Property-Tax Assessment via Second Price Auction: Not That Efficient

I recently wrote up an idea for assessing property values via a second price auction, that I called SPATA. It came out of a discussion about another cool tax-assessment scheme (COST) and was meant to solve some of its problems.

I had been surprised that the authors of COST hadn’t considered it, but of course, they had. After sending them a link, Glen Weyl and Anthony Zhang were kind enough to point out the problems with SPATA, which they had considered and dismissed.

The write:

Qualitatively self-assessment and second-price auctions with modified payment rules do the same thing. But under self-assessment, owners announce correct valuations under a tax equal to what they think the probability of sale is. So maybe something in the ballpark of 5–50% a year, depending on what assets are being discussed. Under second-price auctions, ppl bid honestly only if they have no ownership.

Under SPATA, the owner of the property does not actually need to declare their true value, since both the tax assessment and sales happen at the highest non-owner bid. This is actually fine from a tax perspective, since arguably, the commons should only be compensated for the opportunity cost of leasing the property to the highest bidder. However, it results in a less-efficient allocation of property to owners.

I admit that this took me a while to understand, and took walking through some examples to see the problem. Consider this example:

1. Alice owns a tractor, and is able to generate $1K from it.
2. Bob would be able to generate $1.5K from the tractor, and bids $5K on it.
3. Alice must now pay taxes on the $5K bid, say at 10% tax rate, or $500/year.
4. Alice is still making a profit of $500, and will hold on to the tractor.
5. Even though the commons are compensated, $500 worth of value could be generated if Bob had gotten to the tractor first.

Under COST, Alice would need to declare her value for the tractor, and risk losing it with some probability if she set her price too low.

SPATA also fails to solve the problems of hold-outs for property consolidation. The cool thing about COST is that a developer that wanted to build a road, could immediately purchase all the properties for the listed prices, without having to negotiate with each owner (who are all incentivized to hold out for an unreasonably high price)

The nice thing about SPATA is that it required less market-savvy on the part of the asset owner. Under COST, the owner would need to constantly follow the market and adjust their assessment, or risk losing the asset. However, I think we can achieve a similar outcome via insurance.

In thinking about these systems, I also realized that both COST and SPATA fail to address asymmetric transaction costs, which I will write about in a follow up.