Document 219-2
Filed 10/05/2006
Page 1 of 45
EFFECTS OF AIRCRAFT NOISE ON RESIDENTIAL PROPERTY VALUES:
NAVAL AIR STATION (NAS) OCEANA and NAVAL AUXILIARY LANDING FIELD (NALF) FENTRESS
______________________________________________________________________________ EXPERT REPORT OF JON P. NELSON, Ph.D.
September 26, 2005 ______________________________________________________________________________
Case 1:01-cv-00201-VJW
Document 219-2
Filed 10/05/2006
Page 2 of 45
I. INTRODUCTION AND QUALIFICATIONS 1. My name is Jon P. Nelson. I reside at 642 Glenn Road, State College, PA 16803. I was employed by the Pennsylvania State University from August 1969 to June 30, 2004. I was promoted to the academic rank of Professor of Economics in 1978 and retired in June 2004 with the rank of Professor Emeritus of Economics. I am presently engaged in private consulting and economic research. My research expertise is the areas of organization of industry, public policy toward business, economics of regulation, applied microeconomic analysis, and environmental economics. During my academic career, I regularly taught courses on industrial organization, antitrust and regulatory economics, microeconomic theory, and environmental economics. From 1986 to 1993, I was director of graduate studies in my department. My curriculum vita is attached to this report. 2. I received my Ph.D. degree in Economics in 1970 from the University of Wisconsin. My major fields of concentration during graduate study were industrial organization and econometrics. Industrial organization is the sub-field of economics that studies markets and industries, and evaluates the extent to which they are functioning in an economically efficient manner. Economics of regulated industries, which includes air transportation, is one of the topical areas of industrial organization. Econometrics is the sub-field of economics that studies the application of statistical methods to economic data and mathematical models. 3. During the past 35 years, the focus of my research has been industry and market regulation, including environmental regulation and policy. I have written three books and more than 70 articles on these topics, including articles on the effects of aircraft noise on residential property values. My peerreviewed articles have appeared regularly in scholarly journals such as the Journal of Econometrics, Review of Economics and Statistics, Journal of Law and Economics, Journal of Regulation, Journal of Regional Science, Quarterly Review of Economics and Business, Southern Economic Journal, Empirical Economics, Contemporary Economic Policy, Applied Economics, Antitrust Bulletin, Journal of Environment and Development, and other peer-reviewed journals and collected volumes. 4. My research expertise includes the evaluation of transportation noise and its control. I have published one book and nine articles and chapters on this topic. My book is entitled Economic Analysis of Transportation Noise Abatement (Ballinger 1978). My articles and chapters have appeared in Noise Abatement: Policy Alternatives for Transportation (National Academy of Sciences 1977); Journal of Environmental Economics and Management (1979, 1985); Journal of Transport Economics and Policy (1980, 1982, 2004); Advances in Applied Microeconomics (JAI Press 1981); Resource Management and Environmental Planning (Wiley 1983); and Interagency Symposium on Research in Transportation Noise Proceedings (Stanford 1973). As a mark of importance, these writings have been cited more than
Case 1:01-cv-00201-VJW
Document 219-2
Filed 10/05/2006
Page 3 of 45
2
300 times by other researchers in economics, transportation, and real estate. Two of my articles have been reprinted in collected volumes, which is another mark of their importance. The 1979 article was reprinted in The Environment and Transport (Hayashi et al. 1999) and the 1980 article was reprinted in Classics in Transport Analysis: Air Transport (Forsyth et al. 2002). In January 2004, I published a peerreviewed meta-analysis of airport noise and property values, which was the lead article in the Journal of Transport Economics and Policy (Nelson 2004; attachment 2). 5. During 1976-77, I was a member of the Committee on Appraisal of Societal Consequences of Transportation Noise Abatement, National Academy of Sciences, and I contributed to two chapters in the Committee's final report (DeVany et al. 1977a, 1977b). I received three research grants from the U.S. Department of Transportation to investigate the effects of mobile-source pollution on residential property values, including aircraft noise (1972-73, 1974-75, and 1977-78). I have published several other articles that examine adverse effects of pollution on residential property values, including peer-reviewed articles appearing in the Southern Economic Journal (1977); Journal of Urban Economics (1978); and Land Economics (1981). I have acted as a technical consultant to several organizations concerned with noise and other pollutants, including the Pennsylvania Low-Emissions Vehicle Commission (1993). 6. I have experience with the collection of data on residential property values and the measurement of aircraft noise levels for economic studies; I have performed several econometric analyses of the empirical relationship between aircraft noise and residential property values; I have examined the costs and benefits of aircraft noise abatement policies; and I have extensively reviewed the academic literatures in economics and real estate that deal with the effects of noise on residential property values in the United States, Canada, and other countries (Nelson 1978, 1980, 1982, 2004). 7. This report contains my evaluation of the effects on residential property values resulting from aircraft noise exposure, including the impact of the realignment of F/A-18 C/D fleet squadrons and fleet replacement squadrons (FRS) to Naval Air Station (NAS) Oceana, Virginia, and Naval Auxiliary Landing Field (NALF) Fentress, Virginia. This decision was effective during 1998-1999. The Navy's realignment decision was followed by the publication of additional environmental policies, including a revised AICUZ map for Oceana and Fentress (1999); a draft EIS for the basing of F/A-18 E/F (Super Hornet) aircraft at these fields (July 2002); revised land use compatibility guidelines for Navy installations (December 2002); and a Joint Land Use Study for the Hampton Roads area (Hampton Roads PDC, April 2005). In September 2003, the Navy announced its intentions to homebase eight F-18 E/F fleet squadrons (96 aircraft) and the FRS (24 aircraft) at NAS Oceana. 8. My rate of compensation as an expert is $200 per hour, and my compensation is not dependent on my method of evaluation or the conclusions researched from my evaluation. My opinions
Case 1:01-cv-00201-VJW
Document 219-2
Filed 10/05/2006
Page 4 of 45
3
and conclusions are solely my own. During 1998-2000, I was retained by the plaintiffs in three cases involving restrictions on commercial speech. I have not testified in other cases in the past five years. I have been swore in and accepted as an expert witness by federal and state courts and state regulatory agencies. 9. For this report, I reviewed academic and government studies that analyze aircraft noise control, including many studies of the effects of aircraft noise on residential property values. I also collected basic data on the economies of Virginia Beach and Chesapeake, Virginia. As part of my evaluation, I was provided with copies of (1) Wyle Laboratories, Aircraft Noise Study for the 1995 BRAC Realignment of Navy F/A-18 Aircraft, WR 97-10 (September 1997, February 1998); (2) Wyle Laboratories, Aircraft Noise Study for Naval Air Station Oceana and Auxiliary Landing Field Fentress, WR 94-18 (July 1994); (3) U.S. Department of the Navy, Final Environmental Impact Statement: Realignment of F/A-18 Aircraft and Operational Functions from Naval Air Station Cecil Field, Florida, to Other East Coast Installations (March 1998a); (4) U.S. Department of the Navy, Record of Decision and General Conformity Determination for Realignment of F/A-18 Aircraft and Operational Functions from Naval Air Station (NAS) Cecil Field, Florida, to other East Coast Installations (May 18, 1998b); (5) U.S. Department of the Navy, AICUZ Program Procedures and Guidelines for Department of the Navy Air Installations, OPNAVINST 11010.36A (April 1988) and OPNAVINST 11010.36B (December 2002); (6) U.S. Department of the Navy, Final Environmental Impact Statement (FEIS) for the Introduction of F/A-18 E/F (Super Hornet) Aircraft to the East Coast of the United States (July 2003); and (7) Wyle Laboratories, Noise Study for the Introduction of F/A-18 E/F to the East Coast, Vol. 1, WR 02-08 (April 2003). I examined the web site for NAS Oceana,
II. NOISE EXPOSURE DUE TO ARS-2 ARS-2 substantially increased aircraft operations at NAS Oceana and NALF Fentress, including touch-and-go, FCLP, and nighttime operations. 10. The 1993 and 1995 Defense Base Closure and Realignment Commissions (BRAC) directed the closure of NAS Cecil Field, Florida. During 1998-99, the Department of the Navy (DoN) realigned
Case 1:01-cv-00201-VJW
Document 219-2
Filed 10/05/2006
Page 5 of 45
4
its aircraft and personnel to NAS Oceana, Virginia (156 aircraft) and Marine Corps Air Station Beaufort, South Carolina (24 aircraft). This realignment scenario is ARS-2 among the five alternative scenarios considered for transfer of the F/A-18 aircraft. NAS Oceana is located approximately 12 miles southeast of Norfolk, Virginia, in the City of Virginia Beach. NAS Oceana has two sets of crossing parallel runways three measuring 8,000 feet in length and one measuring 12,000 feet. NALF Fentress is about 7 miles southwest of NAS Oceana in the City of Chesapeake, Virginia, and is used primary by aircraft based at NAS Oceana for Field Carrier Landing Practice (FCLP) patterns. NALF Fentress has one 8,000 foot runway equipped to simulate an aircraft carrier flight deck (Wyle, 1998 at 2-1). Aircraft operations at NALF Fentress normally occur at a pattern altitude of 800 feet (Hampton Roads PDC, 2005 at 2-8). Some of the analysis treats the two airfields together because NALF Fentress is located close to NAS Oceana, and it is used primarily as a practice field for aircraft based at NAS Oceana (Wyle, 1998 at 3-1). NAS Oceana occupies 5,331 acres of land, which increases to 6,820 acres with the Dam Neck Annex. Obstruction clearances and flight easements add an additional 3,680 acres (DoN, 2003 at 3-16; http://www.globalsecurity.org). NALF Fentress occupies a total of 2,560 acres. 11. Wyle Labs defined the 1997 baseline scenario for the ARS-2 analysis as a single-siting of eleven Atlantic Fleet F-14 squadrons (150 aircraft), one FRS, and one Navy F/A-18 adversary squadron at NAS Oceana (Wyle, 1998 at 3-1). The aircraft mix for the baseline included 198 aircraft (Wyle, 1998 at 3-4). Baseline and projected annual operations were computed using the Naval Aviation Simulation Model (NASMOD). Calendar year 1997 (CY97) was the time period for the baseline analysis and CY99 was used for the ARS-2 analysis. However, in the Navy's most recent EIS report (DoN, 2003 at 3-9), the baseline was updated to CY00 in anticipation of the arrival of F/A-18 E/F aircraft. Separate operational data were identified for each airfield in CY97 and CY00. While the projected data in the most recent EIS report are tentative, I have compared the CY00 operational data to the projected CY99 data used in the earlier reports by Wyle. The differences are modest. The Navy claims that noise exposed areas and populations changed as a result of operational changes that occurred after the arrival in 1998-99 of the F/A-18 C/D aircraft (DoN, 2003 at 3-9). However, following the analysis in the Hampton Roads Joint Land Use Study (2005), I use the 1999 AICUZ noise zones for areas and populations. 12. Baseline operations in 1997 at NAS Oceana were about 109,000 annual flight operations (Wyle, 1998 at Table 3-1). F-14 fleet and FRS aircraft accounted for 86% of NAS Oceana baseline operations, including 49,000 touch-and-go operations. NALF Fentress had about 105,000 annual flight operations in 1997 (Wyle, 1998 at Table 3-2). F-14 fleet and FRS aircraft accounted for 59% of NALF Fentress operations, including 55,000 FCLP operations. In 1997, about 7% of NAS Oceana operations (7,000) and 33% of NALF Fentress operations (35,000) occurred at night.
Case 1:01-cv-00201-VJW
Document 219-2
Filed 10/05/2006
Page 6 of 45
5
13. ARS-2 substantially increased aircraft operations at NAS Oceana from the CY97 baseline value of 109,000 (Wyle, 1998 at Tables E-1 and 4-16). The total number of operations in CY00 was 207,722, including 25,294 operations (12.2%) during the nighttime hours (DoN, 2003 at 3-4). The realignment of F/A-18 fleet and FRS aircraft accounted for more than 98,000 new operations in CY00, including FCLP and touch-and-go operations. Total nighttime operations at NAS Oceana rose dramatically to 25,000 annually, or about 68 operations per night. 14. ARS-2 substantially increased aircraft operations at NALF Fentress from the CY97 baseline of 105,000 (Wyle, 1998 at Tables E-1 and 4-17). The total number of operations in CY00 was 135,190, including 54,850 operations (40.6%) at night (DoN, 2003 at 3-4). The realignment of F/A-18 fleet and FRS aircraft accounted for more than 30,000 new operations in CY00, including FCLP and touch-and-go operations. Total nighttime operations at NALF Fentress rose dramatically to 55,000 annually, or more than 150 operations per night.1 The increase in airfield operations from ARS-2 dramatically increased the residential populations exposed to noise levels of 65 dB or greater. More than 23,000 people are exposed to noise levels of 80 dB or more, which is a level of exposure with the potential for serious adverse health and welfare consequences. 15. Wyle Labs computed baseline and ARS-2 noise exposures using the NOISEMAP 6.5 computer program (Wyle, 1998 at 4-34). At a minimum, NOISEMAP has the capability to compute noise levels in 5 decibel increments, referred to as "noise zones" or "noise intervals." The noise metric used for assessing noise exposure in the vicinity of airfields and airports is the Day-Night Average Sound Level (DNL or Ldn), expressed in A-weighted decibels (dB). DNL is an average sound level generated by all aircraft and aviation-related operations during an average 24-hour period, with a 10 dB penalty for nighttime noise events.2 Several other noise metrics are computed that capture special features of
A comparison of projected operations for CY99 and updated values for CY00 indicates that the earlier EIS reports (DoN 1998a, b; Wyle 1998) contained a slight overestimate of total operations at Oceana, and an underestimate of nighttime operations. At Oceana, the CY99 estimate was 219,000 operations compared to a CY00 value of 208,000 (Wyle, 1998 at 4-16). CY99 nighttime operations were estimated at 18,000 annually compared to CY00 operations of 25,000. At Fentress, total operations were 144,000 for CY99 and 135,000 for CY00. Due to capacity constraints at Fentress, nighttime operations were 55,000 for both CY99 and CY00. The DNL is the Equivalent Sound Level (Leq), with a 10 dB penalty for nighttime noise. The Leq is the energy-averaged sound level integrated over a specified time period. Leq provides a single number measure of sound averaged over a time period, where "equivalent" means that the acoustical energy of the steady state sound equals the energy associated with an actual varying sound level. DNL was selected by the EPA as the uniform descriptor of cumulative sound energy that best correlates with health and welfare effects of noise, including annoyance, community noise exposure, and land use compatibility (FICON, 1992b at B-16).
2
1
Case 1:01-cv-00201-VJW
Document 219-2
Filed 10/05/2006
Page 7 of 45
6
military aircraft operations, such as sporadic operations, low altitude flights, high speeds, and the "surprise" effect resulting from high-speed low-altitude operations (Wyle, 1998 at 1-3). 16. Typical background noise levels in urban areas are in the interval 50-60 dB during daytime hours and 40 dB during nighttime (Nelson, 1978 at 22). A 50-60 dB sound corresponds to the noise level from light traffic at 100 feet or an air conditioner at 100 feet. A food blender or garbage disposal at three feet produces a sound level of about 80 dB. Because the ear's pattern of response is more nearly logarithmic in nature, decibels are measured on a logarithmic scale. The psychological perception of loudness roughly doubles with each 10 dB increase (Nelson, 1978 at 16). A 90 dB sound is perceived to be twice as loud as an 80 dB sound, four times louder than a 70 dB sound, and eight times louder than a 60 dB sound. At 1,000 feet, F-14 arrivals or departures produce maximum sounds of 83 and 97 dB, respectively. F-18 arrivals or departures produce maximum sounds of 104 dB and 108 dB, respectively, at 1,000 feet (DoN, 1998a at Figure 3.1-21). Hence, F-18 aircraft are four times as loud as F-14 aircraft on arrival and twice as loud on departure. 17. Human annoyance due to noise is generally taken to be a function of both the overall soundintensity level and fluctuations in the level produced by significant intrusive sounds (EPA, 1971 at 113; Nelson, 1978 at 18). A 3 dB change in sound level represents a doubling of the sound energy. Changes in DNL greater than 3 dB are regarded as an indicator of the need for further analysis of noise impacts (FICON, 1992b at 3-15; DoN, 2003 at 4-43). At noise levels above 65 dB, changes as small as 1.5 dB may be significant due to the effects on human annoyance (FICON, 1992a at 1-4). ARS-2 increased the outdoor noise exposure in the vicinity of the two airfields by 8 to 20 dB (Wyle, 1998 at 4-37; DoN, 1998b at 11). Considerable economic evidence exists that demonstrates that this change in exposure produces damages that will be reflected in residential property values. 18. The 1997 baseline noise contours show two distinct regions of noise exposure one for each airfield (Wyle, 1998 at Figure 3-16). At NAS Oceana, the 65 DNL contour extends about 1.5 miles to the east and west of the NAS boundary at the widest points. At NALF Fentress, the 65 DNL contour extends about 1.5 miles to the west and 3 miles to the east of the NALF boundary at the widest points (Wyle, 1998 at 3-2). Under baseline conditions, 14,300 acres of land are in the DNL 65+ zone. As a consequence of ARS-2, the land area enclosed by DNL 65+ increases to more than 58,000 acres (Wyle, 1998 at Table 4-20). 19. There is a dramatic increase in the number of people exposed to unacceptably high levels of aircraft noise in the vicinity of Oceana and Fentress. About 23,000 people are exposed to noise levels of DNL 80 and above (DNL 80+), which is a severe level of noise exposure with the potential for temporary hearing loss and substantial adverse effects on the community environment (FICON, 1992b at 3-8). An
Case 1:01-cv-00201-VJW
Document 219-2
Filed 10/05/2006
Page 8 of 45
7
additional 97,000 people are exposed to noise levels in the 65 to 79 dB zone, which is a substantial level of noise that interferes with speech, concentration, learning, sleep, and tranquility. A noise level of 65 dB has been accepted by the Navy and other organizations as the maximum noise level at which interference with residential activities and land use normally occurs (DoN, 2002 at 8). 20. Table 1 shows the effects of ARS-2 on off-base land areas and populations. Due to ARS-2, the population within DNL 65+ rose from 28,000 in 1997 to more than 120,000 people in 1999. Table 1 also shows two separate populations that will be the subject of the property value analysis below. First, the population within the DNL 65-79 zone, which rose from 28,000 people in 1997 to 97,000 people in 1999. Second, the population within the DNL 80+ zone that rose to 23,000 people in 1999, compared to only four people in 1997.
Table 1. Noise-Exposed Land Acres and Population near NAS Oceana and NALF Fentress under ARS-2 Baseline Acres (CY97) 9,217 4,428 618 35 0 ARS-2 Acres (CY99) 18,933 13,209 10,817 8,901 6,510 Baseline Population (CY97) 19,823 7,837 366 4 0 ARS-2 Population (CY99) 40,877 32,635 23,732 13,607 9,812 Pop. Change, 97-99 21,054 24,798 23,366 13,603 9,812 Pop. Percent Change 106.2 316.4 not calc. not calc. not calc.
Noise Zone (dB) DNL 65-70 DNL 70-75 DNL 75-80 DNL 80-85 DNL 85+ Totals by Zone: Total: DNL 65-79 Total: DNL 80+
14,263 35
42,959 15,411
28,026 4
97,244 23,419
69,218 23,415
247.0 not calc.
14,298 58,370 28,030 120,663 92,633 330.5 Total: DNL 65+ Source: Wyle (1998 at Table 4-20). Acres are off-station and exclude other military and water areas. Population estimates are based on 1990 Census population densities of about 1.5-2.0 people per acre. The recent EIS report used 2000 Census population densities of about 2.0-2.5 people per acres. Much of the available residential land areas around Oceana are developed, but this is less true at Fentress (Hampton Roads PDC, 2005 at 2-9).
Land use compatibility guidelines adopted by several federal agencies, including the FAA and Department of the Navy, indicate that noise levels above DNL 75 are clearly unacceptable for residential purposes, and levels of DNL 65-75 are normally unacceptable. 21. The realignment of aircraft under ARS-2 substantially increased noise exposure levels in the vicinity of NAS Oceana and NALF Fentress. When noise exposure rises above DNL 65, there is a
Case 1:01-cv-00201-VJW
Document 219-2 8
Filed 10/05/2006
Page 9 of 45
noticeable increase in individual annoyance and community complaints.3 With this in mind, guidelines for compatible residential land use have been established by several federal agencies, including the U.S. Navy (1978, 1988, 2002). The U.S. Department of Housing and Urban Development (HUD) has guidelines for safe noise environments within a building. HUD's "Site Acceptability Standards" show that exterior noise in the 65-75 dB zone is normally unacceptable for guaranteed loans, and 75 dB and higher levels are clearly unacceptable (24 Code of Federal Regulations, Part 51, "Environmental Criteria and Standards," at 340). The Federal Aviation Administration (FAA) has regulations that identify land uses that are normally incompatible with various levels of noise exposure. The FAA's "Land Use Compatibility" standard shows that noise levels of 65-75 dB are incompatible with residential land uses (other than mobile homes), unless noise attenuation measures are used (14 Code of Federal Regulations, Part 150, "Airport Noise Compatibility Planning," at 87). Above 75 dB, residential land use is considered unacceptable due to interference with normal activities, even with the application of noise mitigation measures (GAO, 2000a at 23). The U.S. Department of Defense (DoD) and U.S. Department of the Navy (DoN) have a long-standing program for land use compatibility around military airfields, labeled the Air Installation Compatible Use Zones (AICUZ) program (DoN 1978, 1988, 2002). The Navy's guidelines indicate that residential housing is "discouraged" at 65-69 dB, "strongly discouraged" at 70-74 dB, and "incompatible" with 75 dB and above (DoN, 2002 at 20 and Table 2). 22. In June 1980, the U.S. Federal Interagency Committee on Urban Noise (FICUN) published the Guidelines for Considering Noise in Land Use Planning and Control (FICUN 1980). The FICUN guidelines adopted DNL as the standard noise metric. These guidelines indicate that residential land use is discouraged at DNL 65-69; strongly discouraged at DNL 70-74; and incompatible with DNL 75 and above (FICUN, 1980 at 6). The 1990 U.S. Federal Interagency Committee on Noise (FICON), which included the DoN and FAA, recommended continued use of the DNL metric and the compatibility guidelines. As a consequence, land use compatibility criteria used by the HUD, FAA, DoD, and DoN are similar (FICON, 1992b at 3-21; EPA, 1982b at ii). However, outdoor noise exposure due to lifestyle differences is not directly accounted for by these criteria, which is important for the residential areas that surround the two Naval airfields. 23. An American National Standard for land use compatibility has been established by the S12 (Noise) Accredited Standards Committee of the Acoustical Society of America (ASA), which included
The modified "Schultz curve" displays the percent of exposed persons highly annoyed as a function of DNL (Schultz 1978). It is widely recognized that annoyance due to aircraft noise and community reaction can occur at noise levels that are below 65 dB (FICON, 1992b at 3-17; DoN, 2002 at 8). About 6-14% of the population will be highly annoyed by noise levels in the 60-65 dB zone (FICON, 1992b at 3-6). There is evidence that aircraft noise is more annoying than other transportation noises (Finegold et al., 1994 at 27).
3
Case 1:01-cv-00201-VJW
Document 219-2
Filed 10/05/2006
Page 10 of 45
9
the Department of Transportation and Department of the Navy. The 1998 revision of these guidelines is the Sound Level Descriptors for Determination of Compatible Land Use, ANSI S12.9-1998/Part 5 (ASA 1998). The guidelines were approved by the American National Standards Institute, Inc. (ANSI). For single-family residential land that involves extensive outdoor use, the Standards Committee recommends that DNL 55-65 is marginally compatible and DNL 65+ is incompatible (ASA, 1998 at 3). As indicated above, ARS-2 exposes more than 120,000 people to DNL levels of 65 dB and greater, compared to 28,000 people prior to the realignment of F-18 C/D aircraft to NAS Oceana. 24. Table 2 summarizes the federal agency and ASA guidelines. Federal agencies have established a uniform standard for compatibility of noise levels and residential land use. These agencies, including the FAA and U.S. Navy, regard noise exposure in excess of 65 dB as an environment that is normally unsuitable for residential use. Above 75 dB, the environment is clearly unsuitable for residential use. The ANSI standard of the Acoustical Society of America, which incorporates lifestyle factors, uses 65 dB as the upper limit for extensive outdoor residential living. The ASA standard shows the importance of accounting for lifestyle factors as part of the evaluation of the effects of noise exposure, which is critical for the land areas near the two Naval airfields. As shown below, economic studies of the effects of aircraft noise on residential property values demonstrate a negative effect of noise exposure above a background noise level of 55-60 dB.
Table 2. Residential Land Use: Federal Compatibility Guidelines, Day-Night Sound Level (DNL) Agency HUD: 24 CFR Part 51 FAA: 14 CFR Part 150 DoN: AICUZ (2) FICUN & FICON 65-70 dB 70-75 dB 75-80 dB 80-85 dB Unacceptable No No No No No No No No No 85+ dB
Normally Unacceptable No (1) No (1) No (1) No (1) No (1) No (1)
Residential Use is Incompatible at 65 dB and above ASA: Residential, Extensive Outdoor Use Guideline Notes: (1) Marginally compatible if indoor noise attenuation is used. "No" means non-compatible or not suitable (DoN, 1978 at 4-25). (2) DoN (2002 at Table 2); Web site for NAS Oceana at http://www.nasoceana.navy.mil. The Navy's revised AICUZ guidelines include a broad provision for protection of public welfare (DoN, 2002 at 13).
25. The guidelines for the Navy's AICUZ program clearly recognize the significance of noise exposures beginning at 65 dB and the critical nature of noise exposures of 75 dB or greater. The revised program guidelines (DoN 2002) include a new requirement for long-term analysis of aircraft operations. The Navy's revised guidelines also emphasize protection of general welfare of the community, which is a broader goal than health and safety considerations:
Case 1:01-cv-00201-VJW
Document 219-2
Filed 10/05/2006
Page 11 of 45
10
The study recommendations shall be based on current operations and the best available (5- to 10year) projections of operations that best support long-range planning controls to protect the health, safety and welfare of the community and the future operational integrity of the air installation (DoN, 2002 at 13; emphasis added). 26. Because noisy environments are undesirable to many individuals, existing single-family and other residences in the vicinity of a noisy airfield will be less valuable; that is, the adverse effects of noise will be (negatively) capitalized into the value of residential properties and registered in marketdetermined housing prices and rents. This effect can be measured using econometric methods and economic data on residential values at differing noise exposure levels. The measurement of these adverse effects is part of a large empirical literature on pollution and property values. The next section of this report reviews selected aspects of this literature and develops the general model of the effects of aircraft noise on property values. This is followed by a meta-analysis of the empirical evidence on the effects of aircraft noise on property values, which is used to derive monetary measures of the adverse and lasting effects of noise. Meta-analysis is a quantitative technique used to synthesize a comparable set of empirical results. Lastly, I discuss special features of ARS-2, including the unique operational features of Navy aircraft and the mild climate that affect residential living and lifestyles in Virginia Beach and Chesapeake, Virginia.
III. AIRCRAFT NOISE AND PROPERTY VALUES Economic studies of residential property values have examined the adverse effects on values of numerous pollutants, including aircraft noise. 27. Noise is unwanted or unpleasant sound. At 65 dB and above, the most common human effects associated with aircraft noise are annoyance, speech and learning interference, and sleep disturbance. In turn, these effects disrupt normal daily activities, such as conversation, television viewing, school work, productivity, outdoor recreation and living, and family activities. Annoyance is the adverse psychological response to a given noise exposure, including the anxiety or apprehension that the noise may cause (FICON, 1992b at 3-4). At noise levels above 75 dB, the Environmental Protection Agency (EPA, 1982a at 12) cautions that severe health effects may occur for some portion of the population, including temporary hearing loss. Those persons who are frequently outdoors are of greatest concern, including young children, retired people in warm climates, and persons in certain outdoor occupations (EPA, 1982a at 55; Suter, 1991 at 14). ARS-2 resulted in more than 23,000 people being exposed to severe noise levels of 80+ dB in CY99, compared to only four people in CY97.
Case 1:01-cv-00201-VJW
Document 219-2
Filed 10/05/2006
Page 12 of 45
11
28. Economists regard pollution as a negative "external cost" (externality) of production and consumption activities an uncompensated third-party effect whereby the polluter causes harm, inconvenience, annoyance, and other damages to the polluted. Aircraft noise is a classic example of an external social cost. Because market prices do not directly reflect these damages, the economic activities in question (e.g., aircraft operations) are subject to "market failure." This term means that the level of the economic activity fails to achieve an efficient allocation of resources as measured by the sum of the transaction surplus in the market less the amount of the uncompensated third-party damage costs. In general, market failures are not self-correcting due to costs of transaction and the public-good nature of pollution abatement efforts. 29. "Market failure" therefore reflects the absence of an explicit market for corrective actions that could reduce or avoid third-party damages the market for tranquility is missing.4 This means that direct estimation of demand schedules for tranquility (or externality avoidance) is not possible. As a consequence, economists resort to a variety of methods direct and indirect to measure the uncompensated costs of externalities. These methods include studies using surveys, political referendums, expenditures on mitigation (e.g., soundproofing), impacts on wages, and effects on residential property values (Freeman 1993; Palmquist 1991). Evaluation of the effects of pollution on property values using regression analysis the "hedonic property value" method is discussed in undergraduate and graduate textbooks in environmental economics (Tietenberg, 2000 at 41; Kolstad, 2000 at 313); urban economics (O'Sullivan, 2000 at 368; DiPasquale and Wheaton, 1996 at 189); costbenefit analysis (Boardman, 1996 at 459; Gramlich, 1990 at 71); and real estate valuation (Chinloy, 1988 at 40; Lusht, 1997 at 139). A recent EPA report, Guidelines for Preparing Economic Analyses, also contains a discussion of this method (EPA, 2000 at 77). 30. Consider two residential properties that are identical in all respects, except that the first house is located close to an airfield and the other is not (Nelson 1981). A but for analysis establishes that the adverse environment will result in a market value for the first house that is lower than the market value of the second house. This occurs because potential buyers reduce their demand for the first house relative to the second house, reflecting the discounted present value of the costs of future annoyance,
Airport noise problems have been dealt with by purchase of residential properties. Under the FAA's Part 150 Noise Compatibility Program, 240 airports received over $2.7 billion in AIP-grants for noise-related projects (GAO, 2000b at 80). Many of these projects are directed at noise levels of 75 dB and above. A GAO study of eight airports in 1989 found that $94 million had been spent in Atlanta on purchase of residential properties; $144 million in Los Angeles; and $100 million in Memphis (GAO 1989). The Seattle-Tacoma Airport spent over $260 million acquiring homes and land and insulating homes (GAO, 1998 at 3). The Navy's AICUZ program also provides for the acquisition of land or restrictive easements (DoN, 2002 at 33; EPA 1977b).
4
Case 1:01-cv-00201-VJW
Document 219-2
Filed 10/05/2006
Page 13 of 45
12
speech and learning interference, sleep disturbance, reductions in the quality of outdoor living, and loss of tranquility. A measure of these noise-induced damages is the difference between the marketdetermined value of the first house as compared to the second house. The but for analysis can be extended to analyze different levels of noise exposure (above background), since it has been established that annoyance and other adverse effects of noise rise predictably with increased exposure levels (EPA 1982a; FAA 1985; FICON 1992a, 1992b). Although there is a missing market for tranquility, there is a complementary market wherein individuals register their willingness to pay to avoid different levels of aircraft noise exposure. Hence, consumers reveal the implicit value that they place on tranquility by the explicit choices that they make in the housing market. The willingness to pay for tranquility is a part of the asset price of the "housing bundle," and econometric techniques (regression analysis) are available that unbundle complex products and thereby reveal the implicit or "hedonic price" for tranquility. A large empirical literature has developed using this method. 31. Using the hedonic price model and regression analysis, economists and real estate researchers have empirically investigated the adverse effects on residential properties of numerous forms of pollution. Several surveys of this large literature are available, including Bartik and Smith (1987) on urban amenities; Boardman et al. (1997) for external costs; Donnelly (1991) and Boyle and Kiel (2001) for environmental disamenities; Farber (1998) for health and amenity risks; Smith and Huang (1993, 1995) for air pollution; Nelson (1978, 1982) for highway noise; and Nelson (1978, 1980, 2004) for airport noise. Follain and Jimenez (1985), Freeman (1993), Palmquist (1991, 2005), and EPA (2000) provide useful surveys of methodological issues. This technique also is widely used in the field of real estate valuation, and articles appear regularly in real estate journals that apply regression analysis to housing values and residential amenities and disamenities. For example, recent articles in The Appraisal Journal estimate the value of proximity to water (Benson et al. 2000), adverse effects of a Superfund site (Reichert 1997), and adverse effects of nearby railroad tracks (Simons and El Jaouhari 2004). 32. Econometric studies that investigate the determinants of residential property values are collectively referred to as "hedonic price" or "hedonic property value" studies (EPA, 2000 at 77). It is rare that two residential properties are identical in all respects, except for the pollutant in question. Consequently, in order to isolate a given hedonic price, it is necessary to control statistically for other influences on property values, such as the size of house and lot, quality of construction, design of the house, merits of the neighborhood, quality of local schools, crime rates, governmental services, and so forth. Some of these characteristics will vary little within a given data set, and separate measurement is not required to explain the observed variation in property values. In other cases, the excluded characteristics are uncorrelated ("orthogonal") with the included pollutant, and do not bias the resulting
Case 1:01-cv-00201-VJW
Document 219-2
Filed 10/05/2006
Page 14 of 45
13
estimate of the hedonic price for the pollutant in question. Random errors in the measurement of property values are captured in the regression error term, and do not bias the estimates. 33. Given differences in statistical methods, samples, time periods, and urban locations, empirical studies have not produced a singular value for the effects of aircraft noise on property values. Indeed, it would be very surprising if they did because demand-supply conditions for housing markets tend to be local or regional, rather than national. Nevertheless, I show below that there is a broad scientific consensus on the measurable effects of aircraft noise. All hedonic price studies have shown that aircraft noise has a negative effect on residential property values, and central tendencies clearly exist for the monetary damages. The technique of meta-analysis can be used to synthesis the damage estimates and increase the reliability of summary measures for a set of comparable estimates. Annoyance due to noise exposure can be incorporated in the economic model of housing demand and supply, which underlies hedonic property value studies for aircraft noise. 34. Each house and lot represents a unique combination of characteristics and attributes, which means that the decision to purchase a given property is complex. However, if the characteristics are provided in various combinations, it is possible to estimate an implicit price function that shows how the values of properties vary due to changes (only) in a given characteristic. This is the but for relationship. Formally, let V be a sample of observations on housing prices; S is a vector of structural variables (house size, number of bedrooms, lot size, etc.); L is a vector of locational variables; T is a vector of local taxes; G is a vector of local government services; and E is a vector of local neighborhood environmental-quality variables (Nelson, 2004 at 6). The market-determined asset value of the houses in the sample is given by V = V(S, L, T, G, E). The marginal implicit price for each characteristic represents the increase in expenditures required to obtain one more unit of that characteristic, holding constant other variables that affect the value of a house. The method of regression analysis is used to insure the constancy of other influences on property values. Mathematically, the marginal price is the partial derivative of the V relationship with respect to the j-th characteristic, or MV/MEj. The marginal price function can be linear or non-linear (Nelson, 2004 at 6). 35. Many empirical studies employ a non-linear function for property values. For aircraft noise, the hedonic function can be represented by V = b0Zb1Ab2u1, where V is the property value; Z is the set of physical and locational characteristics (that is, S, L, T, and G); A is subjective annoyance due to aircraft noise; u1 is a stochastic error term; and b0, b1, and b2 are parameters of the relationship. Subjective annoyance can be approximated by the following semi-logarithmic relationship: A = c0ec1(DNL)u2, where DNL is the day-night sound level in decibels; u2 is a stochastic error term; e is the natural log base; and c0
Case 1:01-cv-00201-VJW
Document 219-2 14
Filed 10/05/2006
Page 15 of 45
and c1 are parameters of the semi-logarithmic relationship.5 36. Taking natural logs, and substituting for ln A, yields the following logarithmic relationship: ln V = d0 + d1(ln Z) + d2 DNL + u3, where d2 = b2c1, etc. (Nelson, 2004 at 6). This relationship is estimated using regression techniques. The regression coefficient d2 × 100 (= MV/MDNL C 1/V) represents the percentage decrease in a given property value resulting from a one dB increase in noise exposure on the DNL scale. For this relationship, the marginal implicit price is given by MV/M(DNL) = d2V, and the noise-housing price elasticity is d2(DNL). The main findings of aircraft noise studies can be summarized by defining a Noise Depreciation Index (NDI), which shows the percentage depreciation per dB. 37. The main findings of empirical studies of aircraft noise can be summarized, with some adjustments, by means of a Noise Depreciation Index (NDI), which was originally developed by Walters (1975 at 102). For two properties that differ but for their level of noise exposure, the absolute amount of housing depreciation per dB (the unit cost of noise) is given by D = (difference in the total noise discount) ÷ (difference in noise exposure in dB). Dividing D by the price of the given house (or an average house price), the percentage rate of deprecation is given by NDI = (D ÷ property value) × 100 = (difference in total percentage deprecation) ÷ (difference in noise exposure in dB). This is the same result as d2 × 100 in the logarithmic model of residential property values. 38. Many empirical studies use the logarithmic model, and these studies directly estimate the NDI. The coefficient on DNL multiplied by 100 is the percentage change in a given property value due a one dB change in noise exposure (above background). Depending on the functional form, some studies estimate the marginal implicit price or the noise-housing price elasticity. In these cases, appropriate mathematical relationships are used to recover estimates of the NDI. Some empirical studies use an older noise metric, the Noise Exposure Forecast (NEF). The relationship between DNL and NEF is given by DNL = NEF + 35 dB (EPA, 1982a at B-3). The summaries below reflect this adjustment, and all noise levels are expressed using the DNL noise metric.
IV. EMPIRICAL STUDIES OF AIRCRAFT NOISE AND PROPERTY VALUES 39. Several summaries of empirical studies of civilian aircraft noise and property values are available, including Nelson (1978, 1980, 2004) and FAA (1985). For this report, I will first summarize
This annoyance relationship is not precisely the modified "Schultz curve" (Finegold et al. 1994), which covers all forms of transportation noise. The Schultz curve includes low levels of noise exposure (40-50 dB), which provide the non-linear left-hand tail of the curve. These noise exposure levels are omitted from aircraft noise studies or included in the background level. The semi-log function is due to Bishop (1966).
5
Case 1:01-cv-00201-VJW
Document 219-2
Filed 10/05/2006
Page 16 of 45
15
the NDI results in the FAA's (1985) report on Aviation Noise Effects. Second, I will summarize my recent meta-analysis (Nelson 2004), which updates and extends my earlier reviews for U.S. and Canadian airports. All of the empirical studies cover noise from civilian aircraft and the residential environment at civilian airports. All of the studies apply the hedonic model. The next section of this report, Section V, discusses special features of military aircraft operations and the residential living environment of Virginia Beach and Chesapeake, Virginia. In Section V, I also summarize the results from two survey studies of the effects of aircraft noise on residential property values, which are especially important for noise exposure levels of 80 dB and above. The monetary damages developed below isolate the effect of aircraft noise on property values, holding constant other influences such as accessibility and general growth of nominal housing values. Where the passage of time is important, researchers have included variables to account for general changes in housing market demand and supply and other macroeconomic conditions. The FAA found that aircraft noise decreases the value of residential property by approximately -1% per dB. This estimate is based on results from seven hedonic studies that use data for civilian aircraft and airports. 40. The FAA's (1985 at 100) report on Aviation Noise Effects includes a discussion of the effects of aircraft noise on the value of residential property located around civilian airports. The FAA selected 10 best estimates of the NDI from the range of available values, including three estimates for 1960 and seven estimates for 1967-70. The three NDI values for 1960 are about -2% per dB, which reflects adjustments in housing markets due to early growth of commercial jet aircraft as a mode of transport. The values for 1967-70 include estimates of -1.5% for San Francisco and -1.0% for Washington, D.C., which may reflect climate and other lifestyle considerations in these areas. After omitting the three estimates from the 1960s, the FAA's range of estimates is -0.6% to -1.5% per dB (FAA, 1985 at 101). The FAA (1985 at 101) concluded that noise decreases the value of property by approximately -1% per decibel. As shown below, the FAA's estimate of the NDI is well within the range of values found by more recent empirical studies of the aircraft noise-property value relationship.6 Empirical studies produce a range of estimates for the NDI from -0.3% to -1.5% per dB. The simple mean NDI is -0.75% per dB, which is close to the FAA's estimate. This estimate is based on information from 23 U.S. and Canadian airports and 33 estimates of the NDI.
The FAA notes that "... all research conducted in this area found negative effects from aviation noise ... researchers have been careful to consider [other factors] and to normalize their influences in research studies" (FAA, 1985 at 100).
6
Case 1:01-cv-00201-VJW
Document 219-2 16
Filed 10/05/2006
Page 17 of 45
41. Since the publication of my survey article (Nelson 1980), there have been additional studies of the empirical relationship between aircraft noise and property values. In Nelson (2004), I updated and extended my earlier survey using the technique of meta-analysis. Table 3 displays the results from 20 hedonic studies that cover 23 different civilian airports in the U.S. and Canada, and includes studies using sample data from 1967 to 1995. A total of 33 best estimates of the NDI is presented. Two independent estimates are available for Atlanta, Dallas, Reno, San Francisco, St. Louis, and Washington, D.C. A variety of data are employed, including data for census tracts, census blocks, disaggregated census blocks, and individual housing sales (e.g., Multiple Listing Service data). The NDI estimates also reflect a variety of model specifications and variables, which capture sample differences in housing and neighborhood characteristics, employment accessibility, airport accessibility, governmental services, and other economic and environmental features. As previously mentioned, it is not expected that the NDI will have a singular value, and the range of NDI values is -0.28 to -1.49%. The simple mean NDI in Table 3 is -0.75% per dB (Nelson, 2004 at 14). This average implies that each dB increase in noise exposure will reduce property values by -0.75%, holding constant other housing and land characteristics. This value is in substantial agreement with the FAA's estimate of -1.0% per dB. 42. The hedonic estimates in Table 3 uniformly indicate that aircraft noise is negatively capitalized into residential property values. Twenty-seven of 31 estimates are statistically significant at the 95% level or better (estimates for JFK and La Guardia Airports are missing standard errors). This is a far greater number of significant negative estimates than could possibly be attributed to chance. The coefficient standard errors and other statistical measures (e.g., R-square) are found in my attached article. The technique of meta-analysis can be used to synthesize the set of NDI estimates contained in Table 3. A meta-regression analysis produces a best estimate for the mean NDI of -0.7% per dB. This value is not affected by moderating and mediating variables such as sample size, sample year, mean property value, or airport accessibility. 43. Meta-analysis is a statistical procedure for integrating and synthesizing the quantitative results contained in a set of studies of a given empirical relationship or outcome. That is, meta-analysis is "an analysis of the results of statistical analyses" (Hedges and Olkin, 1985 at 13). According to Hedges and Olkin (1985 at 1), "replication of experimental results has long been a central feature of scientific inquiry, and it raises questions concerning how [best] to combine the results obtained." Metaanalysis was developed in the early 1980s and applied originally in psychology and education (Cook et al. 1992; Cooper and Hedges 1994). It is now widely applied in the physical sciences and other areas of the social sciences. In economics, meta-analysis has been applied to environmental issues (Smith and Huang 1995; van den Bergh 1997), recreation demand (Rosenberger and Loomis 2000), gasoline demand
Case 1:01-cv-00201-VJW
Document 219-2 17
Filed 10/05/2006
Page 18 of 45
Table 3. Summary of Noise Depreciation Indexes (NDI) for 23 Airports Hedonic Study Airport BAH-FAA (1994 at 18) Baltimore BAH-FAA (1994) Los Angeles BAH-FAA (1994) NYC JFK BAH-FAA (1994) NYC La Guardia Blaylock (1977 at 84) Dallas DeVany (1976 at 213) Dallas Dygert (1973) San Francisco Dygert (1973 at 113) San Jose Emerson (1972) Minneapolis Espey & Lopez (2000) Reno Fromme (1978) Washington, DC Levesque (1994 at 207) Winnipeg Mark (1980 at 112) St. Louis Maser et al. (1977) Rochester urban Rochester suburban McMillian (1980) Edmonton Mieszkowski et al. (1978) Toronto Toronto (Etobicoke) Myles (1997 at 21) Reno Nelson (1978) Washington, DC Nelson (1979, 1980) Buffalo Nelson (1979, 1980) Cleveland Nelson (1979, 1980) New Orleans Nelson (1979, 1980) St. Louis Nelson (1979, 1980) San Diego Nelson (1979, 1980) San Francisco Nelson (1979, 1980, 1981) six airports O'Byrne et al. (1985) Atlanta (houses) Atlanta (census blocks) Price (1974) Boston Tarassof (1993 at 83) Montreal DNL 60 & 70 dB 60 & 70 dB 60 & 70 dB 60 & 70 dB 55-90 dB 55-85 dB 60-80 dB 60-80 dB 60-80 dB 60-75 dB 55-70 dB 75+ dB 70 & 80 dB 65+ dB 65+ dB 55-70 dB 55-70 dB 55-70 dB 55-75 dB 55-70 dB 60-80 dB 60-80 dB 60-80 dB 60-80 dB 60-80 dB 60-80 dB 60-80 dB 65-80 dB 60-80 dB 60-80 dB 55-70 dB Sample Size (N) and Data (Year) N = 30; individual houses (1990) N = 24; individual houses (1991) N = 30; individual houses (1993); no std. error N = 30; individual houses (1993); no std. error N = 4264; disaggregated census blocks (1970) N = 1270; census blocks (1970) N = 128; census tracts (1970) N = 198; census tracts (1970) N = 222; individual houses (1967) N = 1596; individual houses (1991-95) N = 28; census tracts (1970) N = 1635; individual houses (1985-86) N = 6553; individual houses (1969-70) N = 398; individual houses (1971) N = 990; individual houses (1971) N = 352; individual houses (1976) N = 509; individual houses (1969-73) N = 611; individual houses (1969-73) N = 4332; individual houses (1991) N = 52; census tracts (1970) N = 126; census blocks (1970) N = 185; census blocks (1970) N = 143; census blocks (1970) N = 113; census blocks (1970) N = 125; census blocks (1970) N = 153; census blocks (1970) N = 845; census blocks (1970) N = 96; individual houses (1979-80) N = 248; census blocks (1970) N = 270; apt. rentals by census tracts (1970) N = 427; individual houses (1989-90) NDI -1.07 -1.26 -1.20 -0.67 -0.99* -0.80* -0.50* -0.70* -0.58* -0.28 -1.49* -1.30* -0.56* -0.86* -0.68* -0.51* -0.87* -0.95* -0.37* -1.06 -0.52* -0.29* -0.40* -0.51* -0.74* -0.58* -0.55* -0.67* -0.64* -0.81* -0.65* -0.65* -0.90*
Uyeno et al. (1993 )Vancouver (houses) 60-75 dB N = 645; individual houses (1987-88) Vancouver (condos) 60-75 dB N = 907; individual condos (1987-88) Notes: Derived from Nelson (2004 at 12), attached. Asterisks indicate statistically significant at the 95% level.
Case 1:01-cv-00201-VJW
Document 219-2
Filed 10/05/2006
Page 19 of 45
18
(Espey 1996, 1998), taxation (Phillips and Goss 1995), and many other empirical areas. Summaries of economic applications of meta-analysis are found in Stanley (2001) and Stanley and Jarrell (1989); see also Finkelstein and Levin (2001). 44. In order to conduct a meta-analysis, several steps are necessary (Hedges and Olkin, 1985 at 9; Cooper and Hedges, 1994 at 9). First, the relevant empirical literature must be identified, including unpublished works and statistically weak results. Second, a common "effect size" must be available or can be constructed from each of the studies. The effect size used here is the noise depreciation index or NDI. Third, from the relevant literature, the investigator must extract and code the indicators and variables that ". . . predict outcomes, potential mediators of effects, and the differences in how outcomes are conceptualized" (Cook and Hedges 1994 at 11). Fourth, statistical techniques are applied to the effect size, which summarizes the set of estimates and explains any systematic differences in effect sizes. Weighting of the effect sizes according to the precision of the estimate is emphasized. Hence, standard errors of the effect sizes are a necessary part of the analysis. 45. My 2004 article includes several levels of analysis. First, I compute simple (unweighted) mean and median NDIs. Second, I compute two weighted means that use as weights the inverse variance of each estimate. Hence, more precise estimates in Table 3 are given greater weight. Both fixed-effects and random-effects means are computed, which allows for different sampling and statistical procedures by empirical investigators. Third, a meta-regression analysis is conducted that seeks to identify systematic differences among the estimates that can be traced to measurable differences in sampling and statistical techniques. In the regression analysis, seven variables were identified that might systematically increase (or decrease) a given NDI estimate relative to the other estimates. These variables are: (1) sample country (USA or Canada); (2) sample time period (before or after 1970); (3) sample size (log of number of observations); (4) data type (census data or individual sales); (5) mean property value (in 2000 US dollars); (6) functional form (logarithmic or linear) and (7) airport accessibility (accounted for in the study or not). The meta-regression analysis reveals that only the sample country and functional form are important statistically (Nelson, 2004 at 19). 46. Table 4 summarizes the results of the meta-analysis. The full set of results is found in the attached copy of my article. The unweighted estimates tend to be somewhat larger (i.e., more negative) than the weighted estimates, reflecting the inclusion of JFK and La Guardia Airports. The fixed-effects and random-effects weighted means are very close in magnitude (-0.58 and -0.59). The meta-regression using the inverse variance weights produces the smallest estimate, -0.51, but this estimate does not employ the optimal set of weights (Saxonhouse 1976). The meta-regression using inverse standard error weights yields a synthetic NDI of -0.67 (std. error = 0.20). The 95% confidence interval associated with
Case 1:01-cv-00201-VJW
Document 219-2
Filed 10/05/2006
Page 20 of 45
19
this estimate is -0.67 ± 1.96 (0.20) = -0.28 to -1.06. Among the estimates reported in Table 3, the 95% confidence interval includes all NDI values with the exception of four values (Baltimore, Los Angeles, Washington, D.C., and Winnipeg). The large negative estimates for JFK and La Guardia Airports are not included in the regression analysis due to lack of standard errors.
Table 4. Summary of the Meta-Analysis of Airport Noise and Property Values Statistical Measure Simple mean NDI (unwt.) Median NDI (unwt.) Fixed-effect mean NDI (weighted) Random-effect mean NDI (weighted) Meta-regression coefficient. (unwt.) Meta-regression coefficient (wt.) NDI value (std.dev.) -0.75 (0.30)* -0.67 -0.58 (0.04)* -0.59 (0.04)* -0.83 (0.31)* -0.51 (0.14)* Comments on Measure Ignores precision; includes JFK and La Guardia Ignores precision; includes JFK and La Guardia Uses inverse variance weights; sampling error only Uses inverse variance weights adjusted for between-study variability of estimates White's correction for heteroskedasticity Wt. regr. with inverse variance weights
Meta-regression coefficient (wt.) -0.67 (0.20)* Wt. regr. with optimal inverse standard error weights Source: Nelson (2004); attached. Asterisks indicate statistically significant at the 95% level.
47. An NDI of -0.67% per dB is consistent with my earlier review of the empirical literature. For example, in Nelson (1980), I found that the simple mean NDI for a sample of twelve airports was -0.62% (Nelson, 1980 at 43). In a pooled statistical analysis for six airports, I reported a mean NDI of -0.55% (Nelson, 1979 at 327). Recent studies for other countries also agree with this result. For example, a recent study for Geneva, Switzerland used geographical information system (GIS) data and found a noise discount of -0.7% per dB (Baranzini and Ramirez 2005). Lastly, studies of airport expansions and closures also have found property value effects from anticipated changes in noise exposure (Jud and Winkler 2005; Konda 2002). While these studies do not estimate an NDI, they do serve to demonstrate that market dynamics do not greatly alter the negative effects. In summary, 27 of 31 estimates of the NDI are negative and statistically significant at the 95% level or better. This is convincing evidence that aircraft noise is negatively capitalized into residential property values. A synthesis of this evidence using meta-analysis yields an NDI of about 0.7% per additional dB (above background noise levels).
Case 1:01-cv-00201-VJW
Document 219-2
Filed 10/05/2006
Page 21 of 45
20
Combing the results in Table 4 and the FAA's (1985) estimate, the best NDI estimates are -0.7% to -1.0% per dB. These values measure the diminution in property values due to a one dB increase in noise exposure (DNL) above the background noise level (50-60 dB), and should be applied to noise exposure for residential properties located in the 65-79 dB zone. 48. In Table 3, the background noise level in most studies is about 60 dB. The maximum noise level in most studies is 80 dB. The change in noise exposure due to ARS-2 was about 8 to 20 dB (Wyle, 1998 at 4-37). Using these values and the NDI estimate of -0.70% per dB, a 5 dB change in noise due to ARS-2 reduces a given property value by -3.5% (= -0.70% × 5); a 10 dB change in exposure reduces a given property value by -7%; and a 15 dB change in exposure reduces a given property value by -10.5%. At the FAA's (1985) higher value of -1.0% per dB, the diminution estimates are -5%, -10%, and -15%, respectively. 49. In order to illustrate the general calculation of monetary damages using the mean NDI, information was obtained on the assessed value of an average residential property in Virginia Beach. According to the Office of Real Estate Assessor, all residences (non-apartment units) had an average assessed value of $129,800 in January 2000 (City of Virginia Beach, 2000 at 6), compared to $125,290 in January 1999. The average assessed value for July 1, 1999 was about $127,545. Hence, based on an average assessed value of $128,000, the diminution in value due to ARS-2 is $4,480 to $6,400 for a 5 dB change; $8,960 to $12,800 for a 10 dB change in exposure; and $13,440 to $19,200 for a 15 dB change. As discussed below, it is an important policy question whether or not properties exposed to 75+ dB can continue to be used for residential purposes. In the next section, I show that the adverse effect of ARS-2 will be at the upper limit of these estimates, or -1% per dB change in exposure. I also demonstrate that the NDI is larger for residential properties exposed to 80 dB or more. 50. Average assessed residential property values in Virginia Beach rose by 3.60% between January 1999 and January 2000. Average assessed values rose to $158,400 in 2003 (FY 2004). Compared to 2000, this is a compound growth rate of 6.9% per annum. The rate of increase from 2003 to 2004 was 11.7%. The rate of increase from 2004 to 2005 was 22.3%. In the past two years, property values rose rapidly in Virginia Beach (City of Virginia Beach, 2005 at 7). However, this factor only affects the general level of nominal values, and does not change the supply and demand for housing in noisy areas. The correct economic issue is the difference in housing values at different noise exposure levels. Suppose that due to ARS-2 a residential property valued at $100,000 in 1998 was reduced in value by 20% or $20,000. Growth of property values over time means that this property might have a value at $133,500 in 2005. A similar $100,000 house in a quiet area would be worth about $167,000. While both nominal values increased by 67%, the difference in relative values is 20% in 1999 and 2005.
Case 1:01-cv-00201-VJW
Document 219-2
Filed 10/05/2006
Page 22 of 45
21
V. SPECIAL EFFECTS OF MILITARY AIRCRAFT NOISE The special effects of military aircraft noise that exceeds 75 dB are not fully captured by the NDI values of -0.7% and -1.0% per dB change in noise exposure. 51. Only three of the hedonic studies in Table 3 include noise exposures in excess of 80 dB, and even the populations exposed to 75-80 dB are limited in most studies. Hence, most of the studies in Table 3 fail to consider severe noise exposure of 80 dB and above. For example, O'Byrne et al. (1985) examine the effects of aircraft noise in the vicinity of Atlanta's Hartsfield International Airport. For 1970 census blocks, only 12% of the observations are in the 75-80 dB zone, and for 1979-80 individual house sales, only 7% of the observations are in the 75-80 dB zone. As shown in Table 1, ARS-2 subjected 23,000 people to 80+ dB. In this section, I examine evidence indicating that property value losses due to miliary aircraft noise will be seriously understated if the special effects of severe noise exposure are not considered. For this purpose, I review the evidence on: (1) survey studies that show that housing markets near airports are segmented by noise exposure in excess of 75 dB; (2) the adverse effects of severe noise levels on human health and welfare; (3) the unique features of Navy aircraft operations; and (4) the residential living environment and lifestyle of Virginia Beach and Chesapeake, Virginia. 52. The discussion that follows is based on the assumption that the land areas exposed to 80+ dB will continue to be used for some residential purposes. It is not clear that this should be the case. Four U.S. federal agencies EPA, HUD, FAA, and the Department of the Navy have established explicit guidelines for noise exposure of 75+ dB, which indicate that such levels of exposure are incompatible with residential land use. For example, in its National Strategy for Noise Control, one of EPA's specific national goals was expressed at follows: Reduce environmental noise exposure of the population to a DNL value of no more than 75 dB immediately, using all available tools, except in those isolated cases where this would impose hardship. This will essentially eliminate risk of hearing loss due to environmental noise, and red
United States Free Forms
Canada Free Forms
United Kingdom Free Forms
Australia Free Forms