'Legacy' cities often lack functional real estate markets. What exists is (de facto) central planning of land, and thus of the built environment. Yet many cities are reluctant to let go of these land assets, as the municipal elders fear a) the loss of control, or b) being accused of having let land go too cheaply.
The dangers of doing so are manifest, as land tends to fall into the 'Dead Hand' of corporations. As property taxes are assessed on the value of the improvements on the land, rather than the land itself, there is minimal disincentive to 'sit' on a parcel of land. The obvious solution is to tax property of the value of the land, rather than the value of the improvement.
The taxation of property on the value of the improvements is a historical quirk. Initially, taxation was assessed on the value of rural 'working landscapes' with large acreages. The value of the property could be assessed by the bushels of grain produced. Clearly, a similar paradigm could not be applied to non-productive urban land, which clearly had value. Initial efforts assess the value of urban land were little more than 'sumptuary taxes' based on the count of windows or the width of frontage on a public street. Over time, advances in economic theory demonstrated that the value of property was proportional to the economic rents derived from ownership, providing a sound basis on which to assess the value of property. Where no rents existed, the owners equivalent rent could be provided by using a comparable building.
SO WHY THE SEPARATION BETWEEN IMPROVEMENTS AND LAND?
In times before the limitations of zoning, there was a fairly direct proportionality between the value of the land and the degree of development on it. Post-zoning, this was no longer true. Land values continued to rise, but the value of improvements was effectively capped by the characteristics of the structure. While it is possible to retrofit a structure to raise its value, the surest method to raise the value of an improvement is expanding the total number of livable square feet.
Property taxes are assessed only on the value of improvements on the property, rather than the total value of the property. On its faces, these seems equitable. A cheap house thus only faces a small tax bill, while an expensive one faces a large bill. But it is very far from egalitarian. The poorer you are, the higher your cost of capital. As is reflected in everything from mortgage rates to the the use of high-interest forms of debt (credit cards, payday loans), and the lower your total loan limit. Thus, your capacity to purchase a home in a desirable location is limited.
You must thus purchase a home in a less desirable location, where costs are lowest. This results in the "Drive 'til you qualify' phenomenon, where lower income households are forced to buy house on the periphery of the metro area. Homes are developed there because value values are lowest, and so the ROI for a land developer is greatest. Consequently, the land to housing ratio is at its lowest, so that the least wealthy households are subject to the (proportionally) highest property tax burden. While seemingly egalitarian, a home value tax is actually a deeply regressive form of taxation.
Further, houses on the periphery are subject to the highest transportation cost burden. The trade-off between housing and travel costs has been theoretically establish since Von Thunen. It is possible to reduce the costs (in both time and money) of transportation by choosing housing in a more central location. However, the savings in travel time costs are typically capitalized in home values, making the more central desirable neighborhoods less affordable, and thus less available...
The regressive nature of this tax structure also applies in central locations. In many urban environment, much of the cost of property is reflected not in the improvements, but in the value of the land. The value of urban land is largely determined by its proximity to other land uses, either through amenities/disamenities of proximity, and the degree of accessibility. In desirable locations, the value of land is often greater than the value of any improvements on the land. Thus, the tax burden, proportionate to the price of the house, is also the lowest. This is especially true in contexts where planning/zoning limit development potential.
The combination of property tax treatment and high purchase prices makes property in central locations into attractive investments. The desireable location location translates into high rents, with minimal regard to the quality of the improvement (the housing itself). History is demonstrably replete with examples where housing has been allowed to depreciate into dangerous and barely habitable conditions while still remaining attractive to prospective tenants, so maintenance costs are minimal.
The same phenomena explains the endurance of central city parking lots. The improvement value of a parking lot is minimal, and thus so are tax costs. So long as the rents from a parking lot exceed the cost of capital and tax costs, downtown real-estate is worth more as parking. Following this logic, many cities have followed the path of Dallas and Detroit by demolishing older structures to make way for parking lots. It is difficult to argue that the 'highest and best use' of central city land is as a barely paved parking space. The importance of old buildings as incubators for new business (by reducing threshold costs) was noted by Jane Jacobs decades ago.
Assessing land values on the basis of land area would be more equitable. While applying an areal tax in the (rural) context of a working landscape, where acreages are high and profits low would be regressive. Thanks to GIS, there is no technical limitation on the ability to do so.
Urban agriculture...flips side too it...can't be used as a parking lot due to access control limitations.
Can't be developed (economically) due to setback requirements-->Parking lot
Can't de parking lot (due to access control limitations)--->Urban agriculture.