landowners already have an incentive to put their land to the most productive use, or to sell it to someone who can (who will thus be willing to pay more for it than it’s worth to the owner). There is no negative externality here to account for.
The article continues, correctly describing the effect a LVT has:
The tax drives the land price down by the capitalised value of the future levies
Yes! This is why Alice only paid 2M$ for the property instead of 3M$. But this isn’t a good thing! It’s effectively a tax on Bob based purely on the fact that he happened to be holding land at the time the LVT went into effect. If he had sold it to Alice previous day, then Alice would be 1M$ poorer. This is completely arbitrary. It does not optimize anything.
Share the Wealth
Another line of reasoning essentially says, “share the wealth”.
When public investment improves the value of a site — for example by building a new road nearby — the benefit comes back to the community in the form of higher tax receipts, rather than ending up as a windfall in the pockets of the owners.
First, when addressing an externality, we only want to tax or subsidize the affected party in order to optimize behavior. In this case, we want to subsidize people in order to incentivize those behaviors which added value to the property, e.g. the building of a new road. That does not imply that we should fund the subsidy by a tax on the individual who received the benefit.
This might be easier to see with a tax instead of a subsidy. Consider a negative externality like CO2. We want to tax its emissions to disincentivize them, and produce the optimal behavior that maximizes economic value in the market. But that doesn’t mean that the most optimal way to spend those tax proceeds is to give them to those individuals negatively impacted by the CO2. It’s possible that human welfare could be increased more by using the money in some completely unrelated way, like treating schistosomiasis.
The second flaw in the quoted argument is this. Suppose the unimproved value of the land is X, and the publicly added value is Y. The correct Pigovian subsidy would be to credit the exact amount of Y back to those individuals who added that value (e.g. the taxpayers who funded that “new road nearby”), as precisely as feasible as possible. A LVT would instead credit the community some arbitrary percentage of X+Y, rather than by the exact amount of Y.
Many LVT advocates seek to reduce the incentive for land speculation, which is defined as:
This is essentially the same as the previous “share the wealth” argument. What people seem to recognizing here is that the profit from speculation is essentially a capturing of the value created by external factors like community investment. They don’t like the idea of a speculator reaping the benefits of other people’s work. But this is simply the same fallacy of linking a tax and subsidy. Just as you don’t necessarily want to give CO2 tax proceeds to those negatively impacted by CO2, you don’t necessarily want to fund the community benefits subsidy by taxing land speculators. Speculation itself is not a negative externality.
Nothing Is Special about Land
Lastly there’s nothing special about “unimproved value of land”. Positive externalities are positive externalities. If the opening credits to a popular ’80s sitcom feature your home, increasing its value by a million dollars, that’s a positive externality. The same Pigovian rationale says we should give a million dollars to the show’s producers. Perhaps the most bizarre thing about the LVT is its core focus on land as opposed to any other arbitrary asset.