oversight

Aviation Noise: A National Policy is Needed

Published by the Government Accountability Office on 1990-09-27.

Below is a raw (and likely hideous) rendition of the original report. (PDF)

                     United States General Accounting Office

                     Testimon y
GAO


For Release           Aviation Noise : A National Policy is Neede d
on Delivery
Expected a t
9 :30 a .m . ED T
Thursda y
September 27, 1990




                      Statement o f
                      Kenneth M . Mea d
                      Director, Transportation Issue s
                      Resources, Community, and Economi c
                      Development Divisio n


                      Before th e
                      Subcommittee on Aviatio n
                      Committee on Public Work s
                        and Transportatio n
                      House of Representative s




GAO/T-RCED-90-112
                                                               GAO Form 160 (12/67)
Mr . Chairman and Members of the Subcommittee :

     We are here today to present our assessment of the economi c
impact that noise restrictions have had and are likely to have o n
the aviation industry . We will also be discussing the polic y
issues that need to be resolved in developing a national nois e
policy . Our work was done at the request of the Subcommittee an d
Representative Vento .

     Passage of the Airline Deregulation Act of 1978 led to a n
increase in the number of aircraft operations and the correspondin g
level of noise at many airports . This noise can impose substantia l
costs on the communities located near airports . Airport
proprietors have reacted to the rising noise level by adoptin g
noise restrictions on aircraft use that may, in some instances ,
restrict access to airports . We have issued several reports an d
testimonies dealing with aviation noise issues . '

       The analysis we are presenting today is based on a survey we
conducted of the 140 U .S . airports that have predominantly je t
service ; our analysis of various studies that have been done of th e
economic costs of noise restrictions ; and an extensive series o f
interviews with airline, airport, and aircraft industry officials .
A complete report on our work will be issued around the end of thi s
year .




'See, for example, Aircraft Noise : Im•lementation of FAA' s
Expanded East Coast Plan (GAO/RCED-88-143, Aug . 5, 1988), Aircraft
Noise :	 Status and Management of FAA's West Coast Pla n (GAO/RCED -
89-84, May 8, 1989), Aircraft Noise :	 Eight Airports' Efforts t o
Mitigate Noise (GAO/RCED-89-189, Sept . 14, 1989), an d
TransportationNoise : Federal Control and Abatemen t
Responsibilities May Need to Be Revised (GAO/RCED-90-11, Oct . 12 ,
1989) . We will also be testifying this afternoon on a related se t
of issues before the Subcommittee on Transportation, Aviation, an d
Materials, House Committee on Science, Space, and Technology .
                                  1
     Overall, our review shows that, in the absence of a nationa l
noise policy, airports will likely continue to implement a variet y
of noise abatement measures in an attempt to respond to the
concerns of their local communities . While these measures in man y
cases are needed to reduce the impact of airport noise on loca l
communities, the lack of coordination among these local airpor t
measures could place an increasingly heavy burden of higher cost s
and inefficient use of aircraft on the nation's air transportatio n
system . A national policy specifying a schedule for phasing ou t
the older, noisier stage 2 aircraft would provide for a more
orderly transition, by a date certain, to a quieter stage 3 flee t
for the nation as a whole . 2 The implementation of such a polic y
could pose trade-offs between meeting the needs of loca l
communities and reducing the burdens which these needs impose o n
the nation's air transportation system .

     Specifically, we found the following :

     -- Based on our survey, the number of airports requiring th e
        use of the quieter stage 3 aircraft will grow slowly unti l
        1995, but will increase rapidly between 1995 and 2000 . Our
        survey shows that, by the year 2000, 41 percent of th e
        nation's 29 largest airports plan to have banned stage 2
        aircraft .

        The costs of delays and inefficient use of aircraft impose d
        on airlines by the uncoordinated adoption of nois e
        restrictions by airports appear to be modest now . However ,

2 Part 36 of Title 14 of the Code of Federal Regulations prescribe s
noise emission standards for the manufacture and certification o f
aircraft . Aircraft which do not meet these standards are commonl y
referred to as "stage 1" aircraft . These relatively noisy aircraf t
were no longer generally allowed to be operated in the Unite d
States after 1984 . Aircraft meeting minimal standards issued i n
1968 are commonly referred to as "stage 2 ." Quieter aircraf t
meeting more stringent standards promulgated in 1977 are commonl y
referred to as "stage 3 ."
                                 2
        we expect the proliferation of these noise restriction s
        after 1995 to cause these costs to grow rapidly .

     -- The aircraft and airline industries believe they have th e
        capacity to phase out stage 2 aircraft by the year 2000 ,
        but said they will need 10 years advance notice to achieve
        this goal . We estimate that the costs of phasing out stag e
        2 aircraft by 2000 are likely to fall between $2 billio n
        and $5 billion, depending on whether airlines replace o r
        retrofit their existing stage 2 fleets . 3

     -- The federal government will have to establish a nationa l
        noise policy now if it wants the aircraft and airlin e
        industries to phase out stage 2 aircraft in an orderl y
        fashion . The proliferation of uncoordinated airport nois e
        restrictions after 1995 is likely to create growing cost s
        in inefficient use of aircraft . A national schedule fo r
        phasing out stage 2 aircraft will permit a more orderl y
        phase-out of stage 2 aircraft and allow the nation t o
        achieve a quiet stage 3 fleet .

     I will now discuss our findings in greater detail .

LOCAL NOISE RESTRICTIONS EXPECTE D
TO PROLIFERATE	 AFTER 199 5

      The Federal Aviation Administration (FAA) is authorized, unde r
49 U .S .C . Section 1431, to regulate aircraft noise . In 1976, FAA
issued regulations under this provision requiring that al l
aircraft used at U .S . airports meet stage 2 standards starting i n
1985 : FAA also required that all aircraft designs newly certifie d
after 1977 meet more stringent stage 3 standards ; as a result, al l
aircraft now being built are stage 3 aircraft . However, the FAA' s


3 A11 values are expressed in constant dollars .
                                 3
regulations do not require the airlines to replace their old stag e
2 airplanes with the quieter stage 3 aircraft . Airlines can buy o r
lease used stage 2 aircraft to expand their fleets . Airports, o n
the other hand, can set more stringent noise standards, and a
number of airports, in response to pressures from local communitie s
affected by aircraft noise, have done so . While noise restriction s
are expected to spread slowly between now and 1995, our surve y
shows that airport officials expect these restrictions t o
proliferate between 1995 and 2000 .

     We surveyed airports to determine what noise restrictions the y
had now, what restrictions they expected to introduce between no w
and 1995, and what restrictions they expected to introduce betwee n
1995 and 2000 . We found that only three of the 140 airports in ou r
survey currently ban the use of stage 2 aircraft . 4 Another 1 6
airports restrict their use by limiting the percentage of stage 2
aircraft used by each airline and/or by banning them at night . By
1995, only one additional airport told us that it is very likely t o
become all stage 3, while another 13 indicated that they are
somewhat likely to . However, by 2000, 34 additional airports ar e
very likely to become all stage 3, including 12 of the 29 larges t
airports . Seventy-seven airports are at least somewhat likely t o
ban stage 2 aircraft by 2000, including 20 of the 29 larges t
airports .

     Representatives of the airline industry believe that if a
substantial number of large airports adopt rules banning stage 2
aircraft, these rules will amount to a de facto ban on stage 2
aircraft anywhere in the country . Aircraft are normally scheduled
to fly to a number of different airports as part of their regula r
routing . If a significant proportion of the airports require al l
stage 3 aircraft, an airline will have to acquire an all stage 3


4They are Orange County, Long Beach, and Burbank airports i n
California .
                                4
fleet (by replacement or retrofitting) or be burdened with an
inefficient routing system designed to steer stage 2 aircraft awa y
from all-stage-3 airports .


UNCOORDINATED AIRPORT NOISE RESTRICTION S
ARE LIKELY TO IMPOSE GROWING COSTS ON AIRLINE S

     The independent and uncoordinated imposition of local nois e
restrictions by the nation's airports is often characterized b y
airline industry officials as a "patchwork quilt" style o f
regulation . Airline industry officials generally believe that thi s
style of regulation threatens the efficient functioning of th e
national transportation system by causing delays and the uneconomi c
use of aircraft to service certain markets .


     Nighttime noise curfews sometimes compound the effects o f
delays of evening flights caused by bad weather . If the delayed
flight is using a stage 2 aircraft, which cannot operate after th e
curfew, the flight either will have to be cancelled or will have t o
be redirected to an alternative airport not subject to a curfew .
If the flight is redirected, passengers are further delayed i n
reaching their destination . Alternatively, if mechanical problem s
ground a stage 3 aircraft, and the only back-up plane is stage 2 ,
the flight might have to be cancelled . An official representin g
the air freight industry noted that nighttime restrictions strik e
directly at the ability of air cargo carriers to provide overnigh t
service .


     Scheduling of aircraft for particular flights sometime s
results in inefficient use of aircraft, because the limited numbe r
of stage 3 aircraft must be used at the airports that require thei r
use, rather than in the markets where their size is mos t
appropriate . For example, airlines cited examples of being force d
to use B757 (stage 3) aircraft to meet noise restrictions whe n
smaller B737 (stage 2) aircraft would be better suited to traffi c
levels in the market . Other carriers cited being required to us e
                                 5
(stage 3) B737-300s instead of more appropriately sized (stage 2 )
DC-9s to meet noise restrictions .


     Air carrier officials also told us that aircraft require les s
thrust and generate less noise when they are less than full y
loaded . As a result, noise restrictions that are specified i n
terms of allowable decibels (rather than in terms of the aircraf t
being stage 2 or stage 3) sometimes cause airlines to fly plane s
with less than a full load to reduce noise . This limits th e
airlines' ability to make the most efficient use of their aircraft .
For example, one carrier flies B737-300s out of Orange Count y
Airport in California with less than full loads to meet the nois e
restrictions there .


     These costs arise as a result of having to meet separat e
restrictions at each airport . They would not arise if the same
amount of noise reduction were achieved as a result of the airline s
being required to meet national requirements that a certai n
percentage of their fleet meet stage 3 standards .


     These costs have caused some air carriers to reduce service o n
some routes involving noise-controlled airports . Carriers reporte d
reducing service, for example, at Orange County, Long Beach, Sa n
Francisco, and Boston as a result of noise restrictions . One
airline abandoned a route because it would have had to mak e
uneconomic use of its stage 3 aircraft, which were larger than th e
route required . While the number of such cases that have bee n
reported to us so far has been small, further service cutbacks ma y
occur as additional airports impose noise restrictions .

     The extent of these costs under present conditions appears t o
be moderate . While the airlines were able to provide examples o f
many of these costs, they were not able to quantify the extent o f
these costs . However, these costs are expected to increase in th e


                                6
future as the number of airports with noise restrictions increases ,
particularly after 1995 .

     Local airport noise restrictions may also affect othe r
airports because the imposition of noise restrictions at on e
airport could cause "dumping" of older, noisier aircraft on othe r
airports . Our survey indicated that 23 airports reported receiving
higher levels of noise as a result of the imposition of nois e
restrictions at other airports . Conversely, 24 other airport s
reported they were experiencing lower levels of noise, suggestin g
that the requirement to use stage 3 aircraft at some airports mad e
those aircraft available at other airports as well .

A BAN ON STAGE 2 AIRCRAFT IS FEASIBLE `
BUT WILL COST $2 .2 - $4 .6 BILLIO N


     The aircraft and airline industries believe they have th e
capacity to phase out stage 2 aircraft by the year 2000, but sai d
they will need 10 years advance notice to achieve this goal . We
estimate that the costs of phasing out stage 2 aircraft by 2000 ar e
likely to fall between $2 billion and $5 billion, depending o n
whether airlines replace or retrofit their existing stage 2 fleets .

Stage 3 Technology Availabilit y

     Air carriers can convert their fleets to stage 3 by replacin g
stage 2 aircraft with new stage 3 aircraft or by retrofitting thei r
stage 2 aircraft with hushkits or new engines . According to
airline officials, 10 years is the minimum lead time needed t o
accomplish such a transition . Although aircraft manufacturer s
presently have a 5-6 year backlog of orders, officials of Boeing ,
McDonnell Douglas, and Airbus (representing 95% of commercia l
aircraft manufactured) told us they believe production rates ar e
capable of meeting a 10-year lead time .


                                   7
     Engine and hushkit manufacturers also told us they believe a
10-year lead time would be adequate since they can increas e
production capacity as demand increases . They also said that th e
necessary retrofit technology either is or will be available .
Technology which has not yet been certified, such as hushkits fo r
the stage 2 DC-9 and B737, is expected to be certified for
production by the FAA within the next few years .

Costs of a Stage 2 Ban by 200 0
Are	 Likely to Be $2 .2 - $4 .6Billio n

      Four organizations have studied the cost of a stage 2 ban .
The four studies were all reported in 1989 and were conducted b y
the FAA ; American Airlines ; AVMARK, Inc . ; and Leeper, Cambridge ,
and Campbell, Inc . (LCC), on behalf of the air cargo industry . 5
The FAA and American Airlines studies reported similar estimates o f
$2 .7 billion and $3 .1 billion, respectively, for the cost o f
meeting stage 3 requirements in the year 2000, assuming a 30-yea r
useful aircraft life . The AVMARK study reported a much highe r
estimate of $59 .6 billion, while the LCC study estimated costs o f
$15 .9 billion for the air cargo fleet alone .

     Different Assum p tions Resul t
     in a Ran e of Cost Estimate s

     The wide range of cost estimates in these studies reflect s
differences in key assumptions concerning the useful life of a n
aircraft, the discount rate used to compare costs and benefit s
occurring in different years, the growth rate of the airlin e
industry, and the extent to which airlines can meet stage 3
standards by using hushkits or new engines rather than by replacin g
aircraft .



5A fifth study, by the International Civil Aviation Organizatio n
(ICAO), did not conduct an original analysis of the U .S . market ;
it reported the results of the FAA study .
                                8
     Two of the studies, by LCC and AVMARK, report very hig h
estimated costs . These high cost estimates result from three ke y
assumptions . First, both LCC and (implicitly) AVMARK assume tha t
aircraft have infinite useful lives, so they would never have to b e
replaced in the absence of a stage 2 ban . Second, AVMARK assumes a
zero discount rate, so that costs occurring in the distant futur e
are weighted just as heavily as costs appearing this year . Third ,
LCC assumes that the growth rate of the air cargo industry will b e
15-20 percent per year . When these assumptions are changed t o
better reflect a consensus of expert opinion on aircraft lif e
spans, discount rates, and growth rates, their estimates change t o
approximately the level of our estimates . We present a detaile d
analysis of the impact of these different assumptions in appendix I
to this statement .

     We have developed our own estimates based on our review of al l
four studies . We used the FAA study as our starting point, but we
used alternative assumptions when we thought they were mor e
appropriate .

     Assuming that stage 3 standards would be achieved by replacin g
non-complying aircraft, we estimated the cost of a stage 2 ban i n
2000 as $4 .6 billion . This is the capital (or present value) cos t
in 1990 of retiring stage 2 aircraft before they would have bee n
retired in the absence of the stage 2 ban . It includes replacin g
2,039 aircraft prematurely at an average cost of prematur e




                                9
retirement of $2 .2 million each . 6 It is based on the FAA model ,
except that three assumptions have been changed :

     -- First, we assumed a real discount rate of 7 .6 percent . Th e
        four studies used discount rates ranging from 0 to 9
        percent ; we concluded that 7 .6 percent was most
        appropriate . 7

     -- Second, we assumed an economically useful life of passenge r
        aircraft of 30 years . This is equal to the 30 years
        assumed by American Airlines, and is the center point o f
        the 25, 30, and 35 year assumptions used by FAA .. For carg o
        aircraft, however, we assumed a longer useful life of 5 0
        years, reflecting the fact that cargo aircraft are used fo r
        fewer cycles per day, and thus can be expected to last fo r
        more years . We believe this is more reasonable than LCC' s
        assumption that cargo aircraft have unlimited economicall y
        useful lives .




6 The cost per airplane of premature replacement is substantiall y
smaller than the price of a new airplane for three reasons : (1 )
these costs are net of operating and maintenance savings whic h
result from using a new airplane ; (2) these are only the costs o f
replacing the aircraft prematurely--they deduct replacement cost s
to the extent that the old aircraft's useful life has been used up ;
and (3) most of these costs will be incurred several years in th e
future, and their present value is less than their value at th e
time that they are incurred . The average cost per aircraf t
replaced varies widely . Some aircraft, such as DC9-50s delivere d
in 1981, would have relatively high replacement costs ($6 .8 3
million each) . However, most of the aircraft that would b e
replaced are older aircraft nearing the end of their useful lives ,
so the cost of premature replacement will be small .
70ur assumed discount rate was the commercial prime rate, plus a
1 .5-percent risk premium, minus the GNP deflator (which measure s
the overall rate of inflation in the entire economy) . We believe
this best measures the real private opportunity cost of money fo r
firms like major airlines .

                                 10
    -- Third, we assumed, using an estimate by the Boein g
       Corporation, that the air cargo portion of the overall U .S .
       civil air fleet will grow at 6 percent per year throug h
       2000 . This is faster than the 1 .9 percent per year growt h
       rate projected by FAA for the industry as a whole, bu t
       slower than the 15-20 percent growth rate for the air carg o
       fleet assumed by LCC . We adopted Boeing's 6-percent
           imate because it seemed reasonable and appeared to b e
        Jased on more substantial analysis than LCC's estimate .
       The faster growth rate for the air cargo fleet, with it s
       longer expected useful life, implies that the costs of a
       stage 2 ban will be increased .

    Alternatives to Aircraft Replacement
    Could Reduce the Cost of a Ba n

      The cost of a stage 2 ban would be significantly reduced i f
alternatives to aircraft replacement were followed . For example ,
new aircraft sell for approximately $25 million to X50 million o r
more . However, existing stage 2 aircraft can meet stage 3
requirements by installing new engines for about $9 million to $1 1
million . In some cases, hushkits can be installed on existin g
engines at a cost of about $1 million to $3 million per aircraft .
According to the FAA, allowing hushkits when available results in a
41-percent savings, on average, versus requiring aircraf t
replacement . The savings are less than the difference in cos t
because new aircraft have lower maintenance and fuel costs than d o
retrofitted aircraft .

     We did our own calculation assuming that airlines would adop t
the cheapest strategy for each type of plane--either replacement ,
re-engining, or hushkitting . This calculation resulted in a
(present value) cost estimate of $2 .2 billion . It include s
replacing 471 aircraft, at an average cost of $0 .3 million, an d


                                11
retrofitting 1,569 aircraft, at an average cost of $1 .2 million . 8
Our analysis suggests that the oldest stage 2 planes are mos t
likely to be replaced, while newer planes are more likely to b e
brought into stage 3 compliance by hushkitting or re-engining . The
exact proportion of aircraft that are retrofitted rather tha n
replaced is uncertain, but we think the actual costs of achievin g
the stage 2 ban are likely to fall between $2 .2 and $4 .6 billion ,
spread out over 10 years . To put these costs in perspective, tota l
industry revenues in 1988 were about $65 billion . Total revenue s
for the 10-year period over which these costs of noise abatemen t
will be borne are likely to exceed $650 billion, so the costs o f
noise abatement are likely to be less than 1 percent of the
industry's revenues .

     A Sta+e 2 Ban b the Year 200 0
     Will Affect Individual Air Carriers Different )


     Although industry officials generally believe a year 200 G
stage 2 ban is achievable, they also believe that the impact of a
stage 2 ban on individual airlines will depend on the size of eac h
carrier's stage 2 fleet and on its financial health . Some airlin e
officials said they are planning to acquire stage 3 fleets befor e
the year 2000 even without a ban ; others believe meeting that dat e
may require them to downsize their operations and could inhibi t
their growth plans .

     The costs of phasing out stage 2 aircraft are likely to b e
borne partly by the airline industry, in the form of lower profits ,
and partly by airline passengers . Since many airline passenger s
are very price-sensitive, the airlines will probably not be able t o
pass on all the costs of the stage 2 ban to their customers an d


8 When retrofitting is an option, average costs of replacement fall ,
because only the oldest aircraft, with the fewest years of usefu l
life (and hence the lowest replacement costs) are replaced . Newe r
aircraft are retrofitted .
                                12
will have to bear a substantial portion of the cost themselves .
Some financially weak airlines may be further weakened by the cost s
of complying with a stage 2 ban .

     Airline passengers are likely to experience both reduce d
service and higher fares as a result of a stage 2 ban . Carrier s
may not find it profitable to replace or retrofit all the old stag e
2 aircraft in service, thus limiting the overall size of thei r
fleets . This is likely to reduce the level of service that can b e
provided and the level of competition on some routes . Some
carriers have stated that they have already reduced service o n
certain routes because they do not have enough stage 3 aircraft t o
meet the noise restrictions at the airports on the route . However ,
because of the relatively low cost of a stage 2 ban, the extent o f
service cutbacks is likely to be small also .

     Fares are likely to rise both because of the costs   of
replacing and retrofitting aircraft, and because of the   reduce d
levels of competition on some routes . However, because   of th e
small relative cost of a stage 2 ban, the size of these   far e
increases is likely to be modest .

POLICY ISSUES FOR CONGRESSIONAL CONSIDERATIO N

     We believe that the key issues that need to be resolved by th e
Congress in formulating a national noise policy are (1) how soo n
stage 2 aircraft should be phased out ; (2) what other actions ar e
needed besides a phase-out of stage 2 aircraft ; and (3) to wha t
extent should federal noise regulation preempt local airpor t
restrictions .

How Soon Should Stage 2 Aircraft be Phased Out ?

     Our interviews with airline and aircraft industry official s
indicate that phasing out stage 2 aircraft by the year 2000 i s
                                13
feasible . Most passenger airlines indicate that they plan t o
phase out stage 2 aircraft voluntarily between 2000 and 2010 .
However, our survey of airports suggests that, in the absence o f
any federal action, a large proportion of the nation's larges t
airports are likely to prohibit the use of stage 2 aircraft by th e
year 2000 . In these circumstances, the airline officials we spok e
with believe that a de facto stage 2 ban will result, becaus e
airlines will not be able to use stage 2 aircraft a~ enoug h
airports to operate such aircraft efficiently .

     Federal policy could delay this phase-out by preempting nois e
restrictions adopted by airports that would effect a year 2000 ban .
The issue for federal policy is whether to adopt the year 2000 as a
reasonable date for phasing out stage 2 aircraft, or whether t o
prevent airports from adopting this or some earlier date . Our
analysis suggests that the cost of adopting this date would amoun t
to about 1 percent of airline industry revenues over the next 1 0
years .

What Other Actions Are Neede d
Besides a Phase-out	 of Stage 2 Aircraft?

     A ban on stage 2 aircraft is an example of noise abatement ,
that is, reducing noise at the source . A variety of othe r
strategies to reduce noise are also available that focus o n
mitigating the impact of noise on the people who are exposed to it ,
for example by soundproofing homes and schools, buying homes tha t
are affected by noise, and improving land-use planning . Nois e
abatement is more cost-effective than noise mitigation when a n
airport is located in an urban area where large numbers of peopl e
are affected by aircraft noise, because the costs of quieting th e
aircraft reduce noise for a large number of people . By contrast ,
the cost of reducing noise exposure by, for example, soundproofin g
homes in a large metropolitan area is so large that this is not a
viable alternative to noise abatement .

                                14
     However, there are limits to a noise abatement strategy .
Aircraft manufacturers do not believe it is possible to mak e
aircraft significantly quieter than the quietest aircraft bein g
built today (though National Aeronautics and Space Administratio n
researchers believe that a long-term research project might develo p
new noise control technologies) . Even these relatively quie t
aircraft cause a high level of noise exposure for people who liv e
close to the airport . FAA estimates that 1 .1 million people wil l
continue to be exposed to excessive noise levels even if stage 2
planes are banned . The only way to reduce aircraft noise t o
reasonable levels close to an airport is through a combination o f
noise abatement and noise mitigation measures . Programs such a s
FAA's Part 150 program, which pays for noise mitigation, wil l
continue to be needed even if a stage 2 ban is adopted .

Should the Federal Governmen t
Preempt Local Airport Restrictions ?

     The federal government has prohibited the use of stage 1
aircraft at U .S . airports, but beyond that has generally left th e
adoption of noise restrictions to individual airports . Airport s
participating in FAA's Part 150 program must have their nois e
control plans approved by the FAA . But airports are not require d
to participate in the Part 150 program, and many do not . FAA may
sue an airport if it regards the airport's noise control program a s
unfairly discriminatory . In a few cases (for example, in a cas e
involving San Francisco's airport) FAA has done so, but generall y
FAA has not contested noise restrictions adopted by the airports .

     Leaving the responsibility for noise regulation to th e
airports has some advantages . Airports vary greatly in ho w
sensitive their neighbors are to noise . One airport may be unde r
great pressure to reduce noise from its neighbors, and 'may be
willing to suffer a substantial reduction in service in order t o

                                 15
reduce noise . Other airports may place a greater value on hig h
service levels than on noise reduction, and may favor a les s
restrictive noise policy . Leaving noise regulation to the airport s
allows each airport to tailor its noise policy to its ow n
individual situation . The benefits of noise reduction are greates t
when the noise reduction is concentrated on airports that are mos t
sensitive to noise .


     On the other hand, our analysis shows that the potential fo r
inefficient use of the nation's air transportation system as a
result of a "patchwork quilt" style of regulation by individua l
airports exists and is likely to become much more significant a s
airport restrictions proliferate after 199_°x . It is much mor e
difficult for the nation's airlines to meet uncoordinated airpor t
requirements--a 50-percent stage 3 requirement at this airport, a
75-percent stage 3 requirement at that airport, and a 100-percen t
stage 3 requirement at yet another--than to meet a nationa l
requirement that a certain percentage of the airline's overal l
fleet in any one year be stage 3 . If an airline meets an overal l
fleet requirement, then it is free to schedule those aircraf t
through its system in a cost-effective and efficient way .


      Establishing a uniform rule, of course, w^uld requir e
preventing airports from adopting rules that were stricter than th e
federal rules . Preemption would prevent airports from tailoring
noise restrictions to each area's sensitivity to noise . I f
airports were preempted from adopting stricter rules, thei r
neighbors would be forced to accept more noise than the airport' s
own rules would have allowed . If the increased noise burden wer e
attributable to federal preemption, plaintiffs might seek to hol d
the federal government liable for damages caused by the extr a
noise .


      The extent to which the federal government would become liabl e
for damages due to aviation noise in the event that it preempte d

                                 16
local restrictions is uncertain . State and local governments, a s
commercial airport proprietors, are responsible for obtaining th e
necessary air easements from surrounding land owners as well a s
fashioning reasonable and nondiscriminatory noise rules for airpor t
operation . Therefore, under present law, states and localities ,
not the federal government, are responsible for injuries resultin g
from airport noise .

     Federal preemption of these local airport noise rules woul d
remove the localities' ability to abate noise and thus the basi s
for liability for injury . As a result, the federal governmen t
could be liable for "takings" and torts arising from airpor t
noise . 9

     Because the federal government has never before been expose d
to liability for regulating commercial airport noise, and becaus e
liability for either a taking or a tort is dependent upon the fact s
in an individual case, we cannot predict whether an injured part y
could successfully sue the federal government for airport noise .
However, we can outline the context within which such a suit migh t
take place .

     In order to show that aircraft flights over private propert y
constitute a taking, the owner must demonstrate that the flight s
are so low and frequent as to cause direct and immediat e
interference with the use and enjoyment of the land . Landowner s
have successfully sued the United States, in its capacity a s
airplane operator, in a number of instances where planes wer e
repeatedly flying at altitudes below 1,000 feet and where the plan e
passed directly over the property .

 The Fifth Amendment to the Constitution prohibits th e
government's taking of private property for public use, that is ,
the direct and immediate interference with the use and enjoyment o f
property, without just compensation . A tort is an injury or wron g
to the person or property of another ; it is the breach of a duty
fixed by law, independent of contract .
                                17
     Under the doctrine of sovereign immunity, the federa l
government is not liable for torts, except where it has waived it s
immunity . The Federal Tort Claims Act (28 U .S .C . Sections 1346 and
2671-2680) waives the federal government's sovereign immunity in a
limited number of cases . With respect to federal regulation o f
airport noise, there are two barriers that a plaintiff would hav e
to overcome before being able to bring a tort suit against th e
federal government . First, no suit may be maintained based o n
discretionary government acts, and second, suits must be based o n
the negligent or wrongful act or omission of a government employee .

     Determinations of whether a particular governmental functio n
is discretionary depend upon the specific facts . Thus it is no t
possible to provide a definite answer on potential federa l
liability . However, as a general matter, the development of nois e
standards that balance local and national interests would appear t o
be discretionary and thus exempt from suit .

      Assuming, however, that FAA's actions were not exempt, a
plaintiff would still have to show that the government employee wa s
negligent . It is not enough that state law provides a remedy i n
tort . Suits based on damages caused by airplane noise cannot b e
brought against the federal government absent a showing that th e
noise was caused by the wrongful act of a government employee .

     The extent of preemption could be limited by "grandfathering "
existing restrictions--that is, allowing airport restriction s
already in place to continue, but preempting any new airpor t
restrictions that were more stringent than the federal rules . Thi s
would ensure that no airport experienced an increase in noise as a
result of the federal stage 2 ban . Grandfathering would, o f
course, reduce the degree of uniformity among airports and the
benefits that uniformity would provide . However, since the numbe r
of airports with existing limitations on stage 2 aircraft i s
                                 18
relatively small, and to date the airlines appear to be managin g
with these "patchwork" costs relatively easily, this may not be to o
costly .

     The Congress might also wish to allow an exemption process b y
which airports that believed their local circumstances justified a
more rapid phasing out of stage 2 aircraft could apply for a n
exemption to allow them to adopt restrictions more stringent tha n
the national rules . This, again, would reduce the degree o f
uniformity in the national system, but would allow some degree o f
variation to reflect local variations in sensitivity to noise .
Such an exemption process would require a weighing of local nois e
concerns against national air commerce objectives, which could b e
difficult to achieve .

CONCLUSION S

     In conclusion, Mr . Chairman, the cumulative effect o f
additional, independently derived, and uncoordinated local airpor t
noise restrictions could create a serious cost burden on th e
nation's air transportation system after 1995 . While the extent o f
these costs is not documented, airlines have stated that th e
current patchwork quilt pattern of local noise restrictions impose s
costs and inefficiencies on the system . Based on our work, thes e
costs appear likely to become much more serious as loca l
restrictions proliferate after 1995 .

     In our view the FAA should make every reasonable effort t o
develop a national noise policy that balances the concerns o f
airports, airlines, local communities, and the nation's ai r
transportation system . A key component of such a national polic y
would be a year-by-year phase-out of stage 2 aircraft, culminatin g
in a ban on stage 2 aircraft by approximately the year 2000 . Our
analysis indicates that the cost of such a ban would be in the $ 2


                                 19
to $5 billion range, depending on whether airlines replace o r
retrofit their existing stage 2 fleets .

     This concludes our statement, Mr . Chairman . We would b e
pleased, at this time, to answer any questions that you or th e
other members of the Subcommittee may have .




                                 20
APPENDIX I                                                   APPENDIX I

                       ANALYSIS OF COST STUDIES

     Four organizations have conducted major studies to determin e
the cost of a stage 2 ban . The studies were all reported in 198 9
and were conducted by the FAA, American Airlines ; AVMARK, Inc ., an d
Leeper, Cambridge, and Campbell (LCC), Inc . l AVMARK is a n
aviation consulting firm whose clients own and operate commercia l
aircraft . LCC is a consulting firm whose study was conducted a t
the request of the Air Freight Association .

     The American Airlines study focused on the nine majo r
airlines ; it did not consider the effect of a ban on smalle r
carriers or the air cargo industry . The FAA and AVMARK studie s
focused on the entire U .S . domestic fleet, both passenger and ai r
cargo . LCC's study focused exclusively on the air freigh t
industry .


DIFFERENT ASSUMPTIONS RESUL T
IN A RANGE OF COST ESTIMATE S


     The FAA, American Airlines, and AVMARK studies each reported a
range of cost estimates . For example, by varying the assumption s
about expected aircraft life and the proposed date of a stage 2
ban, the FAA study estimated the costs under 12 differen t
scenarios . LCC's study, on the other hand, reported a singl e
estimated cost for the air cargo industry .

     Depending on the assumptions used, the estimated costs range d
from a low of $17 million to a high of almost $60 billion . To
illustrate the effect of different study methodologies, we selecte d
a "base case" scenario for comparison purposes, namely, that a

I A fifth study, by ICAO, did not conduct an original analysis o f
the U .S . market ; the study reported the results of the FAA study .
                                  21
APPENDIX I                                                   APPENDIX I


stage 2 aircraft ban will be implemented in the year 2000 . Unde r
this scenario, the cost estimates range from $2 .7 billion (FAA
study) to $59 .6 billion (AVMARK study) . Table I .1 illustrates th e
varying results under the base case scenario as well as under othe r
assumptions .

                   Table 11 : Summary of Studer Result s
                    (in billions of constant dollars )
                                     American
                          FAA        Airlines      AVMARK      LCC

Base case scenario a     $2 .7        $3 .1         $59 .6    $15 . 9

Range of costs :
     High                $5 .8        $3 .1         $59 .9    $15 . 9
     Low                 $0 .017      $0 .53        $22 .5    $15 . 9

a The base case scenario assumes a year 2000 stage 2 aircraft ba n
and a 30-year useful aircraft life .

     Given the common assumptions of the base case scenario, th e
remaining variations in the cost estimates resulted from th e
different cost estimating methodologies employed in the studies, a s
well as from different assumptions regarding the useful life o f
stage 2 aircraft and the expected fleet size in the year 2000 .

STUDIES EMPLOY DIFFEREN T
ESTIMATING METHODOLOGIE S

     The FAA and American Airlines studies each assumed the cos t
associated with a stage 2 ban to be the incremental cost o f
retiring an aircraft . early minus any savings associated wit h
operating new, more efficient, replacement aircraft . For example ,
if an aircraft had to be replaced one year before it would normall y
be replaced, the cost attributable to the ban would be the cost o f
requiring the capital expenditure one year earlier than would hav e
normally occurred, minus the discounted value of savings i n
                                  22
APPENDIX I                                               APPENDIX I


operating and maintenance costs incurred by substituting a ne w
aircraft for an older one . From an economic perspective, this i s
the correct approach to modeling this problem .

     LCC and AVMARK used different approaches . LCC's study state d
that the cost of a ban would be the full capital cost of th e
replacement aircraft minus any operating and maintenance savings .
Although recognizing that aircraft must eventually be replaced, LC C
argued that there are no technical reasons why an aircraft can't b e
maintained for safe use indefinitely and that, consequently, an y
forced retirement should result in the full cost of the replacemen t
aircraft being charged to the ban . AVMARK's study assumed a 30 -
year aircraft life ; however, under its methodology, if an aircraf t
had to be replaced before it was 30 years old--even one yea r
before--then the entire cost of the replacement aircraft wa s
attributed to the ban . Attributing the entire cost of replacemen t
aircraft to the ban resulted in substantially higher cost estimate s
than resulted from the FAA and American Airlines methodologies .

     The methodologies employed by LCC and AVMARK both, in essence ,
assume that used aircraft have an indefinitely long economic lif e
and therefore do not depreciate in value . They implicitly assum e
that old aircraft would never be replaced in the absence of a
government intervention requiring their replacement . Consequently ,
their studies argue that the full cost of replacement aircraf t
should appropriately be charged to the event--the ban--which led t o
the need for replacement . This assumption is not consistent wit h
the fact that old aircraft are constantly being replaced even i n
the absence of any government requirement, primarily because, a s
aircraft age, the costs of maintenance and repair become greate r
than the costs of buying a new airplane .



                                23
APPENDIX I                                                      APPENDIX I

     In our analysis, we followed the methodology of FAA an d
American Airlines, and charged costs of replacement to the stage 2
ban only to the extent that they caused aircraft to be replace d
prior to the expiration of their useful lives .

      In their studies, the FAA, American Airlines, and LC C
discounted future expenditures in order to express them as curren t
dollars . This is a common practice in analyzing expenditures ove r
time . The FAA, American Airlines, and LCC studies assumed discoun t
rates of 7 percent, 9 percent, and 6 .2 percent, respectively .
AVMARK did not discount future costs {i .e ., they assumed a zer o
discount rate) . These assumptions affect costs since a discoun t
rate which is too low tends to overstate both costs and benefit s
 (the benefits include reduced operating and maintenance costs) o f
replacement, while one which is too high understates costs an d
benefits . We assumed that the real discount rate would equal 7 . 6
percent, which is the prime lending rate plus a 1 .5 percent ris k
premium minus inflation as measured by the GNP deflator . We use d
that rate to recalculate the present value of the costs reported i n
each study . Table I .2 illustrates the results .

             Table I 2 : Costs Ad j usted to Present Value Term s
                          (In Billions of Dollars )
                                      American
                          FAA         Airlines       AVMARK         LC C
Reported Costs            $2 .7        $3 .1         $59 .6         $15 . 9
Adjusted Costs             2 .7         3 .3          43 .8          15 . 0

Difference                $0 .0       ($0 .2)        $15 .8         $ 0.9


AIRCRAFT USEFUL LIFE
ASSUMPTIONS AFFECT COST ESTIMATE S

     Assumptions about aircraft useful life are also important i n
estimating the cost of a stage 2 aircraft ban . The useful life o f
                                     24
APPENDIX I                                               APPENDIX I



an aircraft depends on how intensively it is used . A cargo
aircraft, for example, which is flown fewer cycles (one takeoff an d
landing) per day than a passenger aircraft, can be expected to b e
economically useful for more years . This is a significant variabl e
because the number of useful years removed from an aircraft's lif e
by a stage 2 ban is a major determinant in the total cost of such a
ban .

      The American Airlines study assumed a 30-year useful lif e
while the FAA study assumed a 25-35 year useful life . The LCC
study, as noted previously, argued that aircraft can be maintaine d
for safe use indefinitely and therefore did not assign a usefu l
life . The AVMARK study, while assigning a 30-year useful life ,
agreed with the LCC study that aircraft can be maintaine d
indefinitely .

     While the technical life of an aircraft is, as LCC and AVMAR K
suggest, indefinite, the economically useful life is definitel y
limited . The economic life extends only to the point when i t
becomes cheaper to replace the aircraft with a newer aircraf t
rather than to make repairs . The economically useful life i s
likely to become shorter as a result of new and planned FAA Ai r
Worthiness Directives, which will require ever more frequent an d
costly maintenance procedures for older aircraft . Neither LCC no r
AVMARK explicitly considered these additional costs in thei r

studies .

     In our analysis, we adopted the assumption of a 30-year usefu l
life for passenger aircraft, but assumed that cargo aircraft woul d
have a longer useful life of 50 years . This appeared to be th e
longest useful life that is consistent with actual practice .



                                25
APPENDIX I                                               APPENDIX I


EXPECTED AIRCRAFT FLEET SIZ E
CRITICAL TO ESTIMATED COST S

     A critical assumption in the LCC study is that the air freigh t
industry will grow substantially in the next 10 years . LCC
assumed that the fleet of 254 stage 2 aircraft owned by the ai r
freight industry in 1987 would increase 20 percent annually throug h
1992, and thereafter 15 percent annually through the year 2000 ,
resulting in a fleet of 1,933 stage 3 aircraft at that time . This
assumption of rapid growth adds appreciably to the cost of a ban a s
calculated in the LCC study, particularly since LCC charged th e
full, undiscounted cost of replacement aircraft to the ban .

     Several other studies have suggested that the rate of growth
in this industry will be substantially less . The FAA estimate s
that the entire U .S . fleet will grow at the rate of 1 .9 percent
annually . The American Airlines study suggests that fleet growt h
will be only "a few" percent annually, and aircraft manufacturer s
forecast growth in the air cargo fleet at less than 6 percen t
annually .

     We analyzed the impact of LCC's growth assumptions b y
substituting different growth rates . Using a growth rate of 1 . 9
percent annually would result in a fleet of 318 aircraft instead o f
the 1,933 calculated by LCC . The adjusted cost estimate, i n
present value terms, would be $1 .8 billion instead of $10 . 9
billion . A growth rate of 6 percent would result in 511 planes an d
an adjusted cost of $2 .8 billion . In our analysis of costs for the
overall civil transport fleet, we assumed that the air cargo flee t
would grow by the 6 percent per year estimated by Boeing . We
assumed that 1 .9 percentage points of this growth would com e
through purchases of new stage 3 aircraft, as FAA assumed . We


                                26
APPENDIX I                                                 APPENDIX I


assumed that the remaining 4 .1 percentage points would come throug h
purchases of used stage 2 aircraft .

     We adjusted the estimates in the four studies to reflect wha t
seemed to be the most reasonable assumptions, namely :

     -- incremental cost of early retirement (rather than ful l
        replacement cost) ;

     -- a 7 .6 percent discount rate ;

     -- a 30-year useful life for passenger aircraft, but a 50-yea r
        useful life for cargo aircraft ;

     -- a 6-percent growth rate for the air cargo fle,c t, includin g
        1 .9 percent growth supplied by purchases of new stage 3
        aircraft, and 4 .1 percent supplied by purchases of use d
        stage 2 aircraft .

     These assumptions resulted in a cost estimate for replacin g
all stage 2 aircraft by the year 2000 of $4 .6 billion . This is the
present value in 1990 of the costs of premature replacement throug h
the year 2000 .

ALTERNATIVES TO AIRCRAFT REPLACEMENT
COULD REDUCE THE COST OF A BA N

      The cost of a stage 2 ban would be significantly reduced i f
alternatives to aircraft replacement were followed . For example ,
new aircraft sell for approximately $25 million to $50 million o r
more . However, existing stage 2 aircraft can meet stage 3
requirements by replacing the engines for about $9 million to $1 1
million . In some cases, hushkits can be installed on existin g

                                  27
APPENDIX I                                                 APPENDIX I


engines at a cost of about $1 million to $3 million per aircraft .
According to the FAA, allowing hushkits when available results in a
41 percent savings, on average, versus requiring aircraf t
replacement .

     We adjusted our estimate by assuming that any particular typ e
and cohort of aircraft (e .g ., B727s built in 1975) would be eithe r
replaced, re-engined, or hushkitted, depending on whic h
alternative was least expensive, taking into account both th e
initial capital cost and the fuel and maintenance savings tha t
replacement and (to some extent) re-engining provide . Thi s
analysis resulted in a present value cost estimate for a year 200 0
phase-out of stage 2 aircraft of $2 .2 billion .
I