Academic journal article
By Shilling, James D.; Vandell, Kerry D.; Koesman, Ruslan; Lin, Zhenguo
The Journal of Real Estate Research , Vol. 28, No. 4
This paper is concerned with the extent to which rehabilitation tax credits affect the conditional probability of commercial real estate rehabilitation. The analysis suggests that rehabilitation tax credits have been a significant determinant of the conditional probability of rehabilitation in the Boston office market. A significant portion of rehabilitation tax credit investment is investment that would have been invested elsewhere, about 60% to 65% in certain periods, but rising to as high as 90% in other periods. The findings indicate that the rehabilitation tax credit has a significant and substantial influence on the conditional probability of rehabilitation. The findings also reveal that the greatest amount of slippage, not too surprisingly, generally occurs when the tax credit is low and when the gain from rehabilitation before the tax credit is high.
Relatively little has been written about the rehabilitation tax credit, even though it has been a feature of the tax code in the United States 1978. After the Tax Reform Act of 1986 (TRA 86), the rehabilitation tax credit was applied to industrial, commercial, and other income-producing buildings (factories, retail stores, hotels, and motels) built and originally placed in service before 1936.' The amount of the rehabilitation tax credit varies with the age and use of the structure. The credit is allowed for all recognized expenditures provided the expenses are substantial. The primary motivation for the rehabilitation tax credit is to favor the rehabilitation of older, more centrally located buildings that are already served by an existing infrastructure of streets, utilities, and civic buildings as against the construction of new buildings at the urban fringe.
This research uses data from a panel survey of the Boston office market to examine the effectiveness of the rehabilitation tax credit. Past analyses directed at ascertaining the effectiveness of tax credits have produced mixed results.2 Clark (1979) and Hendershott and Hu (1981), for example, use time series observations of equipment investment and the cost of capital to uncover the relationship between tax credits and the level of investment. Their results show only modest effects of investment subsidies. Auerbach and Hassett (1991) use an event-study methodology to examine the effectiveness of investment subsidies. Their findings suggest no statistical relationship between investment spending and changes in the investment tax credit. Cummins, Hassett, and Hubbard (1994), on the other hand, using cross-sectional responses to tax law changes to identify exogenous shocks to firms' investment conditions, find that tax policy has a significant and large effect on investment.
The techniques used in this paper are inspired by the literature on failure time models (e.g., Cox, 1972; and Kalbfleisch and Prentice, 1980). The empirical model relates the conditional probability of rehabilitation to the value of the rehabilitation tax credit and other characteristics. The model is used to forecast the rate of rehabilitation with and without the rehabilitation tax credit. This empirical approach has several advantages. First, it makes possible a much clearer comparison of the relative effects of different factors on the decision to rehabilitate. second, it permits the use of right-censored data (unrehabilitated properties). And third, it avoids certain econometric problems-aggregation and simultaneity bias-that plague the past literature.3
The paper proceeds as follows. First, a simple model of the rehabilitation decision for the conditional probability of rehabilitation is presented, which can be derived as a function of the gain from renovation, including the rehabilitation tax credit. The greater the gain from renovation before the rehabilitation tax credit, the greater the conditional probability of rehabilitation, all else held equal. In addition, tax credits can have a significant effect on the conditional probability of rehabilitation. …