Conditional Harberger Tax Games

Baron Georges-Eugène Haussmann … transformed Paris with dazzling avenues, parks and other lasting renovations between 1853 and 1870. … Haussmann… resolved early on to pay generous compensation to [Paris] property owners, and he did. … [He] hoped to repay the larger loans he obtained from the private sector by capturing some of the increased value of properties lining along the roads he built. … [He] did confiscate properties on both sides of his new thoroughfares, and he had their edifices rebuilt. … Council of State … forced him to return these beautifully renovated properties to their original owners, who thus captured all of their increased value. (more)

In my last post I described abstractly how a system of conditional Harberger taxes (CHT) could help deal with zoning and other key city land use decisions. In this post, let me say a bit more about the behaviors I think we’d actually see in such a system. (I’m only considering here such taxes for land and property tied to land.)

First, I while many property owners would personally manage their official declared property values, many others would have them set by an agent or an app. Agents and apps may often come packaged with insurance against various things that can go wrong, such as losing one’s property.

Second, yes, under CHT, sometimes people would (be paid well to) lose their property. This would almost always be because someone else credibly demonstrated that they expect to gain more value from it. Even if owners strategically or mistakenly declare values too low, the feature I suggested of being able to buy back a property by paying a 1% premium would ensure that pricing errors don’t cause property misallocations. The highest value uses of land can change, and one of the big positive features of this system is that it makes the usage changes that should then result easier to achieve. In my mind that’s a feature, not a bug. Yes, owners could buy insurance against the risk of losing a property, though that needn’t result in getting their property back.

In the ancient world, it was common for people to keep the same marriage, home, neighbors, job, family, and religion for their entire life. In the modern world, in contrast, we expect many big changes during our lifetimes. While we can mostly count on family and religion remaining constant, we must accept bigger chances of change to marriages, neighbors, and jobs. Even our software environments change in ways we can’t control when new versions are issued. Renters today accept big risks of home changes, and even home “owners” face big risks due to job and financial risks. All of which seems normal and reasonable. Yes, a few people seem quite obsessed with wanting absolute guarantees on preservation of old property usage, but I can’t sympathize much with such fetishes for inefficient stasis.

Under CHT, property owners would be tempted to declare values below their actual values, to lower their taxes. For example, they might try to estimate a distribution over the second highest value, and trade a gain from lower taxes against a risk of losing their property temporarily and having to buy it back at a higher price. However, as I’ve discussed, speculators would hunt for signs of owners lowballing declared values, and buy such properties in the hope of selling them back to owners at a higher price. In practice, this would ensure declared values didn’t fall much below actual values.

An owner seeking to leave an existing property soon would typically have to offer a substantial discount over typical nearby declared values, in order to entice buyers to take their property in particular. However, an owner seeking to switch properties, but who is flexible on their exact new property, has a comparative advantage in the above described speculation game. That is, they could gain a substantial discount on their purchase via having a sequence of owners pay them to get their properties back, and then finally switching to the property whose owner refuses to buy it back. For example, if 90% of owners will pay 1% of value to get their property back, a flexible buyer can on average gain a 10% price discount.

Now consider some concrete examples of using CHT to change rules limiting property use:

  1. Min Yard Variance – A property owner asks to build a structure closer to a lot line than is usually allowed.
  2. Less Parking Variance – A store currently required to have X parking spots asks to cut this by Y spots.
  3. New Park – Make a new park out of what is currently an empty lot. Specifies park quality, entrance fees.

Today such decisions tend to be made in part bureaucratically, and in part politically. The bureaucratic part tends to make decisions crude, failing to take into account many relevant details. In contrast, while the political part can consider more detail, it tends to be more random and expensive, an expense that makes politics mostly irrelevant for small changes. Both of these parts consistently create both winners and losers, inducing efforts to “fight” in order to win and not lose. In contrast, while CHT also makes winners and losers, and thus induces fights, it could be a lot less random and expensive, and take into account more local detail.

A land use rule change would usually be proposed by an owner who expects it to increase the value of their property. They would pay an auction fee to put their proposal on the official agenda, and once their proposal became official they’d declare a higher conditional value for their property. Other property owners would learn of this proposal, in part via prediction markets on the chance that this change will happen, and on which properties are likely to be substantially influenced. These other owners might then also declare new conditional property values, either higher or lower. Some property owners would personally track change proposals and change conditional declared values; others would delegate such tasks to an app or a human agent.

In any case, all those who’d gain or lose from the proposal would be tempted to exaggerate their conditional declared value differences from the status quo. This is because such exaggerations might influence the conditional tax revenue asset prices that would be used to decide if the proposal is approved. They’ want to exaggerate a win to help make the win happen, or exaggerate a loss to prevent it from happening.

Owners might also be tempted to trade directly in the speculative markets on conditional tax revenue assets, to distort such prices in order to achieve the same result. These efforts can be seen as attempts to “manipulate” such markets, via trades intended to change prices in ways that favorably influence decisions, even if those trades lose on average financially. We have a substantial literature on such manipulation; we have papers on game theory proofs, lab experiments, and field experiments. And the bottom line is that when market speculators are free to counteract suspected manipulation attempts, such attempts tend on average to increase the accuracy of market prices.

Yes, noise in tax revenue asset prices can introduce noise into this decision process. But the two facts that 1) tax revenue on large property aggregates must exactly equal the sum of the tax revenue on each contained property, and 2) change-conditional values are by default equal to status-quo-conditional values, should greatly reduce the noise on differences in change versus status quo sums. CHT allow rather precise estimates of the effect of land use rule changes on net property values.

But there’s another game that owners could play: social shaming and other coordinated mistreatment of new property owners. For example, existing owners in an area might try to coordinate to mistreat any new owner if the old owner complained that they actually didn’t want to sell. For example, nearby owners might act coldly to a new resident, or refuse to shop at a new store. Nearby owners might similarly coordinate to mistreat new owners who bought a property conditional on a disliked property use rule change. As such mistreatment could actually change the value that potential property owners put on a property, it could be stably reflected in tax revenue asset prices. For example, by generally mistreating new owners who are complained about by old owners, new owners might be forced to pay a premium to gain old owner approval, and in that case local owners could safely reduce their declared property values by the amount of that premium.

Consider the example of a store asking to cut its parking requirement, to allow for more store space. Nearby residents may fear that this will result in the streets in front of their homes being used for overflow parking during peak shopping periods, such as at Christmas. While this may reduce the conditional value of their home property, given this parking change, honest estimates of those reductions might not be enough to compensate for the gains to the store. So the change could still happen. In this case, residents might try to exaggerate the harm they suffer, by reducing their conditional property values even more, and trying to coordinate to punish any new residents who buy such homes via paying such extra low conditional declared values.

However, a new owner is less likely to be mistreated by nearby owners when many new owners coordinate to buy new properties near each other all at once. And one easy way to coordinate is to buy properties conditional on a rule change. So new owners may seek exaggerated conditional value reductions, in order to both buy new property cheap, and also to coordinate with other new owners to buy in the same area at the same time.

More generally, anyone proposing a rule change to improve their property is probably well advised to package that change with changes that would be attractive to outsiders seeking to buy many properties at once in an area. This would make more credible the threat of outsiders taking many local properties at their conditional declared values. Which should reduce exaggerated expressions of harm in conditional declared property values. And if many change proposals did this, rules would tend to favor making it easier for big outsider buyers.

Now consider the example of a proposal for a new park. As managing the park is probably a money-losing proposition, the park owner must be paid to do the job, via a negative declared property value. So a simple proposal to turn an empty lot into a park could cause a great loss for the owner of that lot. If the CHT system limits the size of such losses, such proposals might typically include cash transfers from nearby owners to the owner of the park property. Even so, since the CHT system will likely allow some losses, one may still not want to be the property chosen for a new park. This fact should induce owners to think about whether a park might be a good idea in their area, and if so to proactively propose to turn someone else’s property into a park.

The fact that this CHT system induces these sort of games does not seem ideal, but even so it still looks promising compared to our status quo systems, which also induce many harsh conflicts and games.

Note that two key factors are likely to increase the spread between declared and actual values, and thus the scope for such strategic games. One factor is transaction costs; higher costs of changing property usage would reduce the value of other potential owners, relative to the current owner. The other factors is the default terms of the transaction specified via a CHT system. If you accept the official declared value as a sale offer, but if you and the property owner don’t like the default terms of that transaction, you’ll be tempted to renegotiate to get better terms. But this will introduce frictions and thus transaction costs. If, however, the law supports good transaction defaults and low transaction costs, declared values usually won’t deviate as far from actual values.

GD Star Rating
Tagged as: , ,
Trackback URL: