Financing Residential Development with Special Districts

Stephen Billings University of North Carolina-Charlotte Thomas G. Thibodeau University of Colorado July 23, 2009

Preliminary Draft - please do not cite or circulate without permission. Acknowledgements: This research has beneted from comments by Joseph Nichols and Liang Peng. We thank Erika Gleason and Brady Miller for their outstanding research assistance. All errors are our own.

University of North Carolina, 9201 University City Blvd. Charlotte, NC 28223. Email: stephen.billings@uncc.edu Leeds School of Business, University of Colorado, 419 UCB, Boulder, CO 80309. Email: tom.thibodeau@colorado.edu

Phone: (704) 687-4879, Phone: (303) 735-4021,

Financing Residential Development with Special Districts

Abstract There are four mechanisms typically used to nance the development of residential infrastructure: municipal bond nancing (with bonds securitized by the municipalitys general revenues), the developer (either using construction loans, with equity, or some combination of both), development impact fees, and special district bonds. Special districts are created by property owners to provide specic public services. They are typically governed by a board of directors and have the authority to levy taxes, charge user fees and issue debt. Financing public services, including the development of residential infrastructure, using special districts has become a fairly common practice in some parts of the United States. The Special District Association of Colorado, for example, reports that there are over 1,800 special districts currently operating in the state. This paper empirically examines the extent to which the property tax liability created by nancing residential infrastructure using special district bonds is capitalized in house prices. We compare house prices for single-family detached homes built within development districts to similar properties located outside development districts. Our data consists of over 34,000 transactions of recently built homes sold in the Denver metropolitan area during the 2002 to 2004 period. Our hedonic house price specication includes the usual structural characteristics and controls for the inuence of spatial attributes using Census Block Group neighborhood xed eects. The preferred empirical specication restricts the data to transactions within Census block groups that have numerous sales of recently constructed single-family detached homes located both within and outside development districts. Our results indicate that house prices for homes located within development districts are lower than house prices for similar homes located outside of development districts, but the amount of property tax capitalization is signicantly less than full. Our results also attribute some of the wide variation in estimates of property tax capitalization reported in the literature to a failure to adequately control for neighborhood attributes.

JEL Classication: R3, R5 Keywords: nancing infrastructure, special districts, property tax capitalization

Introduction

Special districts1 are local governments formed to provide a variety of services to residents. They are created by property owners, governed by a board of directors or supervisors, have the authority to issue debt securitized by expected sales and property tax revenues as well as by anticipated tolls, user fees, tap and other impact fees.2 Special districts are playing an increasingly important role in nancing residential infrastructure. Extending water, sewer, and drainage facilities to undeveloped parts of a jurisdiction can be very expensive. Historically, residential infrastructure has been nanced using cost-sharing methods, like general obligation municipal revenue bonds, municipal impact fees or the developers funds (using some combination of short-term construction loans and the developers equity). Cost-sharing methods typically involve bond nancing by an existing general purpose municipal or county government. These general obligation bonds are repaid by all property owners in the jurisdiction. With cost-sharing, the consumers of the new infrastructure are subsidized by all taxpayers in the jurisdiction. Impact fees are upfront fees paid by the developer to cover infrastructure costs. These fees increase the cost of developing residential lots, increase lot prices and eventually increase house prices (relative to cost-sharing alternatives). Some developers have the capital capacity to nance infrastructure development themselves. Residential infrastructure can be nanced with short-term construction loans, with the developers own funds, or some combination of both. Special districts oer local governments and developers the opportunity to nance residential infrastructure with relatively inexpensive long-term debt while passing the obligation to repay that debt directly onto the ultimate consumerthe homeowner. Special districts in Colorado were rst authorized by Title 32 of the Colorado Revised Statute to extend public services to rural and unincorporated parts of the state. They became increasing popular in Colorado following the passage of the Taxpayers Bill of Rights (TABOR) in 1992. TABOR limits the growth in government revenues to the sum of the population growth rate plus the rate of ination. Any increase in government revenues over this limit must either be refunded to taxpayers or approved
Special districts are also referred to as special service districts, special purpose districts, limited purpose districts, municipal development districts and municipal utility districts. 2 For more information on special districts, see Galvan (2007) and Grith (2007).
1

by a popular vote in a general election. TABOR applies to the entire State of Colorado as well as to all counties and municipalities within the state. TABOR has made it exceedingly dicult for local jurisdictions to nance residential development using general obligation bonds, as increasing taxes above the TABOR limit would require the approval of a majority of all residents. What eect does development district infrastructure nancing have on house prices? We expect Colorado house prices for homes built in development districts to be lower than house prices for similar homes built outside development districts. Since TABOR cost-sharing methods, like municipal bond nancing, are no longer a viable nancing vehicle in Colorado. The primary ways to nance the construction of residential infrastructure are either directly by developers or indirectly using development districts. When developers nance the construction of residential infrastructure, they will be reimbursed for this expense when the developed lots are sold. Higher prices for improved lots generate higher prices for completed homes. Alternatively, when development district bonds are used to nance infrastructure, repaying this debt is passed onto homeowners. This future property tax liability should be (negatively) capitalized in house prices. The choice between direct developer and special district nancing is idiosyncratic to individual developers. Specically, the use of development district bonds will depend on a developers understanding of the process for forming a special district, prevailing interest rates on both construction loans and special district bonds as well as current liquidity constraints and tax liabilities. This paper makes three contributions to the literature. First, it examines the extent to which one component of residential property taxesthe development district taxis capitalized into the price of new homes. The existing literature on tax capitalization examines how the entire property tax is capitalized in the price of all housing (both new and existing). Second, most of the existing capitalization literature examines the extent to which the property tax liability associated with ongoing public services (e.g. education, police and re protection, etc.) is capitalized in house prices. Property taxes that pay for local public services may be positively capitalized in house prices, negatively capitalized, or have no inuence on house prices depending on how taxpayers value the benets generated by those services.3 Development district bonds have nite maturities
3

See Billings and Thibodeau (2009) for more discussion on measuring property tax capitalization for net benets.

and nite repayment obligations. This paper examines how that nite-term obligation is capitalized in house prices. Finally, we empirically illustrate the importance of controlling for spatial attributes by estimating parameters for several hedonic house price specications that impose increasingly more restrictive controls for neighborhood amenities. We test for dierences across specications using a spatial application of the Wald test. Our empirical results attribute the wide variation in property tax capitalization rates reported in the literature to a failure to adequately control for neighborhood amenities. The remainder of this paper has ve sections. Section 2 provides a brief review of the relevant literature. Section 3 provides an overview of how special districts are created and provides information on the debt issued by some Denver area development districts. Section 4 describes our empirical approach and data. Sections 5 and 6 provide results and concluding remarks.

Literature Review

There are three branches of literature that are relevant for this research: (1) the literature that compares cost sharing techniques to land use extraction methods for nancing residential development; (2) the literature that examines whether land use extraction methods, like development or impact fees, inuence land and house prices; and (3) the property tax capitalization literature. Brueckner (1997) theoretically compares impact fees to two cost-sharing techniques for nancing development: one where all taxpayers in a jurisdiction pay the infrastructure cost at the time the cost is incurred and a second where all taxpayers nance the cost. He concludes that since impact fees are paid by the eventual consumer (e.g. the homeowner), impact fees are more ecient than cost-sharing nancing techniques. He also concludes that impact fees generate slower rates of growth and yield higher aggregate property values relative to cost-sharing methods. Peiser (1983) compares patterns of residential development in two Texas cities: Houston, a city that uses local municipal utility districts (MUDs) extensively, and Dallas, a city that relies on much larger, regional utility districts (RUDs) for nancing residential development. Peiser argues that economies of scale make RUDs more ecient than MUDs in the long run, but that RUDs also constrain the supply of developable land and increase land prices in the short run. He argues that MUDs permit developers 5

to use less expensive land to develop lots and market competition forces developers to pass those cost savings onto homebuyers. Finally, he concludes that MUDs are particularly eective at supporting rapid, short-term, residential development. A number of papers have empirically examined the inuence that land use extraction fees have on land and house prices.4 Delaney and Smith (1989) compare house prices in Dunedin, Florida, a city that began using impact fees to nance residential infrastructure in June of 1974, to house prices in Clearwater, Largo, and St. Petersburgh, three Florida cities that did not implement impact fees. They use a hedonic house price model to estimate hedonic coecients in these places and then price a xed bundle of housing characteristics across cities. They conclude that the Dunedin impact fee increased house prices by an amount that signicantly exceeded the fee. Singell and Lillydahl (1990) report that house prices in Loveland, Colorado increased by 7% after that city imposed impact fees in July of 1984. Yinger (1998) notes that the Delaney and Smith (1989) study may have overestimated the eect of impact fees because the authors held land prices constant when they priced housing characteristics across the four markets. If impact fees are anticipated by developers, then developers will reduce the price they will pay for raw land. Yinger (1998) also noted that both the Delaney and Smith (1989) study and the Singell and Lillydahl (1990) study did not adequately control for the inuence that neighborhood attributes have on house prices. Using data on undeveloped land prices and prices for new and existing homes in Dade County, Florida, Ihlanfeldt and Shaughnessy (2004) report that a dollar of impact fee reduces land price by an equivalent amount and increases the price of both new and existing housing by $1.60. Beginning with Oates (1969), there are numerous papers that examine whether property taxes are capitalized in house prices.5 Yinger et al. (1988) summarizes the econometric approaches, data (e.g. aggregate vs. micro) and resulting property tax capitalization rates reported by thirty empirical studies of capitalization published prior to 1988. Based on dierent methodologies, types of local governments (e.g. municipalities, school districts, special districts), study areas, the assumption of a 3% discount rate and an innite time horizon of the tax liability, estimates range
See Evans-Cowley and Lawhon (2003) for a more complete review of this literature. See Brueckner (1979), Starrett (1981), Yinger (1982), Yinger et al. (1988), and Zodrow (2006) for analysis and additional references on property tax capitalization.
5 4

from 15% capitalization to 120%. Palmon and Smith (1998) empirically examine tax capitalization using data on MUDs in Houston, Texas. They argue their data overcomes several deciencies in previous studies because the level of public services is held constant across districts but tax rates vary considerably. They nd a 62% property tax capitalization rate for Municipal Utility Districts in Houston.

Development Districts

In Colorado, special districts are local governments created under Article 1 of Title 32 of the Colorado Revised Statute to provide one of the following services: water facilities or services, sanitation facilities or services, ood control, streets, roads, alleys, walkways, transit, parks and recreation facilities, golf courses, re protection, security, insect and animal control, libraries, emergency medical, hospitals, soil and water conservation, television relay and transmission facilities or open space. Metropolitan districts provide two or more of these services. Real estate developers frequently use special/metropolitan districts to nance the construction of the infrastructure needed to support vertical residential and commercial development. Special/metropolitan districts are created in four steps. First, district property owners must petition their local government(s) (e.g. municipal, county) to estabish a district. If the petition is approved, the districts governing body, usually a board of directors elected by property owners, must submit a service plan to all counties and municipalities that have jurisdiction within the districts boundaries. The service plans typically include district maps and agreements that explicitly describe which jurisdiction (e.g. county, municipality, or service district) will provide what services. Third, the proposed district must be approved in an election of residents owning property in the district by either 20% of the eligible voters or by 200 voters, whichever is smaller. Finally, once the special/municipal district is approved by district taxpayers and by all local governments that have jurisdiction, the district is led with the local court and with the county assessor. Once the district is led with the local court, it may issue bonds. Special/metropolitan districts created to nance infrastructure usually have nite lives. We dene a development district as a special/metropolitan district that contains no single7

family properties prior to the creation of the district and provides at least one of the following facilities or services: public improvement, street maintenance, public works, water facilities, sewer, storm drainage or transportation. Table 1 lists 45 high growth development districts created in the Denver metropolitan area between 1980 and 2002 that have issued bonds to nance the construction of residential infrastructure.6 A development district is labeled high growth if it is located in a Census Block Group (CGB) that had at least twenty recently built single-family homes sold between 2002 and 2004 and those twenty properties represented at least twenty percent of the year 2000 CBG housing stock. The table provides the district name, location, size (in square miles), the year the district was created and the development district property tax rate. Nearly three-quarters (33 of 45) of the development districts are located in Douglas and Adams Counties, two rapidly growing suburban counties located south and east of the City/County of Denver. Development district sizes range from 0.35 square miles (224 acres) to over 9 square miles. The average district size is 2.65 square miles. About half (22 of 45) of the development districts were created in the 1980s, 19 were created in the 1990s and 4 were created this decade. The across development district property mill levy in 2003 was 35.7 mills and development district property tax rates ranged from 6 to 90 mills. Table 2 provides descriptive information on the bonds issued by these development districts. The table includes the bond CUSIP, the issue date, maturity, term, the aggregate amount of the bond (or bonds), whether the coupon rate was xed or variable, and, for xed rate bonds, the interest rate. While all of the development districts in Table 1 issued bonds to nance infrastructure, some of these bonds were privately placed. Consequently, descriptive information for these district bonds is not publicly available and Table 2 provides information for the 33 districts that sold bonds to the public. Some of the districts issued multiple bonds. For these districts, the CUSIP, issue date and maturity refer to the most recent issue prior to 2004 while the bond amount is the total amount of all bonds issued by the district. Bond amounts range from $1.2M to $72.2M. The average amount issued was $12.2M. Bond maturities range from six to forty years. Seventy-ve percent of the bonds had maturities between twenty and thirty years. The average maturity was twenty-four years. Most
6 There were ve special/metropolitan districts that were classied as development districts but were excluded from Table 1 because they did not issue any bonds as of 2004.

of the bonds paid a xed coupon. The average coupon was 6.87%. Any concern that the use of development districts is determined by the attributes of a residential development (e.g. land area, number of housing units or infrastructure costs) is addressed in Table 1, which highlights a range of development districts in terms of both size and infrastructure costs. For some districts, there is a big disparity between the date the district is created and the time the bonds are issued. Development district bonds are issued just prior to (and sometimes just after) the infrastructure development begins. Once the bonds are issued, the repayment clock starts and properties must be built, sold, registered on the assessors le and begin generating property tax revenues to repay the debt. Development district bonds are typically created with interest reserve accounts adequate to pay interest for the rst few years of the bond. Interest reserve accounts may be created with terms of up to three years. As an example, the Brighton Crossing Metropolitan District was created in 1985. The development district covers 7.78 square miles in Adams County, Colorado. Within this area, a developer has proposed a residential development encompassing 771 acres. The developer plans to build 3,555 single-family homes on 547 acres and 927 multifamily units on 91 acres. The developer allocated 108 acres to parks, trails and recreation facilities and 25 acres to streets and walks. The water, sewer and ood control infrastructure for this development costs $45.7M. The district issued its rst bond for this development in December 2004twenty years after the district was created. The bond amount was $13.8M and the bond maturity was 30 years. The development district tax rate is 38 mills.

Empirical Approach

This paper empirically examines the eect that nancing residential infrastructure using development districts has on house prices. We employ a hedonic house price model to examine the extent to which development district property taxes are capitalized in house prices.7 We begin with our
Yinger et al. (1988) and Palmon and Smith (1998) incorporate a non-linear specication for estimating property tax capitalization. While this specication is well motivated, it requires assumptions or estimates of annual user costs/rental value for a property. The unavailability of rental data for a comparable stock of new homes limits incorporation of this technique.
7

Table 1: Metropolitan Denver Development Districts

Nearest/in City/Town Aurora Commerce City Brighton Broomeld Brighton Commerce City Commerce City Parker Parker Parker Parker Aurora Commerce City Aurora Denver Aurora Foxeld Denver Littleton Lone Tree Littleton Broomeld Parker Castle Rock Castle Rock Castle Rock Castle Rock Castle Rock Commerce City Parker Castle Rock Commerce City Commerce City Littleton Aurora Aurora Parker Louisville Aurora Aurora Thornton Thornton Denver Parker Castle Rock Area (SqMi) 4.21 2.30 7.78 0.39 6.90 2.30 2.30 1.39 0.64 2.57 0.35 0.76 2.30 0.49 0.47 9.10 0.80 5.36 4.40 4.54 4.37 1.16 0.59 2.81 0.79 1.73 0.73 1.73 2.30 1.47 3.37 2.30 2.30 1.84 4.72 0.60 0.77 1.79 9.33 9.33 1.95 1.95 0.88 0.70 3.37 2.65 Date District Created 1998 2000 1985 1997 1985 1996 2002 1996 2000 1985 1998 1998 1996 1994 1982 1998 1994 1983 1980 1980 1980 1994 1983 1987 1985 1985 1985 1985 1999 1985 1998 2001 1996 1985 1985 1994 1983 1988 1998 1998 1994 1996 1983 1985 1984 1994 2003 Property Mill Levy 42.6 47.9 38.0 20.0 38.0 37.3 42.0 39.5 37.3 39.8 45.2 45.3 50.0 17.0 39.0 10.0 26.5 30.3 20.3 20.3 20.3 27.2 45.0 40.0 35.0 35.0 35.0 35.0 38.5 24.0 45.0 38.0 45.0 70.1 90.0 30.0 27.4 22.0 49.0 49.0 16.5 18.0 39.0 6.0 40.0 35.7

Development District Name Aurora Single Tree Metropolitan District Belle Creek Metropolitan District Brighton Crossing Metropolitan District No. 4 Broadlands Metropolitan District No. 1 Bromley Park Metropolitan District No. 2 Bualo Ridge Metropolitan District Bualo Run Mesa Metropolitan District Canterberry Crossing Metropolitan District Canterberry Crossing Metropolitan District II Cherry Creek South Metropolitan District No. 1 Compark Business Campus Metropolitan District Eagle Bend Metropolitan District No. 2 Eagle Creek Metropolitan District East Smoky Hill Metropolitan District No. 2 Ebert Metropolitan District Gateway Regional Metropolitan District Goodman Metropolitan District GVR Metropolitan District Highlands Ranch Metropolitan District No. 1 Highlands Ranch Metropolitan District No. 3 Highlands Ranch Metropolitan District No. 4 Interlocken Consolidated Metropolitan District Lincoln Park Metropolitan District Maher Ranch Metropolitan District No. 4 Meadows Metropolitan District No. 1 Meadows Metropolitan District No. 2 Meadows Metropolitan District No. 6 Meadows Metropolitan District No. 7 North Range Village Metropolitan District Parker Properties Metropolitan District No. 1 Pinery West Metropolitan District No. 2 Potomac Farms Metropolitan District Riverdale Dunes Metropolitan District No. 1 Roxborough Village Metropolitan District Second Creek Ranch Metropolitan District Sterling Hills Metropolitan District Stonegate Village Metropolitan District Superior Metropolitan District No. 2 Tallyns Reach Metropolitan District No. 2 Tallyns Reach Metropolitan District No. 3 Todd Creek Farms Metropolitan District No. 1 Todd Creek Farms Metropolitan District No. 2 Town Center Metropolitan District Upper Cherry Creek Metropolitan District Villages At Castle Rock Metropolitan District Average

County Adams Adams Adams Adams Adams Adams Adams Douglas Douglas Douglas Douglas Arapahoe Adams Arapahoe Denver Adams Arapahoe Denver Douglas Douglas Douglas Boulder Douglas Douglas Douglas Douglas Douglas Douglas Adams Douglas Douglas Adams Adams Douglas Denver Arapahoe Douglas Boulder Arapahoe Arapahoe Adams Adams Denver Douglas Douglas

10

Table 2: Development District Bonds

Rate 8.000% 8.000%

11

Development District Name Aurora Single Tree Metropolitan District Belle Creek Metropolitan District Brighton Crossing Metropolitan District No. 4 Broadlands Metropolitan District No. 1 Bualo Ridge Metropolitan District Canterberry Crossing Metropolitan District Canterberry Crossing Metropolitan District II Cherry Creek South Metropolitan District No. 1 Eagle Bend Metropolitan District No. 2 Eagle Creek Metropolitan District East Smoky Hill Metropolitan District No. 2 Gateway Regional Metropolitan District Goodman Metropolitan District GVR Metropolitan District Highlands Ranch Metropolitan District No. 1 Highlands Ranch Metropolitan District No. 3 Highlands Ranch Metropolitan District No. 4 Interlocken Consolidated Metropolitan District Lincoln Park Metropolitan District Maher Ranch Metropolitan District No. 4 Meadows Metropolitan District No. 1 Meadows Metropolitan District No. 2 North Range Village Metropolitan District Pinery West Metropolitan District No. 2 Potomac Farms Metropolitan District Riverdale Dunes Metropolitan District No. 1 Roxborough Village Metropolitan District Sterling Hills Metropolitan District Stonegate Village Metropolitan District Superior Metropolitan District No. 2 Todd Creek Farms Metropolitan District No. 1 Todd Creek Farms Metropolitan District No. 2 Upper Cherry Creek Metropolitan District Average

Bond CUSIP 05206NAA8 078341AA2 109330AA8 11132NAA5 119802AA4 138100AE2 138101AB6 164561AC6 269422AA9 269492AA2 275292AD2 36779TAA7 38238EAA1 362374CL2 431096EW2 430905BY4 430903BB9 458761BE2 534550AB4 559827AB6 582903BU3 582903BW9 661766AA3 723295AA9 73768TAA3 768643AA6 779814BG6 85933LAB2 861819BJ8 86820PAK7 88901POA2 889013AS7 915618CC4 Maturity 11/15/2025 12/1/2020 12/1/2034 8/1/2018 12/1/2033 12/1/2031 12/1/2032 12/15/2033 12/1/2018 6/1/2021 12/1/2030 12/1/2010 12/1/2019 12/1/2019 9/1/2012 12/1/2019 12/1/2017 12/15/2019 12/1/2026 12/1/2027 6/1/2029 6/1/2029 12/1/2020 11/1/2032 12/1/2032 12/1/2032 12/31/2021 6/1/2018 12/1/2025 12/1/2013 12/1/2021 12/1/2018 12/1/2011

Issue Date 10/5/2000 10/1/2000 12/17/2004 8/6/1998 12/1/2003 12/27/2001 12/1/2002 5/15/2003 10/1/1999 11/1/2001 8/31/2000 5/15/2000 6/1/1999 12/15/1999 6/1/1997 6/1/1999 4/1/1997 11/15/1999 6/29/2001 6/1/2003 9/1/1989 9/1/1989 8/1/2000 12/30/2002 3/1/2002 1/1/2002 9/1/1993 11/1/1998 11/1/1996 6/18/1998 11/16/2001 6/1/2003 5/15/1996

Term (Years) 25 20 30 20 30 30 30 30 19 20 30 10 20 20 15 20 20 20 25 25 40 40 20 30 30 30 28 20 29 15 20 15 15 24

Rate Fixed/ Amount $1,575,000 $5,215,000 $13,800,000 $4,925,000 $10,130,000 $10,430,000 $7,500,000(a) $10,060,000(b) $7,500,000 $2,000,000 $9,600,000(c) $1,235,000 $5,000,000 $13,165,000(d) $19,415,000(e) $13,075,000(f) $5,390,000(g) $72,215,869(h) $16,610,000 $13,100,000(i) $30,730,000 $15,440,000 $2,800,000 $21,815,000 $4,315,000 $3,800,000 $6,247,629 $2,000,000 $42,330,000(j) $9,190,000 $12,500,000 $2,755,000(k) $1,705,000(l)

Interest Variable Fixed Fixed Variable Variable Fixed Fixed Fixed Fixed Fixed Fixed Fixed Fixed Fixed Fixed Fixed Fixed Fixed Fixed Fixed Fixed Fixed Fixed Fixed Fixed Fixed Fixed Not Reported Fixed Fixed Fixed Fixed Fixed Fixed

7.500% 6.700% (a) (b) 6.875% 6.750% (c) 6.500% 5.000% (d) (e) (f) (g) (h) 7.750% (i) 7.990% 7.990% 8.000% 3.700% 6.850% 6.500% Not Reported 7.750% (j) 4.625% 5.000% (k) (l)

12

Notes for District Bonds Table: a. This amount is the result of two bonds [$2,500,000 at 7.00% and $5,000,000 at 7.375%] b. This amount is the result of two bonds [$3,511,000 at 1.4% and $6,549,000 at 1.4%] c. This amount is the result of three bonds [$1,430,000 at 3.8%, $3,390,000 at 4.90%, and $4,780,000 at 5.00%] d. This amount is the result of 15 bonds ranging from $420,000 to $4,630,000 with interest rates ranging from 4.05% to 5.75% e. This amount is the result of 15 bonds ranging from $470,000 to $5,005,000 with interest rates ranging from 4.5% to 5.75% f. This amount is the result of 11 bonds ranging from $465,000 to $4,890,000 with interest rates ranging from 4.25% to 5.3% g. This amount is the result of 14 bonds ranging from $155,000 to $2,310,000 with interest rates ranging from 4.2% to 6.3% h. This amount is the result of 11 bonds ranging from $205,000 to $46,469,778 with interest rates ranging from 5.0% to 5.75% i. This amount is the result of 3 bonds ranging from $820,000 to $6,245,000 with interest rates ranging from 7.7% to 7.875% j. This amount is the result of 16 bonds ranging from $400,000 to $10,660,000 with interest rates ranging from 4.35% to 5.6% k. This amount is the result of 16 bonds ranging from $120,000 to $250,000 with interest rates ranging from 2.0% to 3.75% l. This amount is the result of 5 bonds ranging from $10,000 to $1,005,000 with interest rates ranging from 5.375% to 6.75%

hedonic house price model. Let the natural log of the price for house i in period t be represented by:
j ln(Pi,t ) = + Xi + Ni + +Gi + Ti + j=1 Di,j + i,t

(1)

where Pi,t = the transaction price for house i in period t; Xi = a vector of structural characteristics for property i; Ni = a vector of neighborhood characteristics for property i; Gi = a vector of government characteristics for property i; Ti = the property tax mill rate; Di,j = 1 if property i sold in period j (j = 1, 2,. . . , nj ), and equals 0 otherwise; = the hedonic equation intercept; = a vector of structural characteristic hedonic coecients; = a vector of neighborhood characteristic hedonic coecients; = the vector of government characteristic hedonic coecients; = the hedonic coecient for the property tax rate; = hedonic coecients for sale period j binary variable (j = 1, 2,. . . , nj ); and i,t = error term for property i in period t. The structural characteristics include the usual variables: lot size, square feet of living space, number of bathrooms, etc.8 Since many of the neighborhood amenities that inuence house prices are not observed, we control for spatial variation in neighborhood characteristics using Census Block Group (CBG) xed eects.9 The incorporation of neighborhood level xed eects control for any neighborhood characteristic that is observable or unobservable (e.g. type and density of land use, distance to the Central Business District (CBD) and household incomes). This limits identication to only intra-CBG variation. Controls using neighborhood eects is well suited for examining infrastructure nancing districts because they lack potential spillovers across development district boundaries. The roads, curb, gutter, water, and sewer constructed within a development district are
The hedonic specication excludes dwelling age since all of the homes are new. As shown by Thibodeau (2003), segmenting single-family home markets into smaller geographic areas controls for spatial autocorrelation in neighborhood attributes.
9 8

13

typically not accessed or used by neighboring residential developments. Given the large number and spatial variation of local governments in Colorado highlighted by Billings and Thibodeau (2009), we also include controls for other, non-development district, special districts and municipalities that provide services. Our data consists of 34,048 transactions of recently built single-family homes sold in the Denver metropolitan area. Properties were recently built and subsequently sold during the 2002-2004 period. The Denver-Boulder-Greeley Consolidated Metropolitan Statistical Area (CMSA) consists of the City/County of Denver, its bedroom communities, and nearby employment centers.10 A transaction is classied as a new property if it was less than or equal to three years old at the time of sale. In addition, the data is limited to properties situated on between 0.05 and 5 acre lots and with transaction prices between $10, 000 and $1, 000, 000. All data on special districts is current as of 2003 and the original government dataset was provided by the Department of Local Aairs, State of Colorado in shapeles for jurisdictional boundaries and electronically for mill levies. It was further supplemented with paper records from les in the Denver, CO oce. All Geographical Information Systems datawork is implemented using the Colorado Central Zone State Plane NAD83 Projection. According to the Colorado Department of Local Governments, there were 672 special districts and 377 metropolitan districts in the Denver metropolitan area in 2004. We classied 84 of these as development districts. Forty-four of these development districts were created since 1993 and the remaining 40 development districts pre-date 1993. During the twenty year period prior to TABOR, development districts in Colorado were created at an annual rate of 1.9 per year. Following TABOR, development districts were created at the annual rate of 4.4 districts per year. More recently created development districts are likely not fully captured in the dataset due to the time lag between the creation of a development district and the sale of a new home. Since property transactions and development districts are based on the entire urban area, the observations included in estimation represent the full choice set for households in the market for
Denver and Broomeld are both cities and counties, integrated into a single government institution and treated as a county in this dataset. The Denver-Boulder-Greeley CMSA consists of Adams, Arapahoe, Boulder, Broomeld, Denver, Douglas, Jeerson, and Weld counties.
10

14

new homes in the Denver metropolitan area. This represents an improvement over existing studies, which commonly limit hedonic estimates to a single county (see Yinger et al. (1988)). The use of a single county has the potential to generate varying elasticities of demand and property tax capitalization depending on the housing stock and on the property tax burden of residential units available in neighboring counties within the same metropolitan area. Ignoring this spatial dependency between neighboring counties will lead to inecient estimators and even biased regression coecients.11 Yinger et al. (1988) note two major criticisms of studies that estimate property tax capitalization: (1) the lack of controls for the public services nanced by the property tax; and (2) failure to adequately control for dierences in neighborhood characteristics associated with dierent property taxing jurisdictions. The rst concern is partially mitigated by focusing on newly constructed homes and property taxation associated with nancing development infrastructure. A series of control variables for the existence of other municipal or non-development special districts will further alleviate this concern. To address the second criticism, we control for variation in neighborhood amenities that inuence residential property values by including xed eects for CBGs. To further control for unobserved neighborhood characteristics that may be correlated with property tax rates, we spatially subsample the full population of single-family sales based on CBGs that experience high rates of new development. We partition the data into three increasingly restrictive subsamples: the rst consisting of all 2002-2004 sales of recently constructed homes built on lots between 0.05 and 5 acres with transaction prices between $10, 000 and $1, 000, 000; a second sample limited to recently built properties located in CBGs that had at least twenty sales over the 2002-2004 period and with those sales representing at least 20% of the CBG housing stock (as reported by the 2000 Census); and a third sample that further restricts the data to only those transactions in CBGs that included at least 20 sales, with these sales representing at least 20% of the year 2000 CBG housing stock located both within and outside development districts.12 The rst sample contains 34,048 transactions with 84 development districts and 654 CBGs. The second
See Klotz (2004) for a discussion of the implications of spatial dependence in regression models. Conceptually, this third subsample involves matching neighboring developments that are similar in size, but vary in their usage of a development district.
12 11

15

sample contains 27,869 transactions with 69 development districts and 81 CBGs. The third sample contains 16,543 transactions with the 45 development districts listed in Table 1 and located in 23 CBGs. Table 3 provides summary statistics for the three samples. The top panel summarizes the data for all 2002-2004 transactions; the middle panel summarizes the data for the CBGs with at least 20 sales and with sales representing at least 20% of the year 2000 neighborhood housing stock; the bottom panel summarizes the data for just those CBGs that have at least 20 sales with sales representing at least 20% of the housing stock and with transactions located both within and outside development districts. For the entire sample of 34,048 sales, there were an average of 52.1 sales per CBG. The sales rate increased to 344.1 sales per CBG for the second sample and to 728.2 sales per CBG for the third. Population densities decreased signicantly as the data were conned to CBGs containing development districts as these places were in the process of building out. The population densities were 3,294 people per square mile for the 654 CBGs; 1,109 for the 81 CBGs and 552 for the 23 CBGs. Spatially disaggregated subsamples better controlled for dierences in the number of homes sold within versus outside a development district for a given CBG. A test for dierences in means between the number of homes sold inside and outside development districts was signicant at the 5% level in only the full sample of CBGs.13 This indicates that these subsamples can adequately control for the size of new developments located both inside and outside development districts within the same CBG. Table 4 provides the denitions of variables employed in various hedonic specications. This table denes four categories of variables: (1) house price; (2) property characteristics including lot size, square feet of living area, number of bathrooms, dummy variables for garage, forced air heat and replaces; (3) dummy variables for sale quarter; and (4) special district variables including development district property tax rate, the number of local governments/special districts serving the property, the total annual property tax liability and the total annual taxes paid to the development district.
13 The top sample generated a p-value of 0.022; the middle sample p-value is 0.266 and the bottom sample p-value is 0.382

16

Table 3: Census Block Group Summary Statistics All Census 2000 Block Groups Mean Number Sold 02-04 Sold in Development District Sold outside a Development District Total Existing SF Homes Population Population Density (per square mile) N=654 52.1 20.8 31.3 448.3 1,478 3,294 Std Dev. 181.4 107.0 109.1 267.0 842 2,940 Min 1 0 0 0 0 0 Max 2,711 1,177 1,715 2,140 6,775 14,183

CBGs with > 20 sales; sales > 20% of housing stock Mean Number Sold 02-04 Sold in Development District Sold outside a Development District Total Existing SF Homes Population Population Density (per square mile) N=81 344.1 151.6 192.5 362.8 1,159 1,109 Std Dev. 408.7 263.5 255.4 306.3 954 1,423 Min 37 0 0 0 0 0 Max 2,711 1,177 1,715 1,912 6,442 6,545

Sales located both in and out of development districts Mean Number Sold 02-04 Sold in Development District Sold outside a Development District Total Existing SF Homes Population Population Density (per square mile) N=23 722.2 403.9 318.2 550.5 1,677 552 Std Dev. 564.2 361.7 366.8 412.6 1,335 598 Min 88 20 34 47 125 5 Max 2,711 1,177 1,715 1,912 6,442 1,635

17

Table 4: Variable Denitions

18

variable Sales Price ln(Sales Price) PriceSqft Acres Bath SqFt SqF t2 (000s) Garage Basement ForcedAirHeat Firepl Sale Year-Quarter DevDistMillLevy NumberJuris Anntax DevDistAnnTax

Description Transaction price in 2002-2004 The natural log of transaction price in 2002-2004 The per square foot transaction price in 2002-2004 Lot size in acres Number of bathrooms Total square footage of living area Total square footage of living area squared (thousands) Indicator variable for a property with a garage Indicator variable for a property with a basement Indicator variable for a property with forced air heating Indicator variable for a property with a replace Indicator variable for each year and quarter of sale, total = 12 variables Development district mill levy. Equal to zero for properties not in development districts The number of local governments serving a property a Annual property taxes Annual property taxes paid to the development district

Local governments serving a property always include the county government and the school district and may include a municipality and other non-development special/municipal districts (e.g. recreation, security, etc.

Tables 5 and 6 provide summary statistics for all transactions separately located within development districts and for all sales located outside development districts. There are 13,582 transactions of recently built single-family homes located within development districts and 20,466 new sales located outside development districts. The mean transaction price for properties located inside development districts is slightly lower than the mean transaction price for properties located outside development districts ($318, 622 vs. $322, 347) but homes built inside development districts are slightly larger and the mean per square foot transaction price is lower for development district homes ($135 per square foot vs. $144 per square foot). Development district homes are located on slightly smaller lots (0.2 acres vs. 0.27 acres) and more of the development district homes come with garages, forced air heat and replaces, but there is likely to be signicant spatial variation in these attributes that is not taken into account in these summary statistics. Properties located within development districts also have more local governments providing services and pay higher property taxes on average. The average tax rate across transactions is 36.7 mills.

19

Table 5: Summary Statistics - Single-Family homes sold between 2002 and 2004, 3 yrs old or newer and built in a development district variable Sales Price PriceSqft Acres Bath SqFt Garage Basement ForcedAirHeat Firepl Sale02qt1 Sale02qt2 Sale02qt3 Sale02qt4 Sale03qt1 Sale03qt2 Sale03qt3 Sale03qt4 Sale04qt1 Sale04qt2 Sale04qt3 Sale04qt4 DevDistMilllevy NumberJuris Anntax DevDistAnnTax Observations mean 318,622 135.46 0.196 2.877 2,362 0.993 0.826 0.998 0.808 0.069 0.080 0.088 0.093 0.071 0.087 0.092 0.090 0.064 0.079 0.094 0.094 36.7 4.40 3,056 1,056 13,582 sd 133,893 30.13 0.165 0.681 727 0.084 0.379 0.049 0.394 0.253 0.272 0.282 0.291 0.257 0.282 0.289 0.286 0.245 0.269 0.291 0.292 15.6 0.99 1,186 535 min 104,222 47.68 0.05 1 796 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 4 3 827 101 max 997,000 401.45 2.37 8 5,424 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 90 7 10,359 4,450

Results

As noted by Yinger et al. (1988), one of the inherent econometric diculties in estimating tax capitalization is the simultaneity between house prices and tax rates. This relationship can occur because higher property values provide more property tax revenue for a given tax rate. Since infrastructure such as water, sewer and roads have similar xed costs for both higher and lower valued properties, lower tax rates will be needed to cover the costs of infrastructure in neighborhoods

20

Table 6: Summary Statistics - Single-Family homes sold between 2002 and 2004, 3 yrs old or newer, and NOT in a development district variable Sale Price PriceSqft Acres Bath SqFt Garage Basement ForcedAirHeat Firepl Sale02qt1 Sale02qt2 Sale02qt3 Sale02qt4 Sale03qt1 Sale03qt2 Sale03qt3 Sale03qt4 Sale04qt1 Sale04qt2 Sale04qt3 Sale04qt4 DevDistMillLevy NumberJuris Anntax DevDistAnnTax Observations mean 322,347 143.54 0.267 2.730 2,249 0.958 0.881 0.894 0.621 0.072 0.076 0.077 0.079 0.069 0.078 0.088 0.092 0.077 0.097 0.099 0.097 0 3.88 2,657 0 20,466 sd 134,436 35.05 0.339 0.650 680 0.200 0.323 0.308 0.485 0.258 0.265 0.268 0.270 0.253 0.269 0.283 0.289 0.267 0.296 0.298 0.296 0 0.96 1,143 0 min 17,500 9.67 0.05 0.5 810 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 2 224 0 max 999,681 439.75 5 7.5 5,496 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0 8 11,034 0

21

with higher property values. In addition, development district tax rates are inuenced by the size of a subdivision as well as the timing or total amount of buildout within the subdivision. Since these characteristics both inuence tax rates and can be inuenced by the price of homes within the development district, this may introduce additional bias. To account for this simultaneity, we estimate property tax capitalization using an instrumental variable for the tax rate.14 Variables incorporated into instrumenting tax rates (mill levies) will capture the size of a development as well as expected buildout. A rst stage regression equation relates mill levies to development district variables for the number of years since the development district was created, the land area of the development district (in square miles), and the number of homes sold within the development district between 2002 and 2004. The predicted value of mill levies (DevDistM illLevy) is incorporated into the hedonic price regressions in Table 7.15 We estimate the inuence that the development district property tax liability has on property values for three dierent subsamples, with each imposing more restrictive controls for neighborhood characteristics. The variables included in the hedonic house price equation include a standard bundle of housing characteristics; dummy variables for transaction sale quarter; a set of dummy variables that indicate whether a property is served by a re, recreation or other non-development based metropolitan special district or municipality; and, for three of the four specications, dummy variables for CBG. In order to control for the noncoterminous boundaries of school districts and CBGs, regression specications include school district dummies.16 Table 7 provides regression results. The rst column in Table 7 provides results for all 34,048 sales without any controls for CBG xed eects but including variables that control for distance to downtown Denver (in miles) and the distance to downtown squared. This model indicates a relatively large negative tax coecient of -0.001651. Evaluated at the mean transaction price for development district homes ($318,622) and the mean development district tax rate (36.7 mills), the average single-family property tax capitalization is 6.06% or $19,306. Residential properties in
14 Palmon and Smith (1998) provide a similar correction for tax rates using exogenous characteristics of Texas MUDs. 15 A Hausman-Wu-Durbin test conrms the endogeneity of the mill levy variable. See Staiger and Stock (1997) for a description of the Hausman-Wu-Durbin test. 16 A total of 28 school districts in Model 2, 11 school districts in Model 3 and 6 school districts in Model 4 contain a boundary that falls within a CBG.

22

Colorado are assessed at 7.96% of their market value, so the mean assessed value for development district properties is $25,362.31. With an average development district property tax rate of 36.7 mills, the average annual development district property tax liability is $930.80. So how much of this annual development district property tax liability is capitalized in house prices? The amount of capitalization depends on two things: (1) the discount rate that homeowners use to convert the future tax liability to a present value; and (2) the remaining term of the tax liability. We assume, for simplicity, that homeowners can take full advantage of the property tax deduction and that homeowners will discount the future property tax liabilities at the after tax mortgage interest rate.17 The average nominal mortgage interest rate for thirty-year xed rate mortgages over the 2002-2004 period was 6.07%. So households in the 28% tax bracket discount future expected development district property taxes at 4.37%. If we further assume that the annual property tax is an annuity (e.g. that mill rates will decrease if property values increase to provide a constant coupon) and discount the annual annuity of $930.80 at 4.37% over 19 years (the average remaining term for the district bonds listed in Table 2), homeowners value the future development tax liability at $11,850. Under these assumptions, Model 1 estimates that 163% of the development district tax liability is capitalized in the price of new single-family homes. This large estimate could be attributed to unobserved neighborhood heterogeneity, where in-ll new homes in bedroom communities such as Boulder, CO are treated comparably to suburban subdivisions. The higher land rents for in-ll development likely biases the tax coecient downward because development districts are commonly used for suburban subdivisions. All of the other estimated coecients in Model 1 are statistically signicant at conventional levels. The second model estimates parameters using the same 34,048 transactions but incorporates better controls for variation in neighborhood amenities by including xed eects for CBGs.18 The estimated coecient for the development district tax rate decreases to -0.000373. With this estimate, the house price discount is reduced to 1.37%, or $4,362, and, using the same set of assumptions used to estimate property tax capitalization for Model 1, the rate of property tax capitalization
See de Bartolome and Rosenthal (1999) for a good discussion of property tax capitalization under dierent assumptions regarding tax and mortgage interest deductions. 18 The distance to downtown and distance to downtown squared variables are no longer necessary given the CBG xed eects.
17

23

Table 7: Hedonic Estimation Results

VARIABLES DevDistM illLevy acres bath sqft sqf t2 (000s) Garage Basement ForcedAirHeat repl Sale02qt2 Sale02qt3 Sale02qt4 Sale03qt1 Sale03qt2 Sale03qt3 Sale03qt4 Sale04qt1 Sale04qt2 Sale04qt3 Sale04qt4 Constant (1) ln(Sales Price) -0.001651*** (0.000108) 0.197906*** (0.011964) 0.036416*** (0.002547) 0.000332*** (0.000008) -0.000007*** (0.000002) -0.021559** (0.008926) 0.148061*** (0.003144) -0.132809*** (0.016900) 0.065762*** (0.002066) 0.011279** (0.004941) 0.015871*** (0.004629) 0.013486*** (0.004516) 0.021278*** (0.004837) 0.033215*** (0.004799) 0.039903*** (0.004583) 0.047148*** (0.004620) 0.041334*** (0.005216) 0.073935*** (0.004750) 0.084557*** (0.004684) 0.101434*** (0.004800) 11.871966*** (0.027067) (2) ln(Sales Price) -0.000373*** (0.000081) 0.198597*** (0.011775) 0.016829*** (0.001899) 0.000284*** (0.000007) -0.000005*** (0.000001) 0.039571*** (0.008311) 0.108965*** (0.002581) -0.032918 (0.023042) 0.034153*** (0.001597) 0.007485** (0.003441) 0.014240*** (0.003226) 0.015169*** (0.003208) 0.017513*** (0.003428) 0.024593*** (0.003467) 0.032951*** (0.003290) 0.035105*** (0.003312) 0.033909*** (0.003797) 0.060766*** (0.003372) 0.070398*** (0.003426) 0.084156*** (0.003563) 11.933098*** (0.623246) (3) ln(Sales Price) -0.000401*** (0.000075) 0.262966*** (0.006151) 0.015211*** (0.001928) 0.000270*** (0.000007) -0.000003** (0.000001) 0.023165*** (0.005958) 0.105255*** (0.002130) -0.019219 (0.025215) 0.033824*** (0.001594) 0.011857*** (0.003479) 0.013447*** (0.003273) 0.011881*** (0.003231) 0.012359*** (0.003512) 0.022144*** (0.003426) 0.028458*** (0.003292) 0.031915*** (0.003340) 0.026993*** (0.003726) 0.054631*** (0.003338) 0.063481*** (0.003357) 0.078876*** (0.003560) 11.733458*** (0.052751) (4) ln(Sales Price) -0.000515*** (0.000073) 0.309493*** (0.007869) 0.015590*** (0.002307) 0.000259*** (0.000008) -0.000003 (0.000002) -0.002943 (0.011949) 0.103633*** (0.002497) -0.044478 (0.031517) 0.029968*** (0.002059) 0.004521 (0.004769) 0.003817 (0.004392) 0.003873 (0.004281) 0.007934 (0.004616) 0.020394*** (0.004704) 0.020822*** (0.004409) 0.017790*** (0.004389) 0.023994*** (0.004768) 0.042615*** (0.004384) 0.045318*** (0.004314) 0.061764*** (0.004477) 11.856366*** (0.038062)

Controls for Other Governments Yes Yes Yes Yes Fixed Eects - Census BG No Yes Yes Yes Observations 34,048 34,048 27,869 16,543 241,534 F-Stat 1,951 2,105 1,583 R2 0.7635 0.8834 0.8803 0.8694 Durbin-Wu-Hausman Test 278.4*** 15.8*** 22.86*** 30.14*** Instruments F-Stat 7,555 2,683 2,564 2,570 *** p<0.01, ** p<0.05, * p<0.10. Model 1 coecients for Distance to CBD and Distance to CBD squared are suppressed but are signicant and signed negatively and positively respectively. Robust standard errors in parentheses.

declines to 36.8%! The estimated coecients for lot size (in acres) are virtually identical in Models 1 and 2, but there are signicant dierences in many of the other estimated coecients. In Model 2, estimated coecients for all variables, except the dummy variable for ForcedAirHeat, are statistically signicant at conventional levels. Including CBG xed eects signicantly improves the models goodness of t with R-squares increasing from 76% of the variance in the natural log of transaction price explained to 88%. Model 3 restricts the data to transactions in CBGs with at least 20 sales and with those sales representing at least 20% of the year 2000 neighborhood housing stock. This subsample increases the magnitude of the the estimated coecients for both the development district tax rate and for lot size. The estimate of the house price discount generated by Model 3 is $4,689 and the estimated development district property tax capitalization is 39.6%. All the estimated coecients for the other variables are statistically signicant at conventional levels except for ForcedAirHeat. Model 4 further limits the sample to those homes built and sold in CBGs with sales located both within and outside development districts. Relative to Models 2 and 3, this Model increases both the estimated property tax capitalization eect and the marginal value of land. The estimated house price discount is 1.89%, or $6,022, and the estimated development district property tax capitalization is 50.8%. The estimated coecients for square feet of living space squared and attached garage become statistically insignicant. Unreported coecients for government variables were statistically insignicant except that re special districts and non-development metropolitan special districts negatively impacted housing prices. In addition, while the rst three models indicate that house prices were increasing throughout the 2002-2004 period, the fourth specication indicates that house prices were basically at for the rst ve quarters of the 2002-2004 period suggesting developers were unable to increase house prices as the rapidly growing developments were being built out. If homeowners were unable to take advantage of the property tax deduction and discounted future development district tax liabilities at 6.07% instead of 4.37%, then present values of future tax liabilities decrease and estimates of property tax capitalization increase to 187% in Model 1, 42% in Model 2, 45% in Model 3 and 58% in Model 4.

25

Two test statistics are reported in Table 7, the Durbin-Wu-Hausman test for the exogeneity of DevDistMillLevy and an F-statistic for the joint signicance of the instruments used to create DevDistM illLevy. The Durbin-Wu-Hausman test conrms the endogeneity of DevDistMillLevy and the need for DevDistM illLevy. The F-statistic for the instruments is greater than 2,500 in all regressions and signicant at the 1% level. As noted by Staiger and Stock (1997), a good rule of thumb for viable instruments is an F-statistic above 10. F-statistics reported in Table 7 alleviate any concern about weak instruments. Table 7 estimated coecients on the development district tax rate variable for Models 2 through 4 range from -0.000373 to -0.000515. The estimated coecients for lot size also increase signicantly when the sample was restricted to properties in CBGs containing sales both within and outside development districts. The estimated coecients for the other structural characteristics were similar across Models 2, 3 and 4. Given the similarity of a number of coecients in Models 2, 3 and 4, the sampling of subsequently smaller numbers of CBGs may represent the same underlying data generating process. The three subsamples given in Models 2 through 4 represent a full model and two spatially restricted models. The restricted models implement an assumption of a structural break based on high growth versus low growth CBGs. In order to test this assumption, a Wald test is used to compare all property variables and the tax variable in Model 4 to the sample of observations in Model 2 and Model 3 that were excluded from Model 4. The Wald test yields a 2 = 208.3 (p = 0.000) and rejects the null hypothesis that Model 4 is structurally equivalent to Model 2 at a 1% signicance level. The Wald test between Models 3 and 4 yields a 2 = 206.4 (p = 0.000) and rejects the null hypothesis that Model 4 is structurally equivalent to Model 3 at a 1% signicance level. As a sensitivity test, we estimated the parameters of Model 4 under a variety of assumptions for the minimum number of transactions as well as for the percent of existing homes used to dene the high growth CBG models. Table 8 shows the robustness of results to variation in the CBGs and development districts adopted to estimate the tax coecient. Estimated tax coecients vary by less than 5% across these four dierent assumptions for determining Model 4. Table 9 provides estimated capitalization eects for the 33 Denver area development districts

26

Table 8: Sensitivity Test Min Transactions/ % of Existing Homes 10/10% 20/20% 30/20% 40/40% No. Census BGs 26 23 20 15 No. Development Districts 55 50 47 43 No. Observations 16,767 16,543 16,086 14,928 Estimated Tax Coecient -0.000517 -0.000515 -0.000547 -0.000547

highlighted in Table 2.19 For each development district, the table lists the average transaction price, the development district tax rate, the estimated house price discount obtained using the parameter estimate from Model 4 (our preferred specication), the annual property tax, the term remaining on the development district bonds, the present value of the future development district tax liability (assuming households are in the 28% marginal tax bracket and can take full advantage of the property tax deduction), and the percent property tax capitalization. The presence of greater than or almost 100% capitalization for some development districts is due to the use of the mean estimated tax coecient for development districts with few years remaining on the terms of the bond as of 2004. Another way to interpret this result is to compute the length of time necessary to justify the estimated capitalization given the annual tax liability and the homeowners discount rate.20 At a discount rate of 4.37%, 7.8 years of annual property tax payments are capitalized in house prices; at 6.07% (assuming the household either does not itemize or cannot use the property tax deduction), 9 years of property taxes are capitalized.

Concluding Remarks

This paper empirically examines the extent to which property taxes associated with development district nancing of residential infrastructure is capitalized in house prices. We estimate the paThe remaining 17 development districts used in estimation contained incomplete bond data to compute capitalization rate for individual development districts. 20 One can calculate the number of years (N) required to equate the present value of future tax liability to equal AR the property value capitalization based on N = ln( (AR+1000d) )/[ln(1 + d)], where AR = assessment rate, which is 0.0796 in Colorado, is the estimated tax coecient, and d is the homeowners discount rate.
19

27

Table 9: Estimated Capitalization Eects

28

Development Special District Name Aurora Single Tree Metropolitan District Belle Creek Metropolitan District Brighton Crossing Metropolitan District No. 4 Bualo Ridge Metropolitan District Canterberry Crossing Metropolitan District Canterberry Crossing Metropolitan District II Cherry Creek South Metropolitan District No. 1 Eagle Bend Metropolitan District No. 2 Eagle Creek Metropolitan District East Smoky Hill Metropolitan District No. 2 Ebert Metropolitan District Gateway Regional Metropolitan District Goodman Metropolitan District GVR Metropolitan District Highlands Ranch Metropolitan District No. 1 Highlands Ranch Metropolitan District No. 3 Highlands Ranch Metropolitan District No. 4 Interlocken Consolidated Metropolitan District Lincoln Park Metropolitan District Maher Ranch Metropolitan District No. 4 Meadows Metropolitan District No. 1 Meadows Metropolitan District No. 2 North Range Village Metropolitan District Pinery West Metropolitan District No. 2 Potomac Farms Metropolitan District Riverdale Dunes Metropolitan District No. 1 Roxborough Village Metropolitan District Sterling Hills Metropolitan District Stonegate Village Metropolitan District Superior Metropolitan District No. 2 Todd Creek Farms Metropolitan District No. 1 Todd Creek Farms Metropolitan District No. 2 Upper Cherry Creek Metropolitan District

Average Sales Price $229,180 $194,474 $225,072 $205,856 $232,914 $255,525 $256,761 $415,308 $266,531 $256,817 $252,733 $248,462 $245,067 $255,983 $369,769 $367,252 $306,667 $399,530 $430,028 $227,679 $302,879 $335,700 $237,077 $366,465 $288,830 $240,710 $317,903 $271,744 $387,839 $395,456 $311,550 $271,846 $256,119

Property Mill Levy 42.6 47.9 38.0 37.3 39.5 37.3 39.8 45.3 50.0 17.0 39.0 10.0 26.5 30.3 20.3 20.3 20.3 27.2 45.0 40.0 35.0 35.0 38.5 45.0 38.0 45.0 70.1 30.0 27.4 22.0 16.5 18.0 6.0

Estimated Price Discount $5,028 $4,797 $4,405 $3,954 $4,738 $4,909 $5,263 $9,689 $6,863 $2,248 $5,076 $1,280 $3,345 $3,994 $3,866 $3,839 $3,206 $5,597 $9,966 $4,690 $5,459 $6,051 $4,701 $8,493 $5,652 $5,578 $11,477 $4,198 $5,473 $4,481 $2,647 $2,520 $791

Annual Property Tax $776.76 $741.50 $680.80 $610.50 $732.14 $757.80 $812.93 $1,497.19 $1,060.79 $347.52 $784.59 $197.78 $517.72 $616.81 $596.97 $592.91 $495.10 $866.30 $1,540.36 $724.93 $843.60 $935.26 $726.55 $1,312.68 $873.65 $862.22 $1,772.95 $648.92 $846.79 $692.52 $409.19 $389.50 $122.32

Remaining Term as of 2004 21 16 30 19 27 28 29 15 17 26 21 6 15 15 8 15 13 15 22 24 25 25 16 28 28 28 17 14 22 9 17 14 7

Present Value of Future Tax Liability $10,535 $8,409 $11,261 $7,772 $11,475 $12,106 $13,221 $16,224 $12,543 $5,337 $10,641 $1,024 $5,610 $6,684 $3,959 $6,425 $4,832 $9,387 $21,493 $10,646 $12,678 $14,056 $8,240 $20,969 $13,956 $13,774 $20,963 $6,690 $11,815 $5,063 $4,838 $4,016 $724

Percent Capitalization 47.7% 57.1% 39.1% 50.9% 41.3% 40.5% 39.8% 59.7% 54.7% 42.1% 47.7% 124.9% 59.6% 59.8% 97.7% 59.8% 66.3% 59.6% 46.4% 44.1% 43.1% 43.1% 57.1% 40.5% 40.5% 40.5% 54.7% 62.8% 46.3% 88.5% 54.7% 62.8% 109.3%

rameters of a hedonic house price model and control for neighborhood attributes by restricting the sample to single-family homes built in CBGs containing numerous sales of properties located both within and outside development districts. We nd that about half of the future property tax liability is capitalized in house prices. Alternatively, homebuyers capitalize between 8 and 9 years of property tax liability in the purchase price. We also nd that our estimates are very sensitive to the sample used, with the most accurate capitalization rates estimated from the sample that provides the best empirical controls for neighborhood attributes. The hedonic house price specication that fails to provide adequate controls for neighborhood amenities (Model 1) yields an estimated 163% property tax capitalization. Limiting the data to rapidly growing residential developments and controlling for variation in neighborhood amenities using CBG xed eects, the estimate of development district property tax capitalization is about 50%.

29

References
Billings, S. and Thibodeau, T. G.: 2009, Intrametropolitan decentralization: Is government structure capitalized in residential property values?, University of Colorado mimeo . Brueckner, J. K.: 1979, Property values, local public expenditure and economic eciency, Journal of Public Economics 11(2), 223245. Brueckner, J. K.: 1997, Infrastructure nancing and urban development: The economics of impact fees, Journal of Public Economics 66, 383407. de Bartolome, C. and Rosenthal, S. S.: 1999, Property tax capatilization in a model with deferred assets, standard deductions, and the taxation of nominal interest, Review of Economics and Statistics 81(1). Delaney, C. J. and Smith, M. T.: 1989, Impact fees and the price of new housing: An emprical study, American Real Estate and Urban Economics Journal 17(1), 4154. Evans-Cowley, J. S. and Lawhon, L. L.: 2003, The eects of impact fees on the price of housing and land: a literature review, Journal of Planning Literature 17(3), 351359. Galvan, S. C.: 2007, Wrestling with muds to pin down the truth about special districts, Fordham Law Review 75, 30413080. Grith, J. C.: 2007, Recent developments in public nance law: Special tax districts to nance residential infrastructure, The Urban Lawyer 39(4), 959983. Ihlanfeldt, K. R. and Shaughnessy, T. M.: 2004, An empirical investigation of the eects of impact fees on housing and land markets, Regional Science and Urban Economics 34, 639661. Klotz, S.: 2004, Cross Sectional Dependence in Spatial Econometric Models with an Application to German Start-up Activity Data, LIT Verlag Berlin-Hamburg-Mnster. Oates, W. E.: 1969, The eects of property taxes and local public spending on property values: An empirical study of tax capitalization and the Tiebout hypothesis, Journal of Political Economy 77(6), 957971. 30

Palmon, O. and Smith, B. A.: 1998, New evidence on property tax capitalization, Journal of Political Economy 106(5), 10991111. Peiser, R. B.: 1983, The economics of municipal utility districts for land development, Land Economics 59(1), 4357. Singell, L. D. and Lillydahl, J. H.: 1990, Housing impact fees, Land Economics 66(1), 8292. Staiger, D. and Stock, J.: 1997, Instrumental variable regressions with weak instruments, Econometrica 65, 55786. Starrett, D. A.: 1981, Land value capitalization in local public nance, Journal of Political Economy 89, 30627. Thibodeau, T. G.: 2003, Marking single-family property values to market, Real Estate Economics 31(1), 122. Yinger, J.: 1982, Capitalization and the theory of local public nance, Journal of Political Economy 90, 91743. Yinger, J.: 1998, The incidence of development fees and special assessments, National Tax Journal 51(1), 2341. Yinger, J., Bloom, H. S., BorschSupan, A. and Ladd, H. F.: 1988, Property Taxes and House Values: The Theory and Estimation of Intrajurisdictional Property Tax Capitalization, Academic Press. Zodrow, G.: 2006, Who pays the property tax?, Lincoln Institute of Land Policy Land Lines 18(2).

31