Are profits from subdivision development higher in areas with more regulations? A case study of South Kingstown, Rhode Island and some implications for land use planning

Abstract

Little is known about the relationship between land use regulations and profits from subdivision development. Using data from a heavily regulated market, this article presents analyses that for the first time determine actual profits from subdivision development.

This study found an average profit, measured by internal rates of return, of 29%, which is statistically greater than what scholars consider normal. Profits decreased as time to complete the subdivision increased. Profits also decreased because of delays due to regulations and voluntary decisions by developers. The findings suggest that higher profits could be attainable in more-regulated areas because regulations may create amenities whose value exceeds the costs of providing them, while the same regulations may make it difficult for outside developers to enter the market thus limiting competition among developers. The findings suggest that policy makers should be cautious about regulations that create amenities when these regulations may inhibit competition among developers.

Keywords:

• real estate
• construction/building
• development
• land use

Notes

¹Normal is defined as the profit required to function in a perfectly competitive market. If profits were to rise above this level, more developers would enter the industry. If profits were to fall below this level, developers would leave the industry.

²The size is also indicative of the fact that, in a given land market, there are fewer transactions on undeveloped land – the starting point for these analyses – than there are on developed land, as noted by Pogodzinski and Sass (1990). Indeed, Colwell and Sirmans (1978) and Colwell and Scheu (1989) performed regressions on undeveloped land prices in samples of 26 and 20 observations, respectively. In contemporary research, Gyourko and Rybczynski (2000) were able to interview only 23 developers in a study of new urbanism published in Housing Policy Debate.

³Unleveraged profits assume that developers invested 100% of their own capital and did not finance any of the investments. In practice, developers invest only a portion of their own capital and use borrowed money to “leverage” their investments and amplify profits.

⁴Other studies that document the costs of regulations include Isard and Coughlin (1957), Rosen and Katz (1981), Johnston, Schwartz, and Hunt (1984, cited in Luger and Temkin, 2000, 137), Glickfield and Levine (1992), Landis (1992) and NAHB (1995). To be sure, regulations may make some land undevelopable so that the land loses value (Brueckner 1990). Even in these situations, however, such price decreases are limited only to the portion of land that became undevelopable, and prices for the remaining developable land will increase (see, e.g., Quigley and Swoboda 2007).

⁵The mild recession of 2001 does not appear to have had a lasting impact on property prices in the study area.

⁶See, for example, Colwell and Sirmans (1993) for the declining marginal value of additional land and Knaap (1985), Nelson and Knapp (1987) and Thorsnes (2000) for the value of other amenities.

⁷For reasons why this is the case, see Mohamed (2009). However, Tanner (2002, 837) notes that bonds that are prepared by local governments may overestimate the costs of improvements. See McDermott (2002, 865) and Hulme (2002, 871–72) for more discussions of performance bonds.

⁸The cost of a bond usually depends on the credit worthiness of the developer and the size of the bond being requested.

⁹The bond also contains estimates for engineering and contingency costs. The former is estimated as 5% of infrastructure costs, and the latter is estimated as 10% or 15% of the sum of infrastructure and engineering costs.

¹⁰The files also provide information on the costs for required off-site improvements. Off-site improvements were required only in three of the 20 subdivisions studied. They usually included minor work, for example, widening a turn into a road that leads to the new subdivision.

¹¹An example is use of the IRR in the presence of alternating positive and negative cash flows. However, other approaches to addressing this problem have their own limitations.

¹²I chose the range of 10–100% because it is sufficiently wide to detect if more than one IRR exists.

¹³Regression results in this article should not be treated as precise; they are intended for expository purposes only. A larger dataset is necessary for more complete analyses.

¹⁴Ideally, one would like to examine the effect on IRRs of the time it took to obtain approvals by performing a bivariate regression between these two variables. However, such an analysis is problematic for two reasons. First, of the eight subdivisions that were bought with approvals ( Table 1, column 11), intermediate developers obtained and paid for approvals for four. For these four subdivisions, the time to obtain approvals was not a concern of the developers who actually built them. Second, because it is reasonable to consider the time to obtain approvals as beginning from the point when the development process began, as defined earlier, the time to obtain approvals for these eight subdivisions was, in effect, negative. Thus, the regression results would not be intuitively interpretable.

¹⁵Bivariate regressions between the IRR and the year the parcel was bought or the year the development process began did not reveal any association between the IRR and these two variables. In both of these analyses, I used an index to represent the year. Note that the year the development process began is different from the time for the development process to begin, as defined above.

¹⁶This is the only instance where I employed more than one independent variable in the regression analyses.

¹⁷“Voluntary” as used here simply means that there were no regulatory reasons for developers to hold onto lots.

¹⁸I do not account for lot size when comparing prices because lots in the same subdivision in the study area tend to be around the same size. Further, lots in the same subdivision have the same features, so accounting for different features when comparing prices in a given subdivision is also not necessary.