Assembly Testimony on Equity in Assessment Practices

H. William Batt

[Testimony presented by H. William Batt, Ph.D. to the New York State Assembly on February 6, 2007. Reprinted from GroundSwell, March-April 2007]

Assessment has historically been the Achilles Heel of the real property tax, and is an important reason why the public resents it. We really have two taxes, one on land value and one on improvements, each with very different dynamics, and until people realize the significance of this, nothing else matters. Buildings depreciate just like cars, computers and refrigerators, 0.5% to 1.5% yearly; land appreciates. Land typically increases in market price because the surplus of community enterprise comes to settle on locations in the form of economic rent. So, unless one revalues frequently, things get quickly out of line. A land value tax (LVT) recaptures that surplus for public use.

The tax on the building part causes many problems: it discourages responsible maintenance, the highest and best use of sites, fosters sprawl development, eviscerates urban cores, and depletes the environment. But taxation of land values is really the superior revenue source, as some eight Nobel prizewinning economists have attested. It conforms to all the textbook principles of sound tax theory -- neutrality, efficiency, progressivity, stability, simplicity, administrability. Yes, the property tax is progressive, despite what some people have said. Especially the land part --tenants pay nothing, and LVT is the only tax that actually fosters economic development and vitality.

There was a time when critics complained that it was difficult to assess the land and the building components separately. Hence, standards for assessment were outrageously low. Now GIS triangulation algorithms are not only more accurate but also cheaper and quicker. It would be worth the State's investment to use its battery of GIS techs to avail itself of and improve upon this work; the payoff would be enormous. It is now possible for the record of every parcel sale to be quickly entered into the database, then passed over to the Assessor's office each night, thereby maintaining a real-time cadastre of a municipality. Present costs per parcel for a complete revaluation are anywhere from $60 (upstate) to $100 (downstate), but the cost can be reduced to less than $10 once the computer technology is instituted. Apportioning land value for condos and coops is just as easily done.

Many American assessors make the mistake of first valuing structures and then treating land components as residuals. The failure to conform to standard practices elsewhere in the world typically leads to over-valuing structures and gives an advantage for depreciation on corporate tax schedules. To this extent, assessors serve the interests of the finance, insurance and real-estate industries. The distortion can be seen in the fact that the aggregate land value proportion in a typical American city is 1/4 to 1/3 of the total tax base, far less than in municipalities elsewhere. Greenwich, Connecticut, a city that employs the building residual method, has a land value proportion of 71 percent. The Mastick Commission Report done for the New York State Legislature in 1932 reports that the aggregate assessed land and improvement component in New York City was $12 Billion for each.

New York State and its people pay a price for not taxing the surplus that accretes to land sites in the form of economic rent. Land, one must recall, has a fixed supply --what economists call inelastic. This means that any tax imposed on land sites is incorporated -- capitalized -- into the market value. The more tax is added, the greater the downward pressure on the land price. This makes housing more affordable up front. It's a simple case of the buyer "paying now or paying later." (There are economic pressures that work in the opposite direction that may stabilize site prices, but that's a secondary point of little consequence here.) By society not recovering the economic rent from locations, market prices rise and so do taxes, witness California and Florida. This rise in values gives some titleholders windfall gains and prices young households out of the market and leads others to keep more housing than they need. It is better to collect the rent and thereby stabilize prices. It helps economic vitality in the long run. For those who find that any burden is difficult "now," they can defer their tax until selling later -- some 24 states have a provision like this.

It is also important to understand the enormously steep gradient of land values. Urban locations can be hundreds of times more expensive than peripheral areas. Since central cities are largely commercial in nature, residential parcels farther out have a much lower land value, and farmers' land value is essentially trivial if not already protected by other save-harmless provisions. The problem is much of high value urban land sits underused or vacant --often as much as a third -- and taxing land value fully typically induces the better use of those parcels. When a property tax is shifted to a land value tax, most homeowners pay less, and derelict high value parcels pay more. They are prompted to develop rather than being held off the market for speculative gain.

The important point is to get the land values assessed at what they should be, and frequent review is important as much as is accuracy. Maryland has state assessors that do one-third of the counties each year, so valuations are never more than three years old. I have been part of a team that simulates how LVT works anyplace in Maryland -- see www.marylandlandtax.org. We're doing similar simulations in New York, though the assessment data is often far poorer. One can see the results for twelve counties at www.newyorklandvaluetax.org. Phasing out taxes on improvements typically gives most homeowners a break. This policy is legal in New York, and it awaits a trial in a city with good land assessments. Amsterdam tried doing this a decade ago, but abandoned it after one year due to poor assessments and lack of public understanding. With good land assessments, a quick adjustment in the computer application will both relieve lots of political and economic pressures and revitalize many moribund urban economies. I have addressed in passing all the bills in question. But it is especially important, I believe, to look at the incidence of each component of any tax on real property - both land and improvements -- because up-to-date studies are needed.

Some colleagues of mine have done some work in Baltimore, but it relies on block and zip code data related to housing, and this is very gross analysis. I know of some studies that have looked at progressivity but have ignored the burdens on renters and on non-residential property. My work (unpublished) shows that households typically pay about half of all property taxes, and the remainder is paid by by non-residential property owners. With LVT, although homeowners are the overwhelming number of titleholders, their burdens are typically small, only the land under their houses. Taxing only land values immediately relieves the one third of households, mostly poor people, who own no land at all.