Harberger Tax

It’s always nice to see trails of thought connect up.

An idea I first encountered and really liked in a 2014 Steve Randy Waldman (interfluidity) post has apparently since acquired a name and a more extended provenance. Waldman’s post, Tax price, not value, presents the idea as a LVT/Georgism-flavored solution to NIMBYism enabled by artificially depressed property tax rates like so:

…There is, of course, a much easier way to gauge what a property would sell for: Solicit from its owner a price.

The price at which an owner would be willing to sell a thing has a particularly valuable characteristic. It limits the burden to alternative users of the exclusion in a property right. If the price is set low, a user harmed by exclusion can simply purchase the thing and have at. If the price is set high, alternative users may be seriously burdened yet be unable to buy access.

So, for the sorts of exclusion that do impose substantial burdens to alternative users, a natural policy intervention would be to require property owners to declare a price at which they commit to sell the property (for some period of time), and levy a tax of some legislatively determined percentage against that actual, actionable price, rather than a hypothetical market value. Property owners could pay as much or as little tax as they choose. When they set their price, they face a trade-off, between the risk of being undercompensated for losing the asset if the price is too low, and an exaggerated tax burden if they set a price so high that the risk of sale is negligible or the required overcompensation extreme. The owner is free to choose how much she values certainty of continued ownership, but she must pay for that.

The price set by the property owner might constitute an option to buy for all comers, or just for the state. (I’m not sure which would be best. What do you think?)

Posner and Weyl talk about essentially the same scheme in Property is Only Another Name for Monopoly and trace it to a 1965 paper by Arnold Harberger (which has a Latin American context/motivation — something about LatAm seems to encourage economics experimentation; probably US economists operating under moral hazard in authoritarian labs?). They’ve since written a book about such ideas I’ve been meaning to read, Radical Markets. The idea seems to be becoming increasingly popular in the Ethereum world as a way to actually set real prices in meaningful markets.

Schemes like this tend to be too simple, but in a good way. Starting incentive and mechanism design from a radical core can lead to meaningfully radical systems. A formula can beget revolutions. Vannevar Bush’s introduction of indirect cost support, the Black-Scholes formula, Vickrey (second-price) auctions come to mind. And if we’re lucky in the future, ranked-choice voting etc.

But for a scheme to have such potential there have to be mathematical rather than merely ideological reasons to prefer it. The Waldman idea stuck with me because it suddenly made Georgism make sense. Land-value taxation as such seems simply like non-property owners fighting an ideological battle with property owners. Tax income or wealth? Where you stand depends on where you sit. How much of each you have or expect to have. But Harberger tax? That elegantly threads the needle with a certain mathematical doomsday logic.

For the record, I’m not a pure Georgian/LVTist. The idea that all wealth derived from property stinks of mercantilist zero-sum thinking to me. I’m too Schumpeterian for that. I think wealth is a process not an asset. But Harberger tax… there’s a there there.

A naked Harberger tax would probably have all sorts of unpleasant consequences, but as the kernel of a more complex scheme, hmmm. A good formula is like construction material. You still have to learn to build with it. What can you build with Harberger taxes? Here’s a website I just found that seems to have some ideas.

Get Ribbonfarm in your inbox

Get new post updates by email

New post updates are sent out once a week

About Venkatesh Rao

Venkat is the founder and editor-in-chief of ribbonfarm. Follow him on Twitter


  1. You’ve already said “A naked Harberger tax would probably have all sorts of unpleasant consequences,” so this isn’t exactly a gotcha, but as a homeowner the prospect of having to do this for myself is extremely nerve-racking; the idea that any idiot with enough money could force me to move all my stuff and put my kids in different schools is hard enough to stomach, and then add to that the pressure of ensuring I set the valuation high enough — and respond swiftly and correctly to dynamics in the housing market — to ensure a reasonably soft landing in the event, and it sounds like a recipe for mental illness. (There are probably some similarities between my stomach-churning counterfactual and the way more economically precarious people in the US are currently forced to live that are worth contemplating.)

    It is a super interesting idea, though, and like you I could see using it as part of a scheme with guardrails against the super-rich forcing folks out of their homes just for the hell of it.

    • I think some other method would be needed for primary residences. Absent a specific reason to move, the value to me of the place my family and I live is always going to be far higher than what the market would actually pay, and it’s not right to charge people tax on that subjective value.

    • the scary thing is that the principal of protecting ‘subjective value’ is exactly how nimbyism gets incredible power today. zoning exists to protect subjective value, tax relief and benefits for mortgage-owners that don’t exist for renters protect subjective value, and on and on.

      i think unless a person lives on incredibly valuable land it’s unlikely that this is a big concern. depending on tax rates themselves, simply adding the subjective value to your home until it is being taxed at the level that would actually make you happy to sell could work. that said, i think such a tax scheme would create such radically different incentives for building housing that the post-change equilibrium would not result in a lot of people being bought out of their homes.

      adding some barriers to primary residences to slow that situation, especially during a ‘transition’ to the new equilibrium over a few decades may be fair, but i get worried that it would be a pretty short road to re-creating prop 13

  2. Wouldn’t it be a stable strategy under the Harberger tax system to push for aggressive devaluation? Let’s say I poison some land s.t. it becomes rotten and worthless, s.t. no one wants to have it and for the same reason the self rating is low and the taxation accordingly. If cybepunk dystopia was bad, what about the shiny Tschernobyl home zone?

    • this is only a big risk if your concept of ‘value’ is so distinct from the rest of society that you would *want* to live on land nobody else likes. oscar the grouch would do very well.

  3. The idea dates back much further—I don’t know that it’s the first instance, but at a minimum, Sun Yat-Sen proposed it in 1918 in his “Memoirs of a Chinese Revolutionary.” Sun, however, understood that it only suffices as a means of evaluating the value of land in a largely-agrarian environment, because a Harberger tax has no means of divining what value comes from land versus the structure built on it.

    I’ll point out that the Georgist position isn’t about playing tug-of-war over value—it’s about the fact that the value of land has no upward limit, and constructs a permanently decelerationist rentier class. Imagine shipwrecking on an abandoned island, and finding that there was exactly one occupant of it, who declared that he owned the island and you had to pay rent—this is where the core intuition comes from to me.

  4. Implementation details are crucial. Despite some ideas in the linked website, the nature of Harberger programs into the system some undesirable properties that are hard to improve upon. For example, in addition to the explicit tax, there is necessarily an implicit value leak in the form of an option which the owner must write. This option also has a market price and thus is an additional (though implicit) tax. In the vanilla model, this implicit tax is leaked away to active property traders and does not contribute to the state tax revenue. I remain open to be proven wrong though, I’d more than welcome a single example where Harberger-like system operates smoothly.

Leave a Comment


This site uses Akismet to reduce spam. Learn how your comment data is processed.