Sergio Copiello*, Laura Gabrielli**

* Università IUAV di Venezia

** Università di Ferrara

Sunto

The title of this contribution is provocative on purpose.

An established cornerstone of real estate appraisal is that properties are composite assets, which market value may be expressed as a linear combination of the land value and the building quality. The literature suggests that the apportionment between the two sources of value represents a significant research strand. Property market value modeling traditionally relies on hedonic functions, which right-hand side features the so-called utility bearing attributes, partly pertaining to location and location-based services (hence, the land), and partly referring to structure, finishes, equipment, and so forth (thus, the building). In the course of time, the enrichment of traditional hedonic functions with time series models has shown that serial dependence plays a significant role towards shaping the property market values. Besides, the development of spatial autoregressive models has raised the awareness of researchers and practitioners that space does matter in order to explain them.

Nevertheless, as far as the first question in the title is concerned, precisely the literature on spatial econometric models has pointed out that several limitations and shortcomings hinder our ability to achieve in-depth knowledge of spatial-related phenomena. Identification issues due to the reflection bias and multicollinearity issues, among others, often prevent to identify all the underlying explanations and especially the causal relationships. The reflection bias occurs when trying to explain the behavior of a unit (let us suppose it is a property located in a downtown area) using the data of the group to which the unit belongs (i.e., all the downtown properties), so the causes and the effects are indistinguishable. Multicollinearity issues are intrinsic to spatial autoregressive models, the property market models as well. Moreover, they are endemic in spatial data analysis, because the potential explanations are typically more related each other than to the dependent variable. Whatever the shortcoming, the results are likely to be biased, barely informative, and perhaps they also lack explanatory power. With regard to the land value, the above issues are exacerbated by sparse and noisy data. As a case in point, let us think to the paucity of information and studies about land leverage in comparison to the plentiful data on construction costs. Therefore, although our knowledge about the real estate market and how it works is ever growing, we still lack the ability to achieve comprehensive insights about several of its facets, precisely that kind of insights that would be required to capture land rent.

As far as the second question in the title is concerned, leaving the above criticisms aside, the topic of land rent should be framed under the scope of the purposes that town planning models and practices aim to pursue. Although land rent sinks its roots in the Ricardian classical theory, under the framework of the neoclassical economic paradigm it is none other than a way scarcity shows itself. Absolute land rent stems from the scarcity of buildable lands in comparison to all the other uses, while differential land rent correlates to the scarcity of well-located land plots suitable for building purposes. Both absolute and relative land rents are also generated by the choices taken when drawing the town plan, and both contribute to optimizing the resource allocation choices in the building domain. During the last thirty years or so, town planners have repeatedly struggled for dictating the targets on the agenda, among which there is one that plays a pivotal role. Initially, it was defined as limiting urban sprawl. Now, a beloved target is to avoid the loss of agricultural land, a phenomenon that is sometimes referred to as soil sealing and land take. Rather surprisingly, the urban containment policies derived from the above target neglect the optimizing role played by the land rent. This may be due to the fact that town planning traditionally relies on command and control regulation. Hence, it mostly employs authoritative and coercive implementation tools, rather than exploiting the virtues of market-based incentives. However, here we suggest that land rent implies, in and of itself, phenomena such as the exploitation of already existing localization opportunities and urban densification.

We deem there are at least two intrinsic pitfalls in the capture of land rent. The first is that the amount of information needed to pursue that purpose effectively is likely to exceed the current knowledge on how real estate markets work, hence only approximate results can be achieved. The second pitfall - probably the most important - lies in the fact that the land rent capture purpose is likely to conflicts with another topical aim, such as tackling and hindering urban sprawl and land take.

Parole chiave: Land rent; Capture; Land value; Land leverage; Real estate modelling; Neoclassical economic paradigm; Scarcity.