I'm afraid "[Northgate] is the most valuable piece of real estate in the whole city, a fact anyone who’s lived there for any amount of time knows intuitively." gives too much credit to speedy intuition. On a second moment's consideration, and I think beyond further doubt, I hazard that the most valuable land in College Station rests under Texas A&M Main Campus.
Sure, the Dixie Chicken's onion rings go well with a reasonably-priced pitcher of yellow beer (though O'Bannon's has greater selection of quality quaffs two blocks away), but a strip of bars sandwiched between apartments and dormitories doesn't scream "high land value", compared to the dirt under the main economic driver of the city. This seems a rather serious bug in the "just ask people their first intuition" process. It ignored the maroon elephant in the room.
Rather, one can take a look at, say, the ENTIRE GOLF COURSE sitting at the (formal) entrance to the university grounds. What would that gluttonous morsel of land sell for, on the open market? It has the surface area of the whole central campus grounds, directly adjacent to it-- one could site an entire second university there, and still leave a few ponds and putting greens. It would even have better road access off the major arteries of Texas & George Bush, rather than fighting through the choke-point on University.
Next, how does the whole LVT system handle the taxation discontinuity of such stonkingly valuable land, once it sits under a tax shelter? Would the Great State of Texas pay itself for the privilege of the land it holds, especially undeveloped? That would make more sense for funding roads and other physical plant needs to spider out into the school's urban cocoon, but state organs have an allergy to paying their own levies.
> Next, how does the whole LVT system handle the taxation discontinuity of such stonkingly valuable land, once it sits under a tax shelter? Would the Great State of Texas pay itself for the privilege of the land it holds, especially undeveloped? That would make more sense for funding roads and other physical plant needs to spider out into the school's urban cocoon, but state organs have an allergy to paying their own levies.
Depends how much you want to (and are able) to revolutionize the system. If you just do some reforms at the edges, probably it stays exempt because the powers that be like it that way. But you're absolutely right to point out that Texas A&M campus is where the most valuable land is!
And you know what? This is exactly the kind of comment I imagine would actually be brought up in one of these meetings--given you and I are both Aggies. The historical record suggests that getting all these contrarian takes out in the open right from the start was the whole point; a key motivation being to move assessment away from the sole opinion of individual self proclaimed experts (such as myself), which was the status quo at the time.
As for this part:
> Would the Great State of Texas pay itself for the privilege of the land it holds, especially undeveloped?
You know if I don't misunderstand you, I think "government paying tax to itself" isn't *such* a crazy idea in principle, even if I agree that I don't expect to see electeds leaping to implement it.
Relevant context --
In Texas property tax is almost exclusively the domain of *local government*--state government doesn't collect any. What the state does do is oversee property tax with the comptroller's annual property value study to ensure quality control. This is b/c the state kicks in funding (from non property tax state revenue) for school districts that are underfunded from local property tax, and they want to make sure local governments aren't intentionally undervaluing and free riding. Local property values are set by the CAD, central appraisal district, which is a fully independent local government that has no authority to either collect taxes or set a tax rate, just set values. The "taxing entities" -- school districts, city and county government, water & utility districts, etc, are the ones who actually set a tax rate and collect taxes (but they have no power to set values). So there's already a good degree of tax flow between government bodies within Texas' borders already, between the three nodes of state, local, and individual taxpayer.
So the answer to "what would happen if the universities lost their tax exempt status" is, following the current system, is the local taxing entities would get that tax money. So rather than paying taxes to itself, the state of texas would be paying money out to local governments. At least under the current system.
But back to the main point -- you're absolutely right that Texas A&M University is the most valuable real estate in town. The tricky bit simply being that it's not for sale and doesn't have any recent transaction price.
But since you opened the door, follow me down the rabbit hole a little further. Your comment now makes me wonder to what degree you could use something like community consensus to crack the so-called "Disneyworld problem" -- ie, how do you value gigantic, "impossible-to-value" pieces of land embodied as one giant contiguous single parcel, owned by a singular entity. IE, Disneyworld. Or certain land-grant Texas Universities.
Texas A&M campus has "streets" and "blocks". Certain parts of it are more highly trafficked and more valuable than others, for various definitions of valuable. Presumably there are people who have various degrees of knowledge and opinion about what parts of the university are more valuable than others, and why.
Even as an internal accounting exercise, it seems that at least a *relative value map* of the most and least valuable locations on campus could be done if you employed enough local knowledge. And certainly if the University were to sell a few bits of its periphery it would readily find willing buyers willing to price their offers for that land in dollars.
The trick is that if the "Disneyworld" in question were to be wholly privatized and sold off, then the chief generator of the awesome land value would to a certain extent go with it -- the land around Disneyworld or Texas A&M is valuable *because Disneyworld/Texas A&M is right there*.
That said, I don't think it's impossible in principle to put a reasonable lower bound value on, say, the stretch of Northside campus just across from Northgate, or perhaps even the Golf Course. Current non-Somers methods of doing this mostly amount to "Go ask a developer how much they would be willing to pay for this and such piece of land if the University was willing to sell it" and then follow the same kind of procedure.
Kind of an undercooked thought but curious what you have to say about it.
Oh right, you're one of the cognoscenti who knows the corporeal form behind the pseudonym. 🥸 Doing quite well, thank you, it has been too long. All my best and prayers for you and the family, as well.
As to these other items, but pittances compared to love and family, yet more readily disposable to the interwebs---- back to land value.
LVT needs three things: [1] a sensible (enough) valuation of land, whose (surplus, un-improved) value can be [2] appropriated in a tax, which [3] establishes a superior levy than the current system. Succeeding in all points accomplishes both better land use, and less tax burden on productive efforts.
The current example (CStat/A&M) poses issues on all three accounts. Which piqued my interest, because my inchoate skepticism (post on topic coming soon! (ish)) stands mostly on the difficulties of problem #2 (an effectively appropriated tax) as pertains to private ownership, rather than #1 or #3, or #2 for public land.
College Station makes a great example, precisely because it's ludicrous in terms of land ownership. Few towns are structured as CStat is, which illuminates edge cases likely to be present elsewhere, in less vivid and thus less visible cases.
I picked on the golf course rather specifically. In days of yore, there wasn't much to College Station, so A&M grew a host of facilities on its own (extensive, lonely) land-- such as a golf course (next to the polo fields, which also held bonfire). A&M even had its own on-campus hotel, shut down about 20 years ago, wrought when the city simply didn't have enough facility with such things. I stayed there myself, once. When long-term campus planning was going on in ~2004, some voices noted that an 18-hole golf course right in the middle of town wasn't a customary best use of such land. For various reasons/politickings, it's still an 18-hole golf course.
Which would make a dandy example of LVT point #3, should it be taxed appropriately: leveraging financial onus to see the finite, unmovable capital of land live its best life, where politics and institutional momentum otherwise impedes a better world. But then, the case also seems to sunder points #1 and #2, being difficult-to-implausible to assay, let alone levy.
---
So it makes a fun exercise on your side of the table, exercising edge cases and unusual conditions to make sure that your land-value processes are robust. That's where I got spicy about Northgate-- not to be contrarian (an attitude to which you can guess, even at many years passing, I am definitely still not averse to)-- but rather to raise the larger issue of how tricky land valuation can be, without test of commercial activity and other skin-in-the-game proof.
In all, I note so much gulf between the world of Somers, and ours. Compact urban areas in the early 1900's wouldn't have faced giant city-sized state campuses in the middle of them; nor would they be expected to fund wild transit systems (which the modern Texas-style concrete jungle certainly is; I presume you're well versed in the stonking financial burden vehicular road systems are), or face our web of squirrelly urban zoning shenanigans. I still think LVT problem#1 is solvable in all reasonable cases, but we haven't made the process simple and obvious.
In practical (?!) and less skeptical terms, I would say the upshot for dealing with A&M's ginormous campus is to establish some version of LVT *internally*. This would resolve various nefarious political wranglings over the proper use of its grounds, and at least illuminate what subsidies are floating its more, ah, esthetic-rather-than-practical land-use choices.
I would definitely agree that quantifying the internal value of on campus locations might be the most valuable use of this system, if somehow the administration could be convinced to prioritize building in the most valuable spots rather than continuing to plop new things randomly around west campus :)
> I would say the upshot for dealing with A&M's ginormous campus is to establish some version of LVT *internally*
I do think an internal land use assessment for the "Disneyworlds" of the world would be interesting, and it feels plausible--I can at least imagine it.
RE: Somers and "then" vs. "now" --- since writing this article I've come to realize the system persisted a lot longer than I expected. I've since found (posted as an edit to the end) that it lasted in Cleveland Ohio until at least 1946. Meanwhile, Scholz outlived Pollock, and went on to contribute to an early edition of the Pennsylvania Assessor's handbook, also from the late 1940's, which in many parts basically reads almost like a complete rehash of Pollock and Scholz's book reviewed in this article, suggesting the system persisted in Pennsylvania into the late 1940's too. There's two more editions of that book that go into the sixties that I'm trying to track down so I can figure out exactly when this system faded away.
A beautifully written post. One minor thing could be the stated population of Saint Paul, MN, in 1896. It looks too high compared to other sources, even if one includes the metropolitan area. Saint Louis, MO, is a city that also experienced a significant decline from its peak, but I don't think it had that population of 1.5M neither.
Thanks Andrew! One thing some folks who read this remarked to me was that they thought the Somers system has a bit of a “strong towns” vibe to it, so I was curious what you, Charles, et al would think of it.
Someone suggested we get in touch with you. We are building eco systems for collective inelligence. Ways to take 10, or 100, or 1 million people and give them a creative voice. There is wisdom in crowds to be extracted and we can use decentralized and high trust transparent systems to extract it. Like this: youtube.com/watch?v=AIF1Rh_JNeE&feature=youtu.be
I don't use whatapp but I also don't have a strong moral affinity for investment speculation.
If there was a way to make money just like that the idea generally does not need sharing, it will share itself until the concept is over traded by high speed AI trading platforms. If it is so good and no one else will figure it out it makes sense to just scale up quietly.
Not having capital to invest makes me uninteresting to someone who lives on commission. Living 10 times zones from the USA also means things work a little differently here.
It all sounds reasonable, but how does it stand up to the determination of a vocal and organized minority to deliberately mess up the valuation process?
It’s a good question and the best answer is to study what happened historically! This thing apparently persisted for some *fifty years* at least, with a multi decade stretch in Cleveland alone. Surely there’s answers in there somewhere. Makes me sad everyone involved is dead.
The surface answer I can give now seems to be to try to include as many people as possible and rely on multiple rounds of well publicized feedback. But a deeper answer needs more understanding of how this actually came into contact with reality.
The issue is that we're not trying to find the market value of the land, we're trying to find the value of the land to the specific person who has the highest market value for the entire property. This is fundamentally impossible and it largely dooms the entire idea of lvt.
It's not impossible to decouple land from improvements. One can always tweak a game algorithm to have as close an approximation to your liking. As long as the incumbent is not disincentivized to stay and the rent of surrounding land is considered, then the LVT can be evaluated (or the evaluation may be slightly lower than LVT, which will be of no harm to the incumbent).
> Land value has sharply diminishing returns to depth
> I need no convincing of this basic principle, I’m just not sure of the exact shape of the curve and how universal it is. I’m also willing to posit some exceptions to the general rule
I take it this is a method that values residences at zero? Something we would only ever apply to parcels that are zoned exclusively for urban retail use?
There are essentially no diminishing returns to depth for residences or for office buildings. Frontage is inherently valuable for retail because it increases exposure to pedestrians. This is not a concern for homes or workplaces.
I do agree that the Somers system is much more optimized for urban land than residential, which I’ve mentioned in other comments. You’re right that frontage doesn’t matter ad much in residential contexts, and Zangerle acknowledges something like that in one of the linked documents at the end. In fact in many residential contexts busy streets can detract from value.
But saying there are NO diminishing returns to depth (or increasing size in general, which it would seem to imply) to any class of land contradicts every dataset I’ve seen. Are you saying land parcels zoned for office and residential increase perfectly linearly in value with size on a $/land sqft basis?
Everything I’ve seen shows they plummet in value on a $/sqft basis even as the total value increases monotonically with size.
If you have some data that contradicts this I’d happily change my mind, and be really surprised!
I think I have to be allowed to say "there are no diminishing returns to depth" to mean that depth in itself has no effect. I'll admit that there is an effect of depth insofar as a mansion that is too large for someone to be willing to cross is getting less value out of each unit of space than a smaller residence would, because of the inconvenience. But to surface an effect, that mansion would need to be occupying something like a full city block.
The system clearly states that road frontage is valuable for its own sake and that, as a consequence, the same area is more valuable if it has more frontage than otherwise. Neither of those claims is true - to any degree - for homes or offices. The depth gradient is a quantification of that phenomenon, and so there is no depth gradient, within the conversational context, for homes or offices. Depth isn't relevant to them.
I'm intrigued to see you say that the unit value of a lot goes down the bigger the lot is. This is definitely true as a matter of sales on the market, in the same way that the unit 𝘱𝘳𝘪𝘤𝘦 of everything goes down the more of it you're willing to buy.
On the other hand, the unit price of everything also goes up the more of it you need to buy. Which way the unit prices moves depends more on the characterization ("need to buy" versus "willing to buy") than the size of the transaction. It's about who's doing a favor for whom, or who's in more of a hurry.
And I don't see how Georgism can survive a perspective that adjusts the value of land according to how much of it there is. The obvious endpoint of that approach is that one guy owns all the land and pays almost no taxes on it.
Hi Michael—thanks so much for taking the time to reply!
> I think I have to be allowed to say "there are no diminishing returns to depth" to mean that depth in itself has no effect.
Absolutely! please express your view in whatever words feel right to you. My aim is simply to engage with the position you actually hold, so any extra clarification you offer is genuinely helpful.
> I'll admit that there is an effect of depth insofar as a mansion that is too large for someone to be willing to cross is getting less value out of each unit of space than a smaller residence would, because of the inconvenience. But to surface an effect, that mansion would need to be occupying something like a full city block.
Great, so we both agree depth does have some influence. If I’m hearing you correctly, you believe this influence is quite modest and only shows up at very large depths (perhaps a whole-block mansion). Is that an accurate reading?
> The system clearly states that road frontage is valuable for its own sake and that, as a consequence, the same area is more valuable if it has more frontage than otherwise. Neither of those claims is true - to any degree - for homes or offices.
Just to be sure I follow: you accept that a parcel needs road access, yet you doubt that additional frontage adds value. Have I captured that correctly?
Happily, that’s a testable claim. As I understand it, the prediction is that, after controlling for other factors, we should never find a statistically significant premium for extra frontage, nor a buyer who says, “I paid more because it had more frontage,” so long as we're talking about homes and offices. If that’s off the mark, please let me know. And if you have data supporting the claim, I’ll be glad to revisit my view.
You also wrote “to any degree.” Am I right that you mean this applies universally: no frontage effect for homes or offices, in any market, anywhere?
> I'm intrigued to see you say that the unit value of a lot goes down the bigger the lot is. This is definitely true as a matter of sales on the market, in the same way that the unit 𝘱𝘳𝘪𝘤𝘦 of everything goes down the more of it you're willing to buy.
Exactly. That diminishing-marginal-acre effect appears in virtually every assessor’s manual and in every dataset I’ve analyzed. It’s one of the first patterns a hedonic regression uncovers, it's very well attested by pretty much everyone in the field.
> And I don't see how Georgism can survive a perspective that adjusts the value of land according to how much of it there is. The obvious endpoint of that approach is that one guy owns all the land and pays almost no taxes on it.
That, too, is a clear hypothesis. Are you imagining a situation where someone aggregates essentially all land (could we define “all” more precisely?) solely to reduce a land-value-tax bill? If so, what counterbalances the well-documented incentive to subdivide high-demand parcels and sell them for profit? Conventional property taxes already appraise large undeveloped tracts at market value, shouldn’t we observe the same roll-up-everything behavior today if the logic holds? I’m curious why a land-value tax would be uniquely vulnerable while a standard property tax is not.
Thanks again, Michael. I appreciate the thoughtful exchange and look forward to any clarifications or evidence you’d like to share!
> Great, so we both agree depth does have some influence. If I’m hearing you correctly, you believe this influence is quite modest and only shows up at very large depths (perhaps a whole-block mansion). Is that an accurate reading?
No, I don't endorse that phrasing. The influence that I grant to depth here applies equally to all distance. Depth is a kind of distance. Walking a long way along the street is just as bad as walking a long way perpendicular to the street. I think it's a mistake to attribute any of this to "depth". (But yes, I am saying this influence is modest, or perhaps more accurately it makes itself felt more in the sizes of homes that people build than in pricing differences between homes.)
> As I understand it, the prediction is that, after controlling for other factors, we should never find a statistically significant premium for extra frontage, nor a buyer who says, “I paid more because it had more frontage,” so long as we're talking about homes and offices. If that’s off the mark, please let me know.
That is a good reading of what I said, yes.
> That diminishing-marginal-acre effect appears in virtually every assessor’s manual and in every dataset I’ve analyzed. It’s one of the first patterns a hedonic regression uncovers, it's very well attested by pretty much everyone in the field.
You seem to have glossed over the material I followed that with. If you're selling land you get diminishing returns from selling more of it at once. If you're buying land, you get diminishing returns -- on your dollars -- from buying more of it at once. But diminishing returns on the buyer's dollars mean increasing returns, to the sellers, on the additional land.
This phenomenon is, for example, what eminent domain is intended to preclude.
> Are you imagining a situation where someone aggregates essentially all land (could we define “all” more precisely?) solely to reduce a land-value-tax bill? If so, what counterbalances the well-documented incentive to subdivide high-demand parcels and sell them for profit?
I'm saying that this structure provides a constant incentive for large landowners to expand their landholdings and for small landowners to divest theirs. Over time, concentration will increase further and further.
What counterbalances the incentive to divide your land and sell part of it is that this is economically inefficient. One guy renting out space in each of his 40 apartment buildings has lower total costs than 40 guys renting out space in one of those buildings each. One guy growing garlic on 400,000 acres of land has lower total costs than 10 guys growing garlic on 40,000 of those acres of land. You expect to see the land put to its most productive use, and it's more productive when it's all owned by the same guy.
> Conventional property taxes already appraise large undeveloped tracts at market value, shouldn’t we observe the same roll-up-everything behavior today if the logic holds?
Yeah, that's a good question. I don't know anything about the actual workings of this market. I'd be interested to know more.
I have a vague impression that landownership in England is in fact extremely concentrated, and that land is not really for sale on the market. You buy a house, you own the house, but you still have to pay rent to the lord who owns the land under it. If this is true, it might suggest that these incentives used to exist and the landownership structure is maintained through inertia, or it might suggest that rolling up the land is a lengthy process and England is an older country than the United States. (A quick search turned up this article: https://www.theguardian.com/money/2019/apr/17/who-owns-england-thousand-secret-landowners-author )
A good thing to check wrt to those two ideas would be "what does landownership look like in other nonrevolutionary European countries?"
> I’m curious why a land-value tax would be uniquely vulnerable while a standard property tax is not.
I stand behind that; a land-value tax is uniquely vulnerable to rolling-up because it taxes people for what their neighbors have. As a landowner, you can make that problem go away by absorbing your neighbors. The state will pay you to do it.
Consider a hotel next to a casino. These have so much synergy that casinos often themselves operate as hotels, but this casino doesn't. The land under the hotel has a high value because of proximity to the casino. Even if there wasn't already a hotel there, you could build one, and it would make a lot of money.
If the hotel's landlord gets tired of paying property taxes and buys out the land under the casino, the land under the hotel is no longer next to a casino - that's a development - and his total property tax bill will fall pretty dramatically. Or he could sell the land to the casino's landlord with much the same effect.
In a standard property tax system, you still collect taxes on the casino regardless of who owns it. In a land value tax system, you collect taxes on the casino (from its neighbors) as long as they don't also own the casino itself. But once everyone unites under a single owner, you no longer tax the casino at all. I claim that this is a much larger incentive toward rolling up than the property tax is.
Cool! Thanks for clarifying your position and the predictions made by your hypotheses. If your expected outcomes are found not to happen as you predict, will you change your mind?
To some degree. Are you referring to anything other than this?
> As I understand it, the prediction is that, after controlling for other factors, we should never find a statistically significant premium for extra frontage, nor a buyer who says, “I paid more because it had more frontage,” so long as we're talking about homes and offices.
I can provide a bit of elaboration on the kind of thing I'd expect:
- In an area with foot traffic, frontage is valuable because you could use it for retail. The classic approach to this is to locate retail on the street and other uses in upper stories of the same building. What you control for matters; if you control for "is the property a residence", a residence that occupies a bunch of valuable frontage will have to pay for it whether they want it or not. (A residence that occupies a bunch of worthless frontage, not so much.) Residences like that should be very rare, but the statistical control will hide that fact.
If you instead control for "how much of the property is frontage", I would expect that properties that are slanted toward frontage should be much less likely to be residences or offices, and properties that are slanted toward depth should be much less likely to be retail.
- Analogously, a buyer might acknowledge that they had to pay more for a house because of its abundant frontage - though I'm having trouble imagining this happening in reality - but I wouldn't expect them to say that the abundant frontage was what made them willing to pay more. If a building already exists, then the facts about it are hard to change. (I do have an example to hand from the other direction. I occasionally visit a shop that has zero frontage. It is located in what was formerly (and partially remains) a factory/office complex; I guess the demand for offices dried up. That shop has to post a lot of signs in more accessible locations (and within its own complex!) advertising where you can find it.)
The class of lot I think would challenge the depth rule is the case where you have an independent house sited down a driveway that gets it behind a street-facing house.
So, imagine you have a 50-foot wide, 100-foot deep lot, and you split it in half, building a home on the front and back halves. Then you carve a ten foot wide driveway off the front lot. Maybe you put a garage for the front house along that driveway as well.
I suspect that despite having a smaller "exclusive" footprint, the front parcel will be worth a bit more than the back one. But not nearly to the extent the Somers table suggests.
I think the truth will be closer to something like, value the two parcels together using the gradient rule; then split that value something like 55/45 between the front and back lots.
Skimming through the book, around page 101 they _kinda_ get at this, with the discussion of lots A and B on the diagram between the imaginary 3rd and 4th Streets... But I'm not seeing anything quite like the form of lot split I'm describing (which I've seen fairly often), where the "back house" is on a flag lot that only fronts the street via a narrow access. (Interestingly, the concept of the "flag lot" seems to be pretty widespread in modern assessment and real estate discourse, but a search in the book doesn't seem to find the word "flag" at all, or at least not that Google's OCR finds.)
Both conducted under John Zengerle in Cleveland, who had tenure throughout the entire period. The 1946 report includes a breakdown of their methodology, which includes a section on "back lots" and alleys -- I'm curious what you think of those and to what degree they address your issue or not.
Well, that Zangerle document has an example of a flag lot of a sort (lot B on page 8), but the "pole" on it is pretty thick (25'). They basically just apply the gradient -- they count the full-depth part of the lot the normal way, and then take the part of B that's behind A, and just subtract the value of A from the full-depth version of that slice of A+B.
I am fairly confident that, whatever the combined value would be of A+B, just running the gradient across like this is crediting too much of the value to A. As long as you have a plot that is practical for some use -- to live on, to conduct business on -- there is some base amount of value to simply having a claim on that plot. The closer you get to a classic flag lot, where the part of the flag lot that runs out to the street is just a driveway next to the front lot, the worse the purely-gradient-based estimate will get.
Additionally -- in residential areas I feel like street frontage can also just matter less in general as long as you have *any* access to the road network and city services. In fact in many places I've noticed that *lack* of street proximity is associated with higher value, the opposite of the traditional Somers assumptions (people like quiet streets and culdesacs, etc). There's some nods from Zangerle on this -- he notes that residential lots don't really benefit from corners -- but otherwise he sticks pretty closely to an Orthodox Somers approach.
This suggests to me the Somers System--taken exactly as described--is most adapted to urban contexts, especially commercial ones, where lots tend to be small, space is at a premium, and road frontage a primary concern. The conventional LEA approach might be better adapted to residential neighborhoods where the lots are all pretty much the same size anyways.
What would you suggest as the best approach for flag lots, in or out of a Somers system context?
Just to be clear I think the Somers system is pretty cool, especially given how easy it is to code this kind of gradient logic with modern computing power.
I think we need some kind of modification that assumes some kind of base value to having the location at all. Like, figure out what the highest-value use is that the plot is feasibly usable for (even if the "implementation" of that use is a "minimum viable product" version of the use), and assign a base value to each lot based on that, and then add onto that value based on the Somers gradient.
So in the case of the flag lot, the fact that you could hypothetically develop (say) a three-story triplex on each of the two lots, puts a floor under the value the lot could possibly have, given how we think six families would value just "living on this block", regardless of whether they're directly on the street. So each lots gets an assignment of $300k for that (or whatever). If the total Somers value for the frontage ought to be $1M, we take $600k of that out for that "base value", and then assign the remaining $400k proportional to how the gradient "shades" each sqft of each lot.
In this case we're assuming that there is an absolute minimum $100k value to a unit you can live in to get a toehold in this location.
This does complicate things, in terms of valuing things like "the smallest possible living unit people will tolerate to live here", and "the smallest possible retail outlet", and so on. Maybe we just don't care enough about this to try to complicate things like that, and could do some kind of adhoc adjustment to flag lots.
If I were going to really work on this (which I don't have the time for) I'd probably start by trying to find sales data where the front and back lots in a pair like this both sold near in time, especially if there's a lot nearby of similar size to the combination of the front and back. If you could collect enough examples of that, it would help provide some points of comparison to get the logic right.
Thinking about it a little more, I wonder as well, whether you could perhaps establish this kind of minimum-viable-unit value in the course of public meetings, as well? Like, have people establish for retail zones, what do they think the value would be to just being able to conduct business here _at all_? Like if you could own just a little hole-in-the-wall operation, a hundred square feet or something? And for residential zones, what would you pay for a postage stamp studio apartment?
Yeah these are all great points! And I wonder how much considerations like these came up in actual Somers proceedings? The historical record seems to suggest that the Manufacturers Appraisal Company had an "Orthodox Somers" method they tried to push everywhere, which *immediately* sprouted all kinds of local warts and hairs the minute it was actually put into practice in a particular location (one of the most common insistences being to employ a locally calibrated depth curve).
The "base lot" value does make sense and jives with practice I've seen -- a lot of residential neighborhoods are valued "by lot" where the presumption is that demand for land (and actual variation in lot size) is pretty inelastic. People just want "a nice house," and plus or minus this much yard within some tolerance isn't really entering the pricing picture at all.
Also the canonical example in actual mass appraisal that comes up all the time for postage-stamp lots is townhomes/rowhomes.
Also don't get me started on valuing condos, who don't technically have their own land attached to the parcel at all, and the parcel is just one of 10 different "building only" parcels stacked vertically on top of each other on the exact same location (presumably you would model them as one combined building, value the land they collectively share, then pro-rate the land back out but... argh. Edge cases upon edge cases! :)
The typical (and in fact overwhelming majority of) Chinese housing has no street frontage. What faces the street is a big wall with a gate and a security guard. Inside that you have maybe a couple dozen apartment buildings. (And a public area for the residents, which might be huge and luxurious or small and bare-bones, depending on the price level of the complex.) All of those couple dozen buildings share the same driveway, for residents with cars.
I already left an independent comment about it, but I'm struggling to see why there would be a depth gradient for residences at all. It obviously isn't the case that rooms in your house located far from the street are less valuable or otherwise less desirable than rooms located near the street. They tend to be more desirable because of the reduced noise.
So I'm definitely open to the idea that the Chinese market is just different and might play out with different dynamics; also I replied to your other comment.
That said, the main thing to understand with depth is that ultimately the depth curve is just one of many ways to capture how land prices change with scale. I've already registered my skepticism that land value varies perfectly linearly with frontage as the Somers system assumes, as well as my skepticism that there's one universal depth curve.
The main observation that I am confident in is that, keeping location fixed, land tends to get more expensive as it gets bigger, but each *marginal* square foot of land gets cheaper as the lot gets bigger. I've seen this pattern repeat over and over again, and every assessor I've ever talked to agrees. If it's true that in China that residential land prices just keep increasing linearly as you make the back yard deeper and deeper, I'm perfectly happy to admit to that if you show me the evidence, I would just find it really surprising!
I'm afraid "[Northgate] is the most valuable piece of real estate in the whole city, a fact anyone who’s lived there for any amount of time knows intuitively." gives too much credit to speedy intuition. On a second moment's consideration, and I think beyond further doubt, I hazard that the most valuable land in College Station rests under Texas A&M Main Campus.
Sure, the Dixie Chicken's onion rings go well with a reasonably-priced pitcher of yellow beer (though O'Bannon's has greater selection of quality quaffs two blocks away), but a strip of bars sandwiched between apartments and dormitories doesn't scream "high land value", compared to the dirt under the main economic driver of the city. This seems a rather serious bug in the "just ask people their first intuition" process. It ignored the maroon elephant in the room.
Rather, one can take a look at, say, the ENTIRE GOLF COURSE sitting at the (formal) entrance to the university grounds. What would that gluttonous morsel of land sell for, on the open market? It has the surface area of the whole central campus grounds, directly adjacent to it-- one could site an entire second university there, and still leave a few ponds and putting greens. It would even have better road access off the major arteries of Texas & George Bush, rather than fighting through the choke-point on University.
Next, how does the whole LVT system handle the taxation discontinuity of such stonkingly valuable land, once it sits under a tax shelter? Would the Great State of Texas pay itself for the privilege of the land it holds, especially undeveloped? That would make more sense for funding roads and other physical plant needs to spider out into the school's urban cocoon, but state organs have an allergy to paying their own levies.
Hey there, how you doing :) Long time no see.
> Next, how does the whole LVT system handle the taxation discontinuity of such stonkingly valuable land, once it sits under a tax shelter? Would the Great State of Texas pay itself for the privilege of the land it holds, especially undeveloped? That would make more sense for funding roads and other physical plant needs to spider out into the school's urban cocoon, but state organs have an allergy to paying their own levies.
Depends how much you want to (and are able) to revolutionize the system. If you just do some reforms at the edges, probably it stays exempt because the powers that be like it that way. But you're absolutely right to point out that Texas A&M campus is where the most valuable land is!
And you know what? This is exactly the kind of comment I imagine would actually be brought up in one of these meetings--given you and I are both Aggies. The historical record suggests that getting all these contrarian takes out in the open right from the start was the whole point; a key motivation being to move assessment away from the sole opinion of individual self proclaimed experts (such as myself), which was the status quo at the time.
As for this part:
> Would the Great State of Texas pay itself for the privilege of the land it holds, especially undeveloped?
You know if I don't misunderstand you, I think "government paying tax to itself" isn't *such* a crazy idea in principle, even if I agree that I don't expect to see electeds leaping to implement it.
Relevant context --
In Texas property tax is almost exclusively the domain of *local government*--state government doesn't collect any. What the state does do is oversee property tax with the comptroller's annual property value study to ensure quality control. This is b/c the state kicks in funding (from non property tax state revenue) for school districts that are underfunded from local property tax, and they want to make sure local governments aren't intentionally undervaluing and free riding. Local property values are set by the CAD, central appraisal district, which is a fully independent local government that has no authority to either collect taxes or set a tax rate, just set values. The "taxing entities" -- school districts, city and county government, water & utility districts, etc, are the ones who actually set a tax rate and collect taxes (but they have no power to set values). So there's already a good degree of tax flow between government bodies within Texas' borders already, between the three nodes of state, local, and individual taxpayer.
So the answer to "what would happen if the universities lost their tax exempt status" is, following the current system, is the local taxing entities would get that tax money. So rather than paying taxes to itself, the state of texas would be paying money out to local governments. At least under the current system.
But back to the main point -- you're absolutely right that Texas A&M University is the most valuable real estate in town. The tricky bit simply being that it's not for sale and doesn't have any recent transaction price.
But since you opened the door, follow me down the rabbit hole a little further. Your comment now makes me wonder to what degree you could use something like community consensus to crack the so-called "Disneyworld problem" -- ie, how do you value gigantic, "impossible-to-value" pieces of land embodied as one giant contiguous single parcel, owned by a singular entity. IE, Disneyworld. Or certain land-grant Texas Universities.
Texas A&M campus has "streets" and "blocks". Certain parts of it are more highly trafficked and more valuable than others, for various definitions of valuable. Presumably there are people who have various degrees of knowledge and opinion about what parts of the university are more valuable than others, and why.
Even as an internal accounting exercise, it seems that at least a *relative value map* of the most and least valuable locations on campus could be done if you employed enough local knowledge. And certainly if the University were to sell a few bits of its periphery it would readily find willing buyers willing to price their offers for that land in dollars.
The trick is that if the "Disneyworld" in question were to be wholly privatized and sold off, then the chief generator of the awesome land value would to a certain extent go with it -- the land around Disneyworld or Texas A&M is valuable *because Disneyworld/Texas A&M is right there*.
That said, I don't think it's impossible in principle to put a reasonable lower bound value on, say, the stretch of Northside campus just across from Northgate, or perhaps even the Golf Course. Current non-Somers methods of doing this mostly amount to "Go ask a developer how much they would be willing to pay for this and such piece of land if the University was willing to sell it" and then follow the same kind of procedure.
Kind of an undercooked thought but curious what you have to say about it.
Oh right, you're one of the cognoscenti who knows the corporeal form behind the pseudonym. 🥸 Doing quite well, thank you, it has been too long. All my best and prayers for you and the family, as well.
As to these other items, but pittances compared to love and family, yet more readily disposable to the interwebs---- back to land value.
LVT needs three things: [1] a sensible (enough) valuation of land, whose (surplus, un-improved) value can be [2] appropriated in a tax, which [3] establishes a superior levy than the current system. Succeeding in all points accomplishes both better land use, and less tax burden on productive efforts.
The current example (CStat/A&M) poses issues on all three accounts. Which piqued my interest, because my inchoate skepticism (post on topic coming soon! (ish)) stands mostly on the difficulties of problem #2 (an effectively appropriated tax) as pertains to private ownership, rather than #1 or #3, or #2 for public land.
College Station makes a great example, precisely because it's ludicrous in terms of land ownership. Few towns are structured as CStat is, which illuminates edge cases likely to be present elsewhere, in less vivid and thus less visible cases.
I picked on the golf course rather specifically. In days of yore, there wasn't much to College Station, so A&M grew a host of facilities on its own (extensive, lonely) land-- such as a golf course (next to the polo fields, which also held bonfire). A&M even had its own on-campus hotel, shut down about 20 years ago, wrought when the city simply didn't have enough facility with such things. I stayed there myself, once. When long-term campus planning was going on in ~2004, some voices noted that an 18-hole golf course right in the middle of town wasn't a customary best use of such land. For various reasons/politickings, it's still an 18-hole golf course.
Which would make a dandy example of LVT point #3, should it be taxed appropriately: leveraging financial onus to see the finite, unmovable capital of land live its best life, where politics and institutional momentum otherwise impedes a better world. But then, the case also seems to sunder points #1 and #2, being difficult-to-implausible to assay, let alone levy.
---
So it makes a fun exercise on your side of the table, exercising edge cases and unusual conditions to make sure that your land-value processes are robust. That's where I got spicy about Northgate-- not to be contrarian (an attitude to which you can guess, even at many years passing, I am definitely still not averse to)-- but rather to raise the larger issue of how tricky land valuation can be, without test of commercial activity and other skin-in-the-game proof.
In all, I note so much gulf between the world of Somers, and ours. Compact urban areas in the early 1900's wouldn't have faced giant city-sized state campuses in the middle of them; nor would they be expected to fund wild transit systems (which the modern Texas-style concrete jungle certainly is; I presume you're well versed in the stonking financial burden vehicular road systems are), or face our web of squirrelly urban zoning shenanigans. I still think LVT problem#1 is solvable in all reasonable cases, but we haven't made the process simple and obvious.
In practical (?!) and less skeptical terms, I would say the upshot for dealing with A&M's ginormous campus is to establish some version of LVT *internally*. This would resolve various nefarious political wranglings over the proper use of its grounds, and at least illuminate what subsidies are floating its more, ah, esthetic-rather-than-practical land-use choices.
I would definitely agree that quantifying the internal value of on campus locations might be the most valuable use of this system, if somehow the administration could be convinced to prioritize building in the most valuable spots rather than continuing to plop new things randomly around west campus :)
> I would say the upshot for dealing with A&M's ginormous campus is to establish some version of LVT *internally*
I do think an internal land use assessment for the "Disneyworlds" of the world would be interesting, and it feels plausible--I can at least imagine it.
RE: Somers and "then" vs. "now" --- since writing this article I've come to realize the system persisted a lot longer than I expected. I've since found (posted as an edit to the end) that it lasted in Cleveland Ohio until at least 1946. Meanwhile, Scholz outlived Pollock, and went on to contribute to an early edition of the Pennsylvania Assessor's handbook, also from the late 1940's, which in many parts basically reads almost like a complete rehash of Pollock and Scholz's book reviewed in this article, suggesting the system persisted in Pennsylvania into the late 1940's too. There's two more editions of that book that go into the sixties that I'm trying to track down so I can figure out exactly when this system faded away.
Strangename Thanks for your comment and support.
Today I’ll be introducing to my subscribers how to start making passive income from investment /business.
I’ll be using this opportunity to lntroduce you to Mr.Taylor my personal Broker for a Profitable and Life-Changing lnvestment.
WhatsApp; +1 (585) 623 0133
Reach out to him now for guidance☝️
Tell him I sent you.
A beautifully written post. One minor thing could be the stated population of Saint Paul, MN, in 1896. It looks too high compared to other sources, even if one includes the metropolitan area. Saint Louis, MO, is a city that also experienced a significant decline from its peak, but I don't think it had that population of 1.5M neither.
You’re right, I think I put an extra zero on there:
https://www.minnesotahistory.org/post/st-paul-a-view-of-the-past-and-bold-predictions-of-the-future
I see what I did now -- the source I was using was giving *state* level figures. Fixed now!
Alfonso Thanks for your comment and support.
Today I’ll be introducing to my subscribers how to start making passive income from investment /business.
I’ll be using this opportunity to lntroduce you to Mr.Taylor my personal Broker for a Profitable and Life-Changing lnvestment.
WhatsApp; +1 (585) 623 0133
Reach out to him now for guidance☝️
Tell him I sent you.
This was really insightful, thanks Lars! I especially like the Sommers Unit, I think there’s something to that.
Thanks Andrew! One thing some folks who read this remarked to me was that they thought the Somers system has a bit of a “strong towns” vibe to it, so I was curious what you, Charles, et al would think of it.
Andrew Thanks for your comment and support.
Today I’ll be introducing to my subscribers how to start making passive income from investment /business.
I’ll be using this opportunity to lntroduce you to Mr.Taylor my personal Broker for a Profitable and Life-Changing lnvestment.
WhatsApp; +1 (585) 623 0133
Reach out to him now for guidance☝️
Tell him I sent you.
awesome. thanks for digging through the history and sharing it.
Max Thanks for your comment and support.
Today I’ll be introducing to my subscribers how to start making passive income from investment /business.
I’ll be using this opportunity to lntroduce you to Mr.Taylor my personal Broker for a Profitable and Life-Changing lnvestment.
WhatsApp; +1 (585) 623 0133
Reach out to him now for guidance☝️
Tell him I sent you.
Someone suggested we get in touch with you. We are building eco systems for collective inelligence. Ways to take 10, or 100, or 1 million people and give them a creative voice. There is wisdom in crowds to be extracted and we can use decentralized and high trust transparent systems to extract it. Like this: youtube.com/watch?v=AIF1Rh_JNeE&feature=youtu.be
The society of problem solvers
Thanks for your comment and support.
Today I’ll be introducing to my subscribers how to start making passive income from investment /business.
I’ll be using this opportunity to lntroduce you to Mr.Taylor my personal Broker for a Profitable and Life-Changing lnvestment.
WhatsApp; +1 (585) 623 0133
Reach out to him now for guidance☝️
Tell him I sent you.
I suggest you get in touch with "The Society of Problem Solvers" and see if they have ways to enable this.
https://joshketry.substack.com/p/humanity-is-under-attack-we-can-build
thank you!
Kalle Thanks for your comment and support.
Today I’ll be introducing to my subscribers how to start making passive income from investment /business.
I’ll be using this opportunity to lntroduce you to Mr.Taylor my personal Broker for a Profitable and Life-Changing lnvestment.
WhatsApp; +1 (585) 623 0133
Reach out to him now for guidance☝️
Tell him I sent you.
I don't use whatapp but I also don't have a strong moral affinity for investment speculation.
If there was a way to make money just like that the idea generally does not need sharing, it will share itself until the concept is over traded by high speed AI trading platforms. If it is so good and no one else will figure it out it makes sense to just scale up quietly.
Not having capital to invest makes me uninteresting to someone who lives on commission. Living 10 times zones from the USA also means things work a little differently here.
It all sounds reasonable, but how does it stand up to the determination of a vocal and organized minority to deliberately mess up the valuation process?
It’s a good question and the best answer is to study what happened historically! This thing apparently persisted for some *fifty years* at least, with a multi decade stretch in Cleveland alone. Surely there’s answers in there somewhere. Makes me sad everyone involved is dead.
The surface answer I can give now seems to be to try to include as many people as possible and rely on multiple rounds of well publicized feedback. But a deeper answer needs more understanding of how this actually came into contact with reality.
Jack Bluman Thanks for your comment and support.
Today I’ll be introducing to my subscribers how to start making passive income from investment /business.
I’ll be using this opportunity to lntroduce you to Mr.Taylor my personal Broker for a Profitable and Life-Changing lnvestment.
WhatsApp; +1 (585) 623 0133
Reach out to him now for guidance☝️
Tell him I sent you.
The issue is that we're not trying to find the market value of the land, we're trying to find the value of the land to the specific person who has the highest market value for the entire property. This is fundamentally impossible and it largely dooms the entire idea of lvt.
https://medium.com/@clayshentrup/the-fatal-flaw-in-land-value-taxation-why-henry-georges-dream-may-be-impossible-68fc0fc3a264
ok
It's not impossible to decouple land from improvements. One can always tweak a game algorithm to have as close an approximation to your liking. As long as the incumbent is not disincentivized to stay and the rent of surrounding land is considered, then the LVT can be evaluated (or the evaluation may be slightly lower than LVT, which will be of no harm to the incumbent).
https://progressandpovertyinstitute.org/decoupling-land-and-improvement-values/
Mihali Thanks for your comment and support.
Today I’ll be introducing to my subscribers how to start making passive income from investment /business.
I’ll be using this opportunity to lntroduce you to Mr.Taylor my personal Broker for a Profitable and Life-Changing lnvestment.
WhatsApp; +1 (585) 623 0133
Reach out to him now for guidance☝️
Tell him I sent you.
Clay Thanks for your comment and support.
Today I’ll be introducing to my subscribers how to start making passive income from investment /business.
I’ll be using this opportunity to lntroduce you to Mr.Taylor my personal Broker for a Profitable and Life-Changing lnvestment.
WhatsApp; +1 (585) 623 0133
Reach out to him now for guidance☝️
Tell him I sent you.
> Land value has sharply diminishing returns to depth
> I need no convincing of this basic principle, I’m just not sure of the exact shape of the curve and how universal it is. I’m also willing to posit some exceptions to the general rule
I take it this is a method that values residences at zero? Something we would only ever apply to parcels that are zoned exclusively for urban retail use?
There are essentially no diminishing returns to depth for residences or for office buildings. Frontage is inherently valuable for retail because it increases exposure to pedestrians. This is not a concern for homes or workplaces.
I do agree that the Somers system is much more optimized for urban land than residential, which I’ve mentioned in other comments. You’re right that frontage doesn’t matter ad much in residential contexts, and Zangerle acknowledges something like that in one of the linked documents at the end. In fact in many residential contexts busy streets can detract from value.
But saying there are NO diminishing returns to depth (or increasing size in general, which it would seem to imply) to any class of land contradicts every dataset I’ve seen. Are you saying land parcels zoned for office and residential increase perfectly linearly in value with size on a $/land sqft basis?
Everything I’ve seen shows they plummet in value on a $/sqft basis even as the total value increases monotonically with size.
If you have some data that contradicts this I’d happily change my mind, and be really surprised!
I think I have to be allowed to say "there are no diminishing returns to depth" to mean that depth in itself has no effect. I'll admit that there is an effect of depth insofar as a mansion that is too large for someone to be willing to cross is getting less value out of each unit of space than a smaller residence would, because of the inconvenience. But to surface an effect, that mansion would need to be occupying something like a full city block.
The system clearly states that road frontage is valuable for its own sake and that, as a consequence, the same area is more valuable if it has more frontage than otherwise. Neither of those claims is true - to any degree - for homes or offices. The depth gradient is a quantification of that phenomenon, and so there is no depth gradient, within the conversational context, for homes or offices. Depth isn't relevant to them.
I'm intrigued to see you say that the unit value of a lot goes down the bigger the lot is. This is definitely true as a matter of sales on the market, in the same way that the unit 𝘱𝘳𝘪𝘤𝘦 of everything goes down the more of it you're willing to buy.
On the other hand, the unit price of everything also goes up the more of it you need to buy. Which way the unit prices moves depends more on the characterization ("need to buy" versus "willing to buy") than the size of the transaction. It's about who's doing a favor for whom, or who's in more of a hurry.
And I don't see how Georgism can survive a perspective that adjusts the value of land according to how much of it there is. The obvious endpoint of that approach is that one guy owns all the land and pays almost no taxes on it.
Hi Michael—thanks so much for taking the time to reply!
> I think I have to be allowed to say "there are no diminishing returns to depth" to mean that depth in itself has no effect.
Absolutely! please express your view in whatever words feel right to you. My aim is simply to engage with the position you actually hold, so any extra clarification you offer is genuinely helpful.
> I'll admit that there is an effect of depth insofar as a mansion that is too large for someone to be willing to cross is getting less value out of each unit of space than a smaller residence would, because of the inconvenience. But to surface an effect, that mansion would need to be occupying something like a full city block.
Great, so we both agree depth does have some influence. If I’m hearing you correctly, you believe this influence is quite modest and only shows up at very large depths (perhaps a whole-block mansion). Is that an accurate reading?
> The system clearly states that road frontage is valuable for its own sake and that, as a consequence, the same area is more valuable if it has more frontage than otherwise. Neither of those claims is true - to any degree - for homes or offices.
Just to be sure I follow: you accept that a parcel needs road access, yet you doubt that additional frontage adds value. Have I captured that correctly?
Happily, that’s a testable claim. As I understand it, the prediction is that, after controlling for other factors, we should never find a statistically significant premium for extra frontage, nor a buyer who says, “I paid more because it had more frontage,” so long as we're talking about homes and offices. If that’s off the mark, please let me know. And if you have data supporting the claim, I’ll be glad to revisit my view.
You also wrote “to any degree.” Am I right that you mean this applies universally: no frontage effect for homes or offices, in any market, anywhere?
> I'm intrigued to see you say that the unit value of a lot goes down the bigger the lot is. This is definitely true as a matter of sales on the market, in the same way that the unit 𝘱𝘳𝘪𝘤𝘦 of everything goes down the more of it you're willing to buy.
Exactly. That diminishing-marginal-acre effect appears in virtually every assessor’s manual and in every dataset I’ve analyzed. It’s one of the first patterns a hedonic regression uncovers, it's very well attested by pretty much everyone in the field.
> And I don't see how Georgism can survive a perspective that adjusts the value of land according to how much of it there is. The obvious endpoint of that approach is that one guy owns all the land and pays almost no taxes on it.
That, too, is a clear hypothesis. Are you imagining a situation where someone aggregates essentially all land (could we define “all” more precisely?) solely to reduce a land-value-tax bill? If so, what counterbalances the well-documented incentive to subdivide high-demand parcels and sell them for profit? Conventional property taxes already appraise large undeveloped tracts at market value, shouldn’t we observe the same roll-up-everything behavior today if the logic holds? I’m curious why a land-value tax would be uniquely vulnerable while a standard property tax is not.
Thanks again, Michael. I appreciate the thoughtful exchange and look forward to any clarifications or evidence you’d like to share!
> Great, so we both agree depth does have some influence. If I’m hearing you correctly, you believe this influence is quite modest and only shows up at very large depths (perhaps a whole-block mansion). Is that an accurate reading?
No, I don't endorse that phrasing. The influence that I grant to depth here applies equally to all distance. Depth is a kind of distance. Walking a long way along the street is just as bad as walking a long way perpendicular to the street. I think it's a mistake to attribute any of this to "depth". (But yes, I am saying this influence is modest, or perhaps more accurately it makes itself felt more in the sizes of homes that people build than in pricing differences between homes.)
> As I understand it, the prediction is that, after controlling for other factors, we should never find a statistically significant premium for extra frontage, nor a buyer who says, “I paid more because it had more frontage,” so long as we're talking about homes and offices. If that’s off the mark, please let me know.
That is a good reading of what I said, yes.
> That diminishing-marginal-acre effect appears in virtually every assessor’s manual and in every dataset I’ve analyzed. It’s one of the first patterns a hedonic regression uncovers, it's very well attested by pretty much everyone in the field.
You seem to have glossed over the material I followed that with. If you're selling land you get diminishing returns from selling more of it at once. If you're buying land, you get diminishing returns -- on your dollars -- from buying more of it at once. But diminishing returns on the buyer's dollars mean increasing returns, to the sellers, on the additional land.
This phenomenon is, for example, what eminent domain is intended to preclude.
> Are you imagining a situation where someone aggregates essentially all land (could we define “all” more precisely?) solely to reduce a land-value-tax bill? If so, what counterbalances the well-documented incentive to subdivide high-demand parcels and sell them for profit?
I'm saying that this structure provides a constant incentive for large landowners to expand their landholdings and for small landowners to divest theirs. Over time, concentration will increase further and further.
What counterbalances the incentive to divide your land and sell part of it is that this is economically inefficient. One guy renting out space in each of his 40 apartment buildings has lower total costs than 40 guys renting out space in one of those buildings each. One guy growing garlic on 400,000 acres of land has lower total costs than 10 guys growing garlic on 40,000 of those acres of land. You expect to see the land put to its most productive use, and it's more productive when it's all owned by the same guy.
> Conventional property taxes already appraise large undeveloped tracts at market value, shouldn’t we observe the same roll-up-everything behavior today if the logic holds?
Yeah, that's a good question. I don't know anything about the actual workings of this market. I'd be interested to know more.
I have a vague impression that landownership in England is in fact extremely concentrated, and that land is not really for sale on the market. You buy a house, you own the house, but you still have to pay rent to the lord who owns the land under it. If this is true, it might suggest that these incentives used to exist and the landownership structure is maintained through inertia, or it might suggest that rolling up the land is a lengthy process and England is an older country than the United States. (A quick search turned up this article: https://www.theguardian.com/money/2019/apr/17/who-owns-england-thousand-secret-landowners-author )
A good thing to check wrt to those two ideas would be "what does landownership look like in other nonrevolutionary European countries?"
> I’m curious why a land-value tax would be uniquely vulnerable while a standard property tax is not.
I stand behind that; a land-value tax is uniquely vulnerable to rolling-up because it taxes people for what their neighbors have. As a landowner, you can make that problem go away by absorbing your neighbors. The state will pay you to do it.
Consider a hotel next to a casino. These have so much synergy that casinos often themselves operate as hotels, but this casino doesn't. The land under the hotel has a high value because of proximity to the casino. Even if there wasn't already a hotel there, you could build one, and it would make a lot of money.
If the hotel's landlord gets tired of paying property taxes and buys out the land under the casino, the land under the hotel is no longer next to a casino - that's a development - and his total property tax bill will fall pretty dramatically. Or he could sell the land to the casino's landlord with much the same effect.
In a standard property tax system, you still collect taxes on the casino regardless of who owns it. In a land value tax system, you collect taxes on the casino (from its neighbors) as long as they don't also own the casino itself. But once everyone unites under a single owner, you no longer tax the casino at all. I claim that this is a much larger incentive toward rolling up than the property tax is.
Cool! Thanks for clarifying your position and the predictions made by your hypotheses. If your expected outcomes are found not to happen as you predict, will you change your mind?
To some degree. Are you referring to anything other than this?
> As I understand it, the prediction is that, after controlling for other factors, we should never find a statistically significant premium for extra frontage, nor a buyer who says, “I paid more because it had more frontage,” so long as we're talking about homes and offices.
I can provide a bit of elaboration on the kind of thing I'd expect:
- In an area with foot traffic, frontage is valuable because you could use it for retail. The classic approach to this is to locate retail on the street and other uses in upper stories of the same building. What you control for matters; if you control for "is the property a residence", a residence that occupies a bunch of valuable frontage will have to pay for it whether they want it or not. (A residence that occupies a bunch of worthless frontage, not so much.) Residences like that should be very rare, but the statistical control will hide that fact.
If you instead control for "how much of the property is frontage", I would expect that properties that are slanted toward frontage should be much less likely to be residences or offices, and properties that are slanted toward depth should be much less likely to be retail.
- Analogously, a buyer might acknowledge that they had to pay more for a house because of its abundant frontage - though I'm having trouble imagining this happening in reality - but I wouldn't expect them to say that the abundant frontage was what made them willing to pay more. If a building already exists, then the facts about it are hard to change. (I do have an example to hand from the other direction. I occasionally visit a shop that has zero frontage. It is located in what was formerly (and partially remains) a factory/office complex; I guess the demand for offices dried up. That shop has to post a lot of signs in more accessible locations (and within its own complex!) advertising where you can find it.)
Micheal Thanks for your comment and support.
Today I’ll be introducing to my subscribers how to start making passive income from investment /business.
I’ll be using this opportunity to lntroduce you to Mr.Taylor my personal Broker for a Profitable and Life-Changing lnvestment.
WhatsApp; +1 (585) 623 0133
Reach out to him now for guidance☝️
Tell him I sent you.
So what sort of curve specifically is the Somers depth curve?
From what I can tell, a custom spline.
Sniffnoy Thanks for your comment and support.
Today I’ll be introducing to my subscribers how to start making passive income from investment /business.
I’ll be using this opportunity to lntroduce you to Mr.Taylor my personal Broker for a Profitable and Life-Changing lnvestment.
WhatsApp; +1 (585) 623 0133
Reach out to him now for guidance☝️
Tell him I sent you.
The class of lot I think would challenge the depth rule is the case where you have an independent house sited down a driveway that gets it behind a street-facing house.
So, imagine you have a 50-foot wide, 100-foot deep lot, and you split it in half, building a home on the front and back halves. Then you carve a ten foot wide driveway off the front lot. Maybe you put a garage for the front house along that driveway as well.
I suspect that despite having a smaller "exclusive" footprint, the front parcel will be worth a bit more than the back one. But not nearly to the extent the Somers table suggests.
I think the truth will be closer to something like, value the two parcels together using the gradient rule; then split that value something like 55/45 between the front and back lots.
Skimming through the book, around page 101 they _kinda_ get at this, with the discussion of lots A and B on the diagram between the imaginary 3rd and 4th Streets... But I'm not seeing anything quite like the form of lot split I'm describing (which I've seen fairly often), where the "back house" is on a flag lot that only fronts the street via a narrow access. (Interestingly, the concept of the "flag lot" seems to be pretty widespread in modern assessment and real estate discourse, but a search in the book doesn't seem to find the word "flag" at all, or at least not that Google's OCR finds.)
Update: I just found some historical exemplars of the Somers system:
This one from 1910:
https://cplorg.contentdm.oclc.org/digital/collection/p16014coll6/id/28073
And this one from 1946:
https://www.railsandtrails.com/Cleveland/Zangerle/Zangerle1946-300t.pdf
Both conducted under John Zengerle in Cleveland, who had tenure throughout the entire period. The 1946 report includes a breakdown of their methodology, which includes a section on "back lots" and alleys -- I'm curious what you think of those and to what degree they address your issue or not.
Well, that Zangerle document has an example of a flag lot of a sort (lot B on page 8), but the "pole" on it is pretty thick (25'). They basically just apply the gradient -- they count the full-depth part of the lot the normal way, and then take the part of B that's behind A, and just subtract the value of A from the full-depth version of that slice of A+B.
I am fairly confident that, whatever the combined value would be of A+B, just running the gradient across like this is crediting too much of the value to A. As long as you have a plot that is practical for some use -- to live on, to conduct business on -- there is some base amount of value to simply having a claim on that plot. The closer you get to a classic flag lot, where the part of the flag lot that runs out to the street is just a driveway next to the front lot, the worse the purely-gradient-based estimate will get.
That seems like a reasonable critique.
Additionally -- in residential areas I feel like street frontage can also just matter less in general as long as you have *any* access to the road network and city services. In fact in many places I've noticed that *lack* of street proximity is associated with higher value, the opposite of the traditional Somers assumptions (people like quiet streets and culdesacs, etc). There's some nods from Zangerle on this -- he notes that residential lots don't really benefit from corners -- but otherwise he sticks pretty closely to an Orthodox Somers approach.
This suggests to me the Somers System--taken exactly as described--is most adapted to urban contexts, especially commercial ones, where lots tend to be small, space is at a premium, and road frontage a primary concern. The conventional LEA approach might be better adapted to residential neighborhoods where the lots are all pretty much the same size anyways.
What would you suggest as the best approach for flag lots, in or out of a Somers system context?
Just to be clear I think the Somers system is pretty cool, especially given how easy it is to code this kind of gradient logic with modern computing power.
I think we need some kind of modification that assumes some kind of base value to having the location at all. Like, figure out what the highest-value use is that the plot is feasibly usable for (even if the "implementation" of that use is a "minimum viable product" version of the use), and assign a base value to each lot based on that, and then add onto that value based on the Somers gradient.
So in the case of the flag lot, the fact that you could hypothetically develop (say) a three-story triplex on each of the two lots, puts a floor under the value the lot could possibly have, given how we think six families would value just "living on this block", regardless of whether they're directly on the street. So each lots gets an assignment of $300k for that (or whatever). If the total Somers value for the frontage ought to be $1M, we take $600k of that out for that "base value", and then assign the remaining $400k proportional to how the gradient "shades" each sqft of each lot.
In this case we're assuming that there is an absolute minimum $100k value to a unit you can live in to get a toehold in this location.
This does complicate things, in terms of valuing things like "the smallest possible living unit people will tolerate to live here", and "the smallest possible retail outlet", and so on. Maybe we just don't care enough about this to try to complicate things like that, and could do some kind of adhoc adjustment to flag lots.
If I were going to really work on this (which I don't have the time for) I'd probably start by trying to find sales data where the front and back lots in a pair like this both sold near in time, especially if there's a lot nearby of similar size to the combination of the front and back. If you could collect enough examples of that, it would help provide some points of comparison to get the logic right.
Thinking about it a little more, I wonder as well, whether you could perhaps establish this kind of minimum-viable-unit value in the course of public meetings, as well? Like, have people establish for retail zones, what do they think the value would be to just being able to conduct business here _at all_? Like if you could own just a little hole-in-the-wall operation, a hundred square feet or something? And for residential zones, what would you pay for a postage stamp studio apartment?
Yeah these are all great points! And I wonder how much considerations like these came up in actual Somers proceedings? The historical record seems to suggest that the Manufacturers Appraisal Company had an "Orthodox Somers" method they tried to push everywhere, which *immediately* sprouted all kinds of local warts and hairs the minute it was actually put into practice in a particular location (one of the most common insistences being to employ a locally calibrated depth curve).
The "base lot" value does make sense and jives with practice I've seen -- a lot of residential neighborhoods are valued "by lot" where the presumption is that demand for land (and actual variation in lot size) is pretty inelastic. People just want "a nice house," and plus or minus this much yard within some tolerance isn't really entering the pricing picture at all.
Also the canonical example in actual mass appraisal that comes up all the time for postage-stamp lots is townhomes/rowhomes.
Also don't get me started on valuing condos, who don't technically have their own land attached to the parcel at all, and the parcel is just one of 10 different "building only" parcels stacked vertically on top of each other on the exact same location (presumably you would model them as one combined building, value the land they collectively share, then pro-rate the land back out but... argh. Edge cases upon edge cases! :)
The typical (and in fact overwhelming majority of) Chinese housing has no street frontage. What faces the street is a big wall with a gate and a security guard. Inside that you have maybe a couple dozen apartment buildings. (And a public area for the residents, which might be huge and luxurious or small and bare-bones, depending on the price level of the complex.) All of those couple dozen buildings share the same driveway, for residents with cars.
I already left an independent comment about it, but I'm struggling to see why there would be a depth gradient for residences at all. It obviously isn't the case that rooms in your house located far from the street are less valuable or otherwise less desirable than rooms located near the street. They tend to be more desirable because of the reduced noise.
So I'm definitely open to the idea that the Chinese market is just different and might play out with different dynamics; also I replied to your other comment.
That said, the main thing to understand with depth is that ultimately the depth curve is just one of many ways to capture how land prices change with scale. I've already registered my skepticism that land value varies perfectly linearly with frontage as the Somers system assumes, as well as my skepticism that there's one universal depth curve.
The main observation that I am confident in is that, keeping location fixed, land tends to get more expensive as it gets bigger, but each *marginal* square foot of land gets cheaper as the lot gets bigger. I've seen this pattern repeat over and over again, and every assessor I've ever talked to agrees. If it's true that in China that residential land prices just keep increasing linearly as you make the back yard deeper and deeper, I'm perfectly happy to admit to that if you show me the evidence, I would just find it really surprising!
Indeed; there’s discussion of “irregular” lots but I didn’t find any treatment of flag lots specifically
Auros Thanks for your comment and support.
Today I’ll be introducing to my subscribers how to start making passive income from investment /business.
I’ll be using this opportunity to lntroduce you to Mr.Taylor my personal Broker for a Profitable and Life-Changing lnvestment.
WhatsApp; +1 (585) 623 0133
Reach out to him now for guidance☝️
Tell him I sent you.