Land use for transport projects: Estimating land value

Land Use Policy 42 (2015) 594–601

Contents lists available at ScienceDirect

Land Use Policy journal homepage: www.elsevier.com/locate/landusepol

Land use for transport projects: Estimating land value Doron Lavee a,b,∗ a b

Department of Economics and Management, Tel-Hai College, Upper Galilee 12210, Israel Pareto Group Ltd., 7 Giborei Israel Street, Netanya 42504, Israel

a r t i c l e

i n f o

Article history: Received 2 April 2014 Received in revised form 11 September 2014 Accepted 22 September 2014 Keywords: Land value Transportation projects Economic feasibility Economic model Elasticity

a b s t r a c t Transportation projects are typically characterized by increased land use, which is a scarce resource of economic value. However, there is a tendency to ignore land value during feasibility studies of transportation projects. This may lead to a reduction in the economic efficiency of a project and to increased land use. This paper presents an economic model, based on the relationship between the elasticity of land price with respect to density, and estimating the future value of land designated for various uses, including transportation projects. The model was applied to transactional data from Israel, and was used for examining the value of land designated for two transport projects within Israel. The conclusions of the study indicate that taking the land value during a feasibility analysis of transportation projects into account, may lead to the consideration of other alternative plans, which may prevent the excessive use of land. © 2014 Elsevier Ltd. All rights reserved.

Introduction Land is a scarce, non-renewable natural resource and is one of the most valuable non-produced assets for most countries (OECD, 2008). However, the price of land is exposed to various market failures and political decisions which may influence the way land is used (FAO, 2007). This problem is reflected particularly in the case of transportation projects, which are characterized by high consumption of land, imposing environmental and esthetic costs (Litman, 2004). It is estimated that 1.5–2.0% of the world’s total land surface is occupied by transportation infrastructure, primarily for roads and parking lots. In urban areas, 30–60% of the land is taken by transportation infrastructure, and in some cases, this figure can reach 70% (Rodrigue, 2013). Land designated for transportation purposes has significant economic value, mainly since many transportation infrastructures are located in close proximity to key destinations, in areas with high-value land (Litman, 2005). Continuous growth in demand for various land use (residential, employment, public, etc.), as well as the need to develop transportation infrastructure for transportation services, create a constant conflict when determining land uses. This situation leads, in most cases, to land resources that are

∗ Correspondence to: Pareto Group Ltd., 7 Giborei Israel Street, Netanya 42504, Israel. Tel.: +972 9 8361000; fax: +972 9 8857667. E-mail addresses: [email protected], [email protected] http://dx.doi.org/10.1016/j.landusepol.2014.09.020 0264-8377/© 2014 Elsevier Ltd. All rights reserved.

directed to transportation projects at the expense of alternative designation which could have been determined for the land. Often, land designated for transportation is incorrectly considered a sunk cost with little current value, even though there are alternative potential uses for the land (Litman, 2005, 2012; Litman and Doherty, 2009). Several studies indicate that when examining the feasibility of various transportation projects, there is no consideration of the land value, or the land is attributed a low economic value, unless it is necessary to purchase the land (Delucchi and Murphy, 2005; Ketcham and Komanoff, 1992). In fact, land value is not taken into account during economic feasibility analysis of transportation projects in many countries, such as Ireland (Goodbody Economic Consultants, 2004), the EU countries (TRIP, 2004) and New Zealand (Booz Allen Hamilton, 2005). In Denmark, economic feasibility analyses of transportation infrastructure projects include land value only if it is required to purchase it (TRIP, 2004). In the United States, land value is not taken into account when estimating transportation projects (Delucchi and Murphy, 2005). In Canada, land designated for roads is often regarded as a sunk cost (Victoria Transport Policy Institute, 2009). Underpricing these lands reduces the economic efficiency of such projects and may lead to high land consumption and loss of open spaces (Braid, 1995; FAO, 2007; Poole, 1997; Roth, 1996). The distortion created when assessing the feasibility of transportation projects to be built on land that the developer (usually the state) is not required to pay for, causes a lack of efficiency when comparing various alternatives for a certain project, and when comparing different projects that compete for the same budget source.

D. Lavee / Land Use Policy 42 (2015) 594–601

This leads to economic inefficiency in the allocation of resources of the economy, due to disregarding the land’s opportunity cost. This opportunity cost results from the damage caused to the economy by preventing alternative use of the land. Therefore, a correct land valuation is an essential part of a sustainable management of land resources when conducting economic feasibility analyses of transportation projects. In some cases, an active land market exists for land designated for transportation projects. In many other cases, however, there is no active land market and the market price is difficult to estimate (Belli et al., 1998). Non-market valuation techniques, such as the Hedonic Price Method (HPM), Contingent Valuation Method (CVM) and Travel Cost Method (TCM), are common methods used to estimate the value of non-marketable land. However, such methods require a specific valuation of the location in question. The aim of this paper is to create a uniform method for estimating land value for all transportation project feasibility analysis, allowing a comparison between projects, or between different alternatives for each project on a uniform basis. In the economic literature, the conventional method of land valuation in the case of non-marketable land is based on the values of adjacent lands (e.g., Booz Allen Hamilton, 2005). For example, an area bordering a residential area is likely to be expanded into a residential neighborhood, an area bordering an industrial area is likely to be expanded into industrial area and so forth. This method can be technically challenging, since often areas adjacent to roads have a variety of land uses (Woudsma et al., 2006). Delucchi and Murphy (2005) argue in their study that when utilizing land for transportation projects, the cost that should be taken into account is the value of the land in the free market, meaning, the price the government would pay for the land were it not already the owner. According to a study conducted by Anas et al. (1997), the economic value of land should be considered in terms of price and taxation based on similar land which is in use, in order to prevent a distortion which would increase the land used for transportation at the expense of other uses. It should be noted that land used for applications such as industry or municipal residence, generates tax revenues, while land used for transportation typically does not. Reference to the price of land and tax implications is particularly significant in urban areas where land prices are high, because there is competition between the different uses (Vickrey, 1997). Litman and Doherty (2009) note in their study that the opportunity cost of land used for transportation in urban centers ranges between the value of adjacent land and urban periphery land. However, a main problem remains when the land designated for transportation is not close to any kind of built-up area. Moreover, the construction of a transportation project is expected to set the land designation. This not only prevents alternative short and medium term land uses, but also damages the possibility to choose a different land use in the long term. The land valuation methods described above provide a current land valuation. However, we argue that based on the above, future rather than current land values should be considered when estimating transportation projects. Classical models of urban economy assume that land value declines with the distance from urban centers (e.g., Alonso, 1964; Beckmann, 1969; Papageorgiou and Casetti, 1971). Many empirical studies have reported on this relationship as well. Yet, parallel to the decrease of land value, population density decreases with distance from the urban centers as well (Sun and Tu, 2005; Woudsma et al., 2006). The main reason that can explain this phenomenon is that transportation costs and travel time to city centers, which constitutes employment centers, increase with the distance from the urban centers (Cadwallader, 1996; Solow, 1972). Therefore, population density is an important parameter that can reflect land value, not only at city level, but also at state level. Higher

595

density indicates a high demand for land, resulting in an increase in land prices in the area. Indeed, areas with higher density are usually characterized by higher land prices (Grimes and Liang, 2007). Furthermore, according to the economic literature, higher density leads to economic benefits resulting from economies of scale, known as agglomeration benefits (see Duranton and Puga (2004) for a literature review), which may be reflected in higher levels of rent and land prices. This supports the assumption that higher density is accompanied by higher land prices. Indeed, studies estimating land value consider the density parameter in some form. Woudsma et al. (2006), for example, estimate the land value occupied by transportation projects in Canada, by dividing the country into three geographic categories, according to the level of urban development: large urban areas with high density, small urban areas and rural areas. Hirshhorn (2003) proposes a system that will map the state into regions based on population density in each area (low, medium, or high), when the estimated value of the land is the average value for each region, and in urban areas with high-density, land value is estimated by the distance from urban centers. The main purpose of this research is to offer an economic model for estimating the future land value for transportation projects. The model is based on the HPM and uses data from transactions of land with different characteristics. Some of the parameters that determine the land value change over time, and some, such as geographical location and land designation, are constant. Since there is a global trend of increasing population density, which is a key factor in explaining the variation in land prices over time, we rely on the expected or planned density in a given area for estimating future land value. We demonstrate that the parameters changing over time can be represented by the density parameter. Thus, the presented model provides a basis for assessing the future land value based on land characteristics and the expected level of density in the area. The paper continues as follows: Second section demonstrates the econometric model for estimating land value. Third section applies the econometric model on Israeli data, while fourth section presents two case studies of applying the model on transportation projects. Fifth section presents a discussion and the last section concludes the paper.

The model When estimating the value of a specific land, a large number of variables should be examined. An appraiser’s estimation, which gives a specific value to a certain land, examines many aspects such as location, proximity to services and/or institutions, attractiveness of the region and so forth. The demand for land is affected by various variables, which varies in importance from person to person. Estimating land value when analyzing the feasibility of a transportation project requires certain data uniformity. In addition, an appraiser’s estimation is problematic, as it displays the current land value without reference to the future land value from the perspective of the economy. That is, there is no reference to the fact that the land is a limited resource while the demand for land is growing. Lavee and Baniad (2013) estimated the value of non-marketable land in Israel, based on various measures, such as the distance from Israel’s central region and socio-economic state. In this study, we assume that population density can represent an increase in demand. The purpose of the model is to estimate the land value based on the estimation of the relationship between land value and population density in the area where the land is located. The model presents a uniform method for examining the total amount of land, while maintaining the highest level of accuracy as possible. The estimation is carried out by running a lateral regression on data of marketable land sale transactions, when the

596

D. Lavee / Land Use Policy 42 (2015) 594–601

dependent variable in the regression is the price of land per square meter stipulated in the sale agreement. The explanatory variable is the population density in the area where the land is located. In addition, the following control variables that affect the price of land were introduced to the regression: Land designation (construction or commercial and industrial), location of the land in the metropolitan area as well as the socio-economic level. In addition, the country is divided in the model to n different regions, depending on density levels in these areas.1 The estimation method is described by the following regression equation: ln(p) = ˛ + ˇ0 ln(d) + ˇj ln(d)Aj + k ck + ε

(1)

where p is the current price of land per square meter, d is the population density of the land in the area, Aj are j dummy variables representing the different regions of the country (j = 1, . . ., n), and ck are the socio-demographic variables.

possible to estimate the future land value given its present value. In particular, the future land value vf is: f = p · [1 + (ˇ0 + ˇj ) · dˆ j ]

(2)

where dˆ j is the expected rate of change in the density level in area j. Our assumption is that the other explanatory variables do not change over time. The expression within the parentheses is the future value multiplier. Thus, given the current price p, it is possible to predict the future value by using the multiplier. It is important to note that the proposed model considers the feasibility based on the aspect of land value, and does not address various factors that are specific to a particular project. In addition, the model does not address various environmental externalities arising from the land, which include, inter alia, attractive landscape, improved air quality, lower temperatures, noise reduction and shading (Haruvy and Shalhevet, 2004). Discounting

Explanatory variables Density and location: The area in which a transportation project is located is the main parameter when estimating the value of the land used for the project. Land values vary depending on the relationship between the demand and supply of the land in each region. Accordingly, land prices range between very low for land in peripheral areas, to highly expensive land within metropolitan areas. The division of the regions reflects in part the differences in the density level in each of these areas. These differences may be large. We assume that the impact of the change in density on land prices may vary between the different regions. In particular, we expect that the higher the density level, a further increase in density will lead to a stronger influence on the price. Therefore, Eq. (1) estimates the elasticity of price of land relative to the density of each of the different areas. Land designation: Land value is affected by its designation, thus we must differentiate between land intended for construction and land intended for trade and industry. In addition, land on which several units can be built will be more expensive than land on which only a single unit can be built. This variable is essential when evaluating the future value of non-marketable land for transportation projects. In this case, the assessment shall be made according to the alternative designation of the land, which is determined by the land adjacent to the project. In some cases, transportation projects are adjacent to different land uses. In these cases, it is possible to refer to the various uses by the proportion of each use. Metropolitan: Housing unit values in metropolitan areas in Israel are often higher than housing unit values in other areas (Sayag, 2010). Therefore, it is necessary to examine whether transactions that take place in one of the metropolitan areas have a different land price from the price of a land with similar characteristics in other areas. Socio-economic level: The socio-economic level has a profound influence on land value (Kok et al., 2011). For instance, there are communities located in peripheral areas with high land prices. Therefore, the socio-economic level constitutes a measure for the attractiveness of an area and its ability to provide services to its residents, and therefore affect the land value. The estimation will provide a predictive model of the land value given the explanatory variables and the expected level of maximum density. According to the model, the land value elasticity relative to the density in area j equals to ˇ0 + ˇj . By using this elasticity, it is

1 It should be noted that the presented model is a logarithmic model and therefore the density variable coefficients represent elasticity as detailed below.

In order to use the future land value in cost-benefit analysis of transportation projects, it is necessary to discount the future value to the present time. Therefore, the discounting rate and required period should be determined. Various studies have estimated different interest rates to calculate the capital cost for estimating transportation projects. In Canada, Lall (1992) used a 10% real interest rate, which corresponds to the rate usually taken into account in evaluating government projects in Canada. Newbery (1995) implements an interest rate of 8% as recommended in the UK. Other studies use various rates that reflect different assumptions. For example, Boucher (1996) uses a real interest rate of 6% based on early studies of Haritos (1973) and Haritos and Shlachter (1982), which based the capital cost on loans interest rate. Delucchi (1998) used a real interest rate of 3% and 7% in order to calculate the capital costs associated with road infrastructure in the U.S. In Arizona, the alternative cost of capital roads (5%) was measured by reference to the commercial return rate (Mansour-Moysey and Semmens, 2001). In Israel there is a Transportation Project Appraisal Procedure (The State of Israel, 2012), which stipulates a 7% annual interest rate for discounting for transportation projects. Applying the model in Israel Transportation infrastructure in Israel Land is a scarce natural resource, especially in Israel, which is characterized by a small area with high population and economic growth (Shoshany and Goldshleger, 2002). In addition, Israel is characterized by a high variability in population density between central Israel and the periphery, which is reflected inter alia in significant gaps in the transportation infrastructure scope provided in each area, as well as in the level of difficulty in designating land for infrastructure development in the demand areas. Transportation infrastructure plays an important role in Israel, both in terms of economics and in terms of social affairs. Establishing a high quality and efficient transportation infrastructure to connect the periphery with the large urban centers, will greatly contribute to minimizing the gap. Israel’s small size facilitates the construction of the necessary transport infrastructure and their related costs (Ben-David, 2000). Transportation infrastructure investment in Israel includes design, development and maintenance of infrastructure. The budget covers three main areas: the development of railway infrastructure, the development of urban public transport, and the development and maintenance of roads. Out of these, the weight

D. Lavee / Land Use Policy 42 (2015) 594–601

of investment in roads is the greatest (Sofer, 2013). Thus, a gap is created between the public transportation infrastructure and road infrastructure, encouraging the use of private transportation, which is the primary means of travel in Israel today. The number of vehicles in Israel has increased over the last decade in over 50%, and with it, the kilometers traveled, causing considerable congestion on the road network (Israeli Ministry of Finance, 2011). In recent years, the Israeli government investment in the development of transportation increased significantly, from the conception that supply of transportation services, as a public product by the government, is essential for continued economic growth. The main objectives forming the basis of Israel’s transportation policy are (Israeli Ministry of Finance, 2011): • Meeting the increased travel demand, mainly by upgrading and expanding existing inter-urban roads and highways. • Development of advanced public transport systems such as light rail and Bus Rapid Transit (BRT) and improving the public transportation system in order to increase accessibility and reduce congestion, air pollution and road accidents. • Improving the efficiency of the roads and vehicles allocation, through the establishment and expansion of toll roads and fast lanes and improvement of the economic incentives that are reflected in the tax system. • Linking the periphery and center of the country through a system of highways and railways. • Increasing competitiveness and removing barriers in aviation, automotive market, public transport, shipping and seaports. In the Transportation Project Appraisal Procedure issued by the Israeli Ministry of Transport (The State of Israel, 2012), there is no uniform format for calculating the value of the land used for the project. Often, the cost of land is taken into account only when there is a need to purchase it (sometimes through expropriation), or when it is necessary to compensate affected landowners in the projects surrounding area. Applying the model to Israeli data The model presented in second section is applied to Israeli data from 1998 to 2009. We estimated the elasticity of land value with respect to the density in the various areas in the country. The explanatory variables in the regression are calculated as follows: Density: Calculating current density levels was carried out by using data from the Israeli Central Bureau of Statistics (CBS) regarding the distribution of land according to land use as well as on data on the number of existing housing units in each locality. For each locality, the area defined as residential area was examined. The reference to this area remained constant over the years, under the assumption that municipal territories do not change. The density variable was calculated by dividing the number of housing units during the relevant year by the area designated for residential purposes in the same locality. Israel has a National Master Plan (NMP 35) which defines the target density in every municipality. The NMP 35 represents the government’s goal for the desired density throughout Israel. The working assumption in this study is that the future land value will be determined by the expected future density.2 Although there is no certainty that this density will be realized, this presents an examination of the significance of this decision and the expected impact on the land value. Hence, in order to estimate the future land value we must evaluate the

2 The government’s decision regarding density does not depend on the evaluation of the change in land value as was done in this study. It should be noted, however, that this research and other research might influence government policy.

597

increase in land prices, in accordance with the future change in density (according to NMP 35). Thus, when estimating the future land value we use this program to calculate future change in density. The expected rate of change in density is the difference between the current density to the density defined by NMP 35. Location: The country was divided into 11 regions according the NMP 35. Despite the detailed distribution into areas, it should be noted that there are some areas where the variance of the price of land in the region is very high. Metropolitan: As a metropolitan is a factor which may lead to increased land values, it is necessary to examine whether a specific transaction was carried out in one of the metropolitan areas. The metropolitan areas taken into account are Be’er Sheva, Jerusalem, Tel-Aviv and Haifa. Socio-economic level: The socio-economic level was determined by the CBS ranking ranging from 1 to 10, where 1 is the lowest level and 10 is the highest level, and divided into three groups: 1–4, 5–6 and 7–10 (Table 1). The regression is based on data of real estate transactions from the Israel Lands Administration. Each transaction included several details, the most relevant of which are location, land designation, date of the transaction, area size and the sum of the transaction. To these details, we added the density data for each transaction (calculated as the average annual density for each year in the different locations) as well as data on location, socio-economic level, metropolitan and land designation in the form of dummy variables. The final database included 3777 observations. The regression equation is: ln(p) = ˛ + ˇ0 ln(d) + ˇ1 ln(d)A1 + ˇ2 ln(d)A2 + ˇ3 ln(d)A3 + ˇ4 ln(d)A4 + ˇ5 ln(d)A5 + ˇ6 ln(d)A6 + ˇ7 ln(d)A7 + ˇ8 ln(d)A8 + ˇ9 ln(d)A9 + ˇ10 ln(d)A10 + 1 m + 2 S1–4 + 3 S7–10 + 4 R + 5 T + ε

(3)

where p is the additional price per square meter in US$ and d is the rate of change in the density of housing units per 1000 m2 according to NMP 35. Ai is the index for 10 dummy variables for 11 regions (A1 – region 1, A2 – region 2 and so forth up to A10 – region 10, region 11 is the base group) and represents the change in comparison to d. m is the dummy variable for metropolitan (1 – metropolitan, 0 – not metropolitan), Si is the index of the two dummy variables for three socio-economic levels (1–4 and 7–10). R is the dummy variables of land designated for saturated construction and T is the dummy variables of land designated for commercial use. The regression results are shown in Table 2. The adjusted R2 explains 61.7% of the variance change in price. When assessing the value of land for transportation projects, one must consider the alternative designation of the land. The alternative designation is determined by the current designation of adjacent land, or the designation expected according to statutory plans. Current land value is generally estimated by appraiser estimation. However, feasibility analysis of transportation projects should take into account the future land value. The presented model provides the estimated future land value. In the case of Israel, we ˆ based on the expected will use the rate of change in density d, density defined by NMP 35. The results suggest that the density positively affects the price of land, i.e., as the density rises the price is expected to rise as well. These results are in line with the conclusions presented in a literature review. If we divide Israel into 3 main regions, South, Center and North, it can be seen that each area is characterized by a different elasticity of demand with respect to density. In the South area (regions 2, 3, 4), price elasticity, in relation to density, is close to zero and thus is inelastic. In region 3, there is a negative effect of density on the

598

D. Lavee / Land Use Policy 42 (2015) 594–601

Table 1 Variable statistics. Variable

Mean

Minimum

Maximum

Definition

Price Density

280.468 (378.714) 4.27061 (1.817)

1.55 0.86

6562.68 12.25

Price per square meter in US$ Rate of change in the density of housing units per 1000 m2 according to NMP 35

Region 1 Region 2 Region 3 Region 4 Region 5 Region 6 Region 7 Region 8 Region 9 Region 10 Metropolitan

No. of 0 3642 3691 3372 3648 2711 3488 3454 3465 3376 3478 2953

No. of 1 135 86 405 129 1066 289 323 312 401 299 824

Socio-economic level 1–4 Socio-economic level 7–10 Saturated construction Construction for trade Ground level construction

2798 2427 2166 3581 2030

979 1350 1611 196 1747

Table 2 Results of the regression equation (3). Variable

Coefficient

ˇ0 + ˇj

Constant Density (ln) (Region 11) Region 1 Region 2 Region 3 Region 4 Region 5 Region 6 Region 7 Region 8 Region 9 Region 10 Metropolitan Socio-economic level 1–4 Socio-economic level 7–10 Saturated construction Construction for trade Ground level construction

2.664** (0.092) 0.087* (0.037) 2.037** (0.090) 0.109 (0.110) −1.052** (0.085) 0.744** (0.097) 1.482** (0.056) 2.044** (0.093) 2.027** (0.083) 2.061** (0.071) 0.798** (0.067) 0.847** (0.071) 0.680** (0.067) −0.369** (0.066) 0.603** (0.039) 0.597** (0.063) 0.901** (0.086) 0.323** (0.064)

– – 2.124 0.196 −0.965 0.831 1.569 2.131 2.114 2.148 0.885 0.934 – – – – – –

Adjusted R2 * **

0.617

Significantly different from zero at the 95% confidence level. Significantly different from zero at the 99% confidence level.

price, that is, as the density increases, the price drops. This may be explained by the fact that the city of Be’er Sheva differs from cities in the center of Israel, since it has large land reserves, in addition to a weak real estate market characterized by low land prices. Additionally, the surrounding areas are characterized by lower density and relatively high prices. In region 2, the result is not significant, since the differences between it and the base region (region 11) are not large. In both these areas there are few observations (since less real estate transactions were carried out), and they have similar features – a large area and low density, less cities and more agricultural communities. In the Center (regions 5, 6, 7, 8), the price is elastic in relation to density, as an increase of one percent in density will result in an increase of about 2 percent in price, without regarding other variables. In the North (regions 9, 10, 11) the price is inelastic relative to density, i.e., when density rises by one percent, the price goes up in less than one percent. Regarding the effect of the metropolitan variable, the result shows that there is a difference affecting the value of land between metropolitan and non-metropolitan. Since this variable is included in the constant, it is not affected by density, but rather the price

Eilat The Negev without Be’er Sheva Be’er Sheva, Omer, Meitar, lehavim Ashkelon district District of Ramle and Rehovot Jerusalem Tel Aviv Sharon district and Petah Tikva Haifa District Jezreel and Akko 1 for Be’er Sheva, Jerusalem, Tel-Aviv and Haifa. 0 for other.

starts at a higher point if the region is located in a metropolitan area. This result is in accordance with Sayag (2010). Also the socio-economic variable is not affected by density. According to the results, as the area is inhabited by a population with a higher socio-economic rating, the land starts at higher prices. This result is in accordance with Kok et al. (2011). The three remaining dummy variables regarding land designation (saturated, ground level or commercial construction) have an impact on the land value as well – if the area is designed for saturated construction, the land price will be higher. Apart from the density variable, all the other variables will be included in an appraiser assessment. In summary, the effect of density on land changes according to geographical location. In the center of Israel, the effect of density is the highest, in the North there is a weak positive effect and in the South, density hardly affects the price. These differences can be explained by the supply of land in the different areas. In the Center, the supply of land available for construction is limited, in the North the supply of available land is relatively high, while in the South the supply of available land is the highest. Methods for estimating land value The division into regions, as was made in the previous section, leads to biased results, since there is high variability within the regions. Therefore, for calculating the value of land designated for transportation projects, specific individual appraiser estimation should be performed for the relevant land, to prevent this bias. The appraiser’s estimation (in US$ per square meter) is multiplied by the multiplier found for each region and the expected change in density: Vl = P1 × [1 + Mi × %C]

(4)

where Vl is the land value, P1 is a specific appraiser value according to an individual assessment of the land by an appraiser presented as US$ per square meter, Mi is the multiplier according to the regression results, and %C is the percentage change in density, calculated by the change between the current density in the region and the future planned density (according to NMP 35). For each project, the following required data should be obtained: Land area: When examining the area on which the project is located, the land affected by the project should be examined as well. If there is residual land that cannot be used after the construction of

D. Lavee / Land Use Policy 42 (2015) 594–601

the project, or there is wayside affected by the project, they should be taken into consideration when estimating the land area. Location: The region of the land, as defined in the previous section, on which the project will be constructed. Alternative land designation: In order to consider what the alternative designations of the land are on which the project will be constructed, existing and future programs in the vicinity of the project should be examined. If there is an expectation for future expansion, the value of land should be discounted accordingly, to embody the current true value. An appraiser estimate: Preparing detailed appraiser estimation to value the current price of land, according to the relevant designations. Density: Calculating the current density level in the relevant area, by dividing the number of housing units by the number of dunams3 designated for residence, and determining the future density according to NMP 35. After collecting the required data, the density multiplier should be determined according to the specific region as given in Table 2. It should be mentioned that in cases where there are external effects on land value, such as proximity to focal points that may affect land value (e.g., sea view), it is appropriate to consider them during the valuation of land value. This, however, is out of the scope of this study. Case studies of applying the model in transportation projects in Israel In this section, we present two examples of implementing the methodology for estimating the value of land for major transport projects in densely populated areas in Israel: (1) a comparison of two alternatives for the Mesubim Interchange, and (2) estimating land value for Route 531. The two projects represent two types of dilemmas: the first project illustrates a construction project of an interchange, where one alternative, which is more expensive, can be constructed with a limited use of land, or a second less expensive alternative, but with a more intense use of land. The second project represents a major road that can be built above ground with intensive use of land, or, alternatively, can be constructed by an underground tunnel. Since these projects include alternatives that allow a more limited use of land, it is possible to perform a cost-benefit analysis between these options. However, to do so, at first it is required to evaluate the benefits of the alternative land as shown in this paper. Mesubim Interchange Mesubim Interchange connects Highway 4, one of the longest roads in Israel, with Route 461, one of the major access roads to Tel Aviv from the east, and serves as a southern entrance to the cities of Ramat Gan and Tel Aviv, as well as the entrance to the city Or-Yehuda. The Mesubim Interchange is built in a diamond interchange configuration and includes a bridge on Highway 4, which passes over Route 461, this enables the traffic on Highway 4 to flow without interruption. Right turns in the interchange are made without a traffic light, but left turns, as well as driving on Route 461, are directed by a traffic light. The transportation project examines two alternatives for upgrading Highway 4 in Mesubim Interchange: 1. An extended alternative – building a 3-level interchange that includes the shifting of the travel lanes of Highway 4 to the east, and the construction of two new bridges for this purpose. 2. A limited

3

1 dunam equals 1000 square meters.

599

alternative – expanding the interchange in the existing planned situation based on the existing route of travel lanes on Highway 4, and adding a third level. Data Land area: In this case, the only effected area is the area of the road and rights of way. The area size of the extended alternative is estimated at 275 dunam (275,000 m2 ). The estimated area size for the limited alternative is estimated at 86 dunam (86,000 m2 ). Location: Tel Aviv – Region 7. Alternative land designation: The land designated for construction is currently defined as land with an agricultural designation. However, the interchange is in close proximity to the “Ramat Ef’al” neighborhood in Ramat Gan, and already today, there is an interest in expanding the neighborhood towards Highway 4. Current expectation regarding the construction is that saturated construction of 8 units per dunam will be established. Discounting: Mesubim Interchange is adjacent to a residential neighborhood in the city of Ramat Gan. In the event that these lands were being marketed, it is very likely that this area would have been quickly occupied. Development time of design and marketing in the central regions of Israel is estimated at 10 years. That is, during this time buildings will be built and reach the target density as defined in NMP 35. Therefore, the land value should be discounted back 10 years. Current and future density: Current density in Ramat Gan is 6.8. According to NMP 35, the anticipated future density in this region is 12 (an increase of 76.4%). An appraiser estimate: For both alternatives, the appraiser land value is US$ 752 per square meter (saturated construction). Results The results of the estimation equation (Eq. (4)) for the two alternatives for Mesubim Interchange are presented in Table 3. The results enable the value of the land when assessing the alternatives to be taken into account. It should be noted that there are additional benefits, such as benefits from urban development and accessibility. When assessing the feasibility of a transportation project, the benefits arising from it should be examined against the costs in order to come to a decision as to which alternative to choose. Route 531 Route 531 is a suburban freeway in the southern Sharon region of Israel, crossing the southern Sharon from west to east and connecting between Highway 4 and Highway 6 (the Trans-Israel Highway). In the future, Route 531 is planned to continue west, – up to Highway 20 (Ayalon Highway), a major intercity freeway in Gush Dan. In this case study, we examine the planned road segment between Highway 4 meeting with Highway 20. Data Land area: In this case, the only effected area is the area of the road and rights of way. The estimated area size is 1250 dunam (1,250,000 m2 ). Location: Sharon District – Region 8. Alternative land designation: Route 531 has been listed in national master plans for many years, so we must examine the designation in the near vicinity. The majority of the road passes in close proximity to private homes, but part of the road passes by a saturated construction neighborhood. Discounting: The segment of Route 531 being examined passes through a desirable area. In the event that these lands were being marketed, it is very likely that this area would have been quickly occupied. As mentioned earlier, the land value should be discounted back 10 years.

600

D. Lavee / Land Use Policy 42 (2015) 594–601

Table 3 Calculating land value for Mesubim Interchange and Route 531. Price per m2 (US$)

Calculated land value per m2

Land area (m2 )

Project

Alternative

Region

Mesubim Interchange

Extended Limited

7 7

752 752

1966 1966

275,000 86,000

540,801,250 169,123,300

274,915,932 85,973,709

Route 531

–

8

1000

4651

1,250,433

5,816,514,142

2,956,820,847

Current and future density: Current density in the region is 3.7. According to NMP 35, the anticipated future density in the area is 10 (an increase of 170%). An appraiser estimate: The appraiser land value is US$ 1000 per square meter. Results The results of the estimation equation (Eq. (4)) for Route 531 are presented in Table 3. The results indicate that the calculated land value is extremely high, and cannot be ignored when assessing the feasibility of the project. Thus, consideration of land value may affect decisions regarding transportation projects. For example, it may cause a project to become not economically viable, or may possibly encourage reducing the area of land designated for the project, or lead to the feasibility of tunneling the project. Discussion Israel has an orderly procedure for evaluating transportation projects. This procedure takes into account many parameters such as the impact of the project on load transport, safety aspects, environmental protection, vehicle maintenance, etc. So far, policy makers failed to include in this procedure the cost of land in a proper manner. This paper aims to present policy makers a useful model that can be used for decision making. The results of the model and the findings obtained from the presented case studies, indicate that land value is indeed substantial and may have a significant impact on the decision-making process when selecting the optimal alternative for transportation projects, or may even lead to the cancelation of a project. Ignoring the true value of the land may lead to misguided decisions and a lack of attention to the consideration of reducing land use and as a result to intensive land use, which could have been prevented, in such projects. Nowadays, there is a struggle between real-estate entities and environmentalists regarding the preservation of open spaces. Realestate bodies argue that there is a shortage of land designed for housing, commerce and industry, and therefore the designation of open spaces should be changed for the benefit of urban development. Due to this dilemma, an educated use of limited land for transport projects, which are adjacent to existing uses, will increase the amount of land available for needed urban expansion and may delay expansion at the expense of open spaces and nature reserves. Summary and conclusions Transportation projects are characterized by high consumption of land, which is a valuable and a limited resource. Thus, it is important to internalize the true land value during a feasibility analysis of transportation projects, to allow the most suitable and efficient use of the land. Generally, feasibility analyses of transportation projects do not take into account land value. As transportation projects utilize land in the long-term, where the designation of these lands cannot be changed, it is necessary to address the future value of the land. This study presented a model for estimating the future value of land designated for transportation projects. The model is based on

Total price of land (US$)

Discounted value (US$) (to present value)

estimating the relationship between land value and density in the area in which the project is located, taking into account other characteristics as well. In particular, the elasticity of land price relative to density is estimated. Given the expected change in density and based on the estimated elasticity, the future land value can be estimated. This future value is to be discounted to the current period, and thus can be used during a feasibility analysis of transportation projects. The objective of this study is to allow a comparison between alternatives for transportation projects, and to take into account the economical land cost, thus ensuring that the chosen option will be the most economic alternative and prevent excessive use of land. The presented methodology was applied to two case studies of planned transportation projects in Israel. According to the case studies results, it can be seen that land constitutes a substantial cost, which cannot be ignored during a feasibility analysis of a transportation project. Adding the cost of land when examining the viability of transportation projects may lead to the examination of various alternatives, such as tunneling or reduced use of land, which may conserve precious land. Acknowledgements This paper is based on research funded by the Israel Ministry of Transport, Economics and Planning Division. The authors also would like to thank Hadas Joseph-Ezra and Sefi Bahar for comments and editing. References Alonso, W., 1964. Location and Land Use; Toward a General Theory of Land Rent. Harvard University Press, Cambridge, MA. Anas, A., Arnott, R., Small, K., 1997. Urban spatial structure. Working Papers No. 357. University of California Transportation Center. Beckmann, M.J., 1969. Distribution of urban rent and residential density. J. Econ. Theory 1 (1), 60–67. Belli, P., Anderson, J., Barnum, H., Dixon, J., Tan, J.P., 1998. Handbook on Economic Analysis of Investment Operations. Operational Core Services Network, Learning and Leadership Center, The World Bank, Washington, DC. Ben-David, D., 2000. Transportation infrastructure. Haaretz June 15 http://www.tau.ac.il/∼danib/articles/TransBarakEng.pdf Booz Allen Hamilton, 2005. Surface Transport Costs and Charges Study. Ministry of Transportation, New Zealand www.transport.govt.nz/current/issues Boucher, M., 1996. Highway costs and revenues in Quebec: evidence and analysis. In: Proceedings, CTRF Annual Meeting. Braid, R.M., 1995. Use of land for roadways in a growing Mills-de Ferranti urban area. J. Urban Econ. 37, 131–160. Cadwallader, M., 1996. Urban Geography: An Analytic Approach. Prentice Hall, Upper Saddle River, New Jersey. Delucchi, M., 1998. Motor Vehicle Infrastructure and Services Provided by the Public Sector. Annualized Social Cost of Motor-Vehicle Use in the U.S., 1990–1991, vol. 7. Institute of Transportation Studies, UCD-ITS-RR-96-3 (7). Delucchi, M., Murphy, J.J., 2005. Motor Vehicle Infrastructure and Services Provided by the Public Sector. Report #7 in the series: The Annualized Social Cost of Motor-Vehicle Use in the United States, based on 1990–1991 Data. Institute of Transportation Studies, UCD-ITS-RR-96-3 (7) rev. 2. Duranton, G., Puga, D., 2004. Micro-foundations of urban agglomeration economies. In: Henderson, J.V., Thisse, J.F. (Eds.), Handbook of Regional and Urban Economics, vol. 4, 1st ed. Elsevier, Amsterdam, pp. 2063–2117 (Chapter 48). FAO, 2007. Land evaluation: Towards a revised framework. Land and Water Discussion Paper 6. FAO, Rome. Goodbody Economic Consultants, 2004. Parameter Values for Use in Cost-Benefit Analysis of Transport Projects. Ballsbridge Park, Ballsbridge, Dublin.

D. Lavee / Land Use Policy 42 (2015) 594–601 Grimes, A., Liang, Y., 2007. Spatial determinants of land prices in Auckland: Does the metropolitan urban limit have an effect? Motu Working Paper 07-09. Haritos, Z., 1973. Rational Road Pricing Policies in Canada. Canadian Transport Commission, Ottawa. Haritos, Z.J., Shlachter, J., 1982. Transport costs and revenues in Canada, Ottawa. Log. Transport. Rev. 18 (4), 323–337. Haruvy, N., Shalhevet, S., 2004. Optimal allocation of agricultural crops in the open spaces in Israel: economics, financial and environmental considerations. J. Finan. Manage. Anal. 17 (2), 103–107. Hirshhorn, R., 2003. Concepts and Practical Values of Land Costs and Capital Charges for a “Full-Cost Accounting” of Transport Infrastructure in Canada. Prepared for Economic Research Branch, Transport Canada. Israeli Ministry of Finance, 2011. Economic policy and fiscal policy for the financial years 2011–2012. http://www.mof.gov.il/BudgetSite/StateBudget/ Budget2011 2012/Lists/List2/Attachments/2/mediniut.pdf Ketcham, B., Komanoff, C., 1992. Win-Win Transportation: A No-Losers Approach to Financing Transport in New York City and the Region. KEA, New York. Kok, N., Monkkonen, P., Quigley, J.M., 2011. Economic geography, jobs, and regulations: the value of land and housing. Program on Housing and Urban Policy Working Paper No. W10-005. University of California, Berkeley. Lall, A., 1992. Transportation infrastructure costs in Canada. RCNPT, Ottawa. Lavee, D., Baniad, G., 2013. Assessing the value of non-marketable land: the case of Israel. Land Use Policy 34 (1), 216–281. Litman, T.A., Doherty, E., 2009. Transportation Cost and Benefit Analysis: Techniques, Estimates and Implications, 2nd ed. Victoria Transport Policy Institute, Victoria, British Columbia. Litman, T., 2004. Transportation Cost and Benefit Analysis. Victoria Transport Policy Institute, Victoria, British Columbia. Litman, T., 2005. Transportation Land Valuation: Evaluating Policies and Practices that Affect the Amount of Land Devoted to Transportation Facilities. Victoria Transport Policy Institute, Victoria, British Columbia. Litman, T., 2012. Transportation Land Valuation – Evaluating Policies and Practices that Affect the Amount of Land Devoted to Transportation Facilities. Victoria Transport Policy Institute, Victoria, British Columbia. Mansour-Moysey, N., Semmens, J., 2001. Value of Arizona’s State Highway System. Transportation Research 1747. Transportation Research Record, ISSN 03611981. Newbery, D.M., 1995. Royal Commission report on Transport and the Environment – economic effects of recommendations. Econ. J. 105 (432), 1258–1272.

601

OECD, 2008. Estimation of the stock of land in OECD countries. KIM Young-Hwan National Accounts Division, Statistics Directory. In: Working Party on National Accounts, 14–16 October 2008. Papageorgiou, G.J., Casetti, E., 1971. Spatial equilibrium residential land values in a multicenter setting. J. Region. Sci. 11, 385–389. Poole, R., 1997. Privatization: a new transportation paradigm. Ann. Am. Acad. Pol. Soc. Sci. 553, 94–105. Rodrigue, J.P., 2013. The Geography of Transport Systems, 3rd ed. Routledge, New York. Roth, G., 1996. Roads in a Market Economy. Avebury Technical, Ashgate Publishing, England, UK. Sayag, D., 2010. Measuring the local house price movements and estimating the price elasticity. Working Paper Series No. 51. The Israeli Central Bureau of Statistics and the Hebrew University of Jerusalem. Shoshany, M., Goldshleger, N., 2002. Land-use and population density changes in Israel – 1950 to 1990: analysis of regional and local trends. Land Use Policy 19 (2), 123–133. Sofer, A., 2013 June. More roads, more congestion – regarding the budgeting policy of transport infrastructure in Israel, 2000–2012. Adva Center (in Hebrew) http://www.adva.org/uploaded/transport%202013.pdf Solow, R.M., 1972. Congestion, density and the use of land in transportation. Swedish J. Econ. 74 (1), 161–173. Sun, L., Tu, Y., 2005. Urban land value with site density restriction: an empirical study of Beijing, China. In: Paper Submitted for AREUEA Annual Conference, Boston, USA, January 2006. The State of Israel, 2012. The Transportation Project Appraisal Procedure (Nohal Prat) 2012. Transportation Project Appraisal Guidelines, Jerusalem. TRIP (Transport Research on Environmental and Health Impacts and Policy), 2004. Policy and Project Appraisal at Regional Level. Centre for Transport Research on Environmental and Health Impacts and Policy, Denmark, Copenhagen. Vickrey, W., 1997. Public Economics; Selected Papers by William Vickrey. Cambridge University Press, UK. Victoria Transport Policy Institute, 2009. Transportation Cost and Benefit Analysis II – Roadway Land Value. Victoria Transport Policy Institute, Canada http://www.vtpi.org/tca/tca0507.pdf Woudsma, C.G., Litman, T., Weisbrod, G., 2006. A report on the estimation of unit values of land occupied by transportation infrastructures in Canada. Final Report, submitted to Transport Canada, Ottawa.