FTC Staff Comment Before Massport Concerning Program for Airport Capacity Efficiency

COMMISSION 	AUTHORIZED
V8'T003o
Massport's Proeram for Airport Capacity Efficiency
Comments of the Bureaus of Economics, Competition, and Consumer
Protection of the Federal Trade Commission 1
Washington, D.C.
February 29, 1988
1. Introduction and Summary
A. Interest and Experience of the Federal Trade Commission Staff
The staff of the Federal Trade Commission ("FTC staff") is pleased to
respond 	to 	Massport's 	in\llitation 	to 	comment 	on 	its 	Proposal 	for 	Airport
Capacity Efficiency ("PACE"). 	The FTC staff has a longstanding interest in
aviation 	issues, 	including 	the 	pricing 	of 	landings 	at 	airports. 	FTC 	staff
research 	has 	led 	to 	the 	publication 	of 	a 	report 	on 	slot 	allocation, 	and 	a
report on airline deregulation. 2 	In addition, FTC staff has participated in
several 	administrative 	proceedings 	involving 	airport 	access. 3 	These
1 These comments represent the views of the Federal Trade Commission
staff, and 	do 	not 	necessarily represent 	the views of 	the Commission or of
any 	individual 	Commissioner. 	The 	Commission 	has, 	however, 	voted 	to
authorize the Bureaus to submit these comments.
2 D. Koran and J. Ogur, Airport Access Problems: Lessons Learned from
Slot Regulation by the FAA, Staff Report to the Federal Trade Commission,
May 	1983; and 	J. 	Ogur, C. 	Wagner, 	and 	M 	Vita, 	The 	Deregulated 	Airline
Industrv: 	A 	Review 	of 	the 	Evidence, 	Staff 	Report 	to 	the 	Federal-=rra-de
Commission, January 1988.
3 For example, the FTC staff filed comments in Slots Transfer Methods,
Before 	the 	Federal 	Aviation 	Administration 	("FAA"), 	Docket 	No. 	24105,
August 	1984; 	Elimination 	of 	Airport 	Delavs, 	Before 	the 	FAA, 	Docket 	No.
24206, 	1984; 	Slot 	Allocation 	Alternative 	Methods. 	Before 	the 	FAA, 	Docket
No. 	24110, 	1984; 	Discussion 	Authoritv 	for 	Agreement 	to 	Shift 	Schedules.
Before 	the 	U.S. Department of Transportation, Docket 	No. 44634, 	February
23, 	1987; 	and 	Charges 	for" 	the 	Use 	of 	M~troDolitan Washington 	Airports,
----.-- -

proceedings 	raised 	issues 	that 	are 	similar 	to 	the 	issues 	addressed 	by
Massport in PACE.
Although 	the 	benefits 	of 	airline 	deregulation 	are 	substantial, 	these
benefits 	c:ould 	be even 	larger 	if 	airports 	priced 	landings 	more 	efficiently.
Research shows that deregulation has provided billions of dollars of benefits
to society.( 	However, the price of a landing does not reflect its full costs,
leading to excessive delays and reduced consumer benefits. s 	Current pricing
of landings is inefficient because it wastes resources, such as extra fuel used
by 	aircraft 	waiting 	to 	take 	off 	or 	land, 	and 	wastes 	the 	time 	of 	delayed
passen gers. 	Efficient 	pricing 	of 	landings 	would 	bring 	the 	benefits 	of
deregulation closer to their potential.
We 	believe 	that 	P~CE 	would 	be 	a 	significant 	step 	toward 	more
efficient pricing of airport services at Logan International Airport (Logan).
PACE 	recognizes 	:hat 	the 	cost 	of 	providing 	a 	landing 	varies 	with 	factors
other 	than 	aircraft 	weight. 	In 	particular, 	the 	proposal 	recognizes 	that 	a
landing during the congested morning or afternoon peak period may impose
greater 	social 	costs 	than 	does 	a 	landing 	during 	an 	uncongested 	off-peak
hour. 6 	Hence, PACE contains proposals to make a portion of the landing fee
Before the FAA, Docket No. 25204, April 13, 1987.
(The 	benefits, 	which 	have 	been 	estimated 	at 	$15 	billion 	per 	year,
result 	primarily 	because carriers 	are 	now 	able 	to set 	prices 	and 	schedules
that 	reflect 	what 	consumers 	want 	and 	the 	cost 	of 	providing 	it 	(see 	S.
Morrison and C. Winston, The Economics of Airline Deregulation, Brookings,
1986; and Ogur, Wagner, and Vita, 1988, Section II).
5 See 	Morrison 	and 	Winston 	( 1986) 	for 	a 	discussion 	of 	these
inefficiencies:
6 The social costs 	of a 	landing are the costs 	to society, 	including 	the
costs 	to 	the 	airport, 	to 	aircraft 	operators, 	to 	passengers, 	and 	to 	residents
living near the airport.
2

independent of 	weight. 	Some of 	those proposals 	would 	make 	that 	portion
variable by time of day.
In 	the Appendix 	to these comments, we use Massport's assumptions 	to
derive sample estimates of 	the benefits and costs of 	the agency's 	proposed
pricing changes. 	Based on Massport's assumptions, making a portion of the
landing fee independent of weight would yield benefits that are greater than
costs.
We also offer suggestions that may further increase the net benefits of
PACE. 	In 	particular, we suggest 	that Massport consider adjustments to the
PACE landing fees so that they more closely approximate the added cost of
an 	additional 	landing. 7 	Such 	landing 	fees 	will 	generate 	the 	greatest 	net
social benefits.
Finally, we offer for discussion a system of priority and 	postponable
service as an alternative to restricting the number of landings below airport
capacity. 	This approach would avoid restricting flights in good weather and,
unless administrative costs were prohibitive, would increase net benefits.
B. The Program
In PACE, Massport proposes to change the basis on which it calculates
landing fees 	at Logan. 8 	Current fees, 	which are based on aircraft weight,
are $1.31/thousand pounds (1,000 Ibs.), with a minimum of $25 per landing.
To generate the same revenues, Massport offers several alternative pricing
7 In economic terms, this is the marginal cost of a landing. 	As we will
discuss 	below, 	marginal 	cost 	includes 	the 	resource 	costs 	imposed 	on 	the
airport, 	the delay costs imposed on 	passengers, and the 	noise costs 	imposed
on nearby residents.
8 There is no takeoff fee at Logan, nor is one proposed in PACE.
3

schemes. 	Under 	one, 	Massport 	would 	charge 	a 	two-part 	price 	of 	$91 	+
$0.45/1,000 Ibs. 	(PACE, pp. 22-23). 	This proposal would make a portion of
the landing fee ($91) fixed for all types of aircraft, and thus independent of
aircraft weight. 	The fixed portion of the landing fee would be the same for
all hours of the day. 	In the Appendix, we present sample estimates of the
benefits and costs of this proposal.
Massport 	also 	presents 	several 	alternative 	proposals. 	One 	alternative
pricing scheme would charge $91 + $0.72/1,000 Ibs. during peak periods, and
$50 + $0.72/1,000 Ibs. during non-peak 	periods (PACE, pp. 27-28). 	Another
alternative scheme would charge $300 + $0.45/1,000 Ibs. during peak periods,
and only $0.45/1,000 Ibs. during 	non-peak periods (PACE, 	p. 31). 	In 	these
two proposals, the portion of the landing fee that is independent of aircraft
weight 	would 	vary 	by 	time 	of 	day, 	being 	greater during 	peak 	hours 	than
during 	off-peak 	hours. 	In 	addition, 	Massport 	proposes 	an 	alternative
.allocation scheme 	that 	would 	give large aircraft priority for 	85 percent of
the operations (takeoffs 	and 	landings) during 	peak 	periods. 	Rights 	to the
remaining 	operations 	would 	be 	auctioned 	to 	small 	aircraft 	and 	essential
service 	(PACE, 	Exhibit 	M). 	Finally, 	Massport 	proposes 	to 	assess 	general-
aviation 	operations 	a 	fee 	of 	$65 	for 	each 	use 	of 	the 	general-aviation
terminal at Logan (PACE, p. 21). 	Where a vaila ble inf orma tion permits, we
will 	draw qualitative conclusions 	regarding 	the benefits and 	costs of these
alterna ti ve proposals.
-C._Estimated Benefits and Costs
In 	the 	Appendix, 	we 	provide sample calculations 	of 	benefit 	and 	cost
estimates. 	Our 	calculations 	are 	based 	in 	large 	part 	on 	the 	data 	and
assumptions provided 	in PACE. 	As a result, our calculations are dependent
4

on 	the 	accuracy 	of 	that 	information. 	Where 	we 	have 	made 	additional
assumptions ourselves, we have chosen conservative assumptions in order to
understate 	the 	net 	benefits 	of 	the 	proposed 	pricing 	change. 	We 	should
emphasize, 	however, 	tha"t\these 	figures 	are 	simpl y 	sample 	estimates 	dra wn
from the available information.
D. Priority and Postponable Service
As we understand 	the PACE proposals, they would restrict the number
of 	landings on days with good weather, even though it appears that 	Logan
can 	provide 	the 	current 	level 	of 	landings 	without 	excessive 	delays.9 	We
offer for discussion a system that would restrict landings only on days with
bad weather. lO
II. Landing Fees and Economic Efficiency
Based 	on 	Massport's 	assumptions, we estimate 	that 	changing 	from 	the
current fee of $1.31 per thousand pounds of aircraft weight to the proposed
fee 	of 	$91 	per 	landing 	plus 	$0.45 	per 	thousand 	pounds 	of 	weight 	would
increase 	net 	social 	benefits. 	We 	also 	estimate 	that 	offering 	priority 	and
postponable service would further increase net social benefits.
9 That is, when visual-flight-rule (YFR) condi-t-ions prevail.
10 That 	is, 	when 	instrument-flight-rule 	(IFR-I 	or· worse) 	conditions
prevail. 	We recognize that the FAA and other airports would have important
roles to play in any system of priority and postponable service. 	In principle,
a 	scheme 	to 	offer 	such 	service 	could 	also 	apply 	to 	days 	when 	wind
conditions reduce Logan's capacity below the level of demand.
5

Net social benefits would be maximized by a landing fee equal to the
added 	cost 	of 	providing 	an 	additional 	landing 	in 	each 	time 	period. l1
Although part of the added cost varies with aircraft weight, another part of
that 	cost 	is 	incurred 	because 	the 	landing 	takes 	place, 	preventing 	other
aircraft 	from 	using 	the 	runway at 	that 	time. 	The 	part 	of 	the 	added 	cost
that is unrelated to weight varies with the amount of congestion at different
times of the day. 	Hence, economically efficient landing fees vary with the
time 	of 	day 	in 	addition 	to 	varying 	with 	aircraft 	weight, 	unless 	any
administrative costs 	incurred 	in 	implementing 	the time-sensitive fees 	would
exceed the benefits obtained. 12
A. The Added Cost of A Landing	,
The added cost of a landing includes: I) the resource costs imposed on
the airport, 2) the .delay costs imposed on aircraft operators and passengers,
including 	the 	delay 	costs 	imposed 	on 	other 	flights, 	and 	3) 	the 	noise 	costs
imposed on residents living near the airport. 	Economically efficient landing
fees cover all of these costs.
11 For a discussion of the benefits of setting price equal to the added
cost 	of 	service 	(marginal-cost 	pricing) 	see 	E. 	Mansfield, 	Economics. 	Fifth
Edition, 1986, Chap. 25.
12 Massport's 	current 	use 	of 	aircraft 	weight 	as 	the 	sole 	basis 	for
landing fees is similar to the practice in airports around the world (see A.
Walters, "Airports -- An 	Economic Survey," Journal of Transport Economics
and 	Policy, 	May 	1978, 	PP. 	125-160). 	Some 	deviation 	from 	marginal-cost
pricing may be justified at airports that have excess capacity and that must
cover their costs from revenues (see S. Morrison, "The Structure of Landing
Fees at 	Uncongested 	Airports," Journal 	of Transport 	Economics and 	Policv,
May 	1982, 	pp. 	151-159). 	According 	to 	Massport, excess 	capacity existed 	at
Logan for a period following deregulation (PACE, p. 14), but no longer does
during 	peak 	periods 	under 	IFR-l 	or 	worse conditions. 	Moreover, Morrison
(1982) shows that, even if excess capacity still existed at Logan, landing fees
based solely on weight would be inefficient.
6

1) Airport Resource Costs
Massport incurs added resource costs to provide a landing during both
peak 	and 	off-peak 	periods. 	These 	costs 	are 	generally 	broken 	down 	into
costs 	of 	maintenance, 	operations, and 	administration. 13 	Aircraft 	weight 	is
significantly 	related 	to 	such 	costs,14 	and 	so 	efficient 	landing 	fees 	should
vary with aircraft weight.
2) Delay Costs
Under Massport's current landing fees, operators of and 	passengers on
peak-period 	flights 	incur significant costs 	in 	the 	form 	of 	lost 	time 	due 	to
delays.lS 	Delay 	costs 	are 	imposed 	because 	one 	aircraft 	on 	a 	runway
precludes 	runway 	use 	by 	other aircraft 	at 	the 	same 	time. 16 	All 	types 	of
13 S. Morrison, "Estim<Jtion of 	Long-Run 	Prices and 	Investment 	Levels
for Airport Runways," Research in Transportation Economics, 1983, pp. 103-
130.
14 The 	added' cost 	of 	maintenance, 	operations, 	and 	administration 	is
bigger 	for 	heavy' aircraft 	than 	for 	light 	aircraft. 	For 	the 	mid-1970s,
Morrison 	(1983) 	estimated 	that 	the 	typical 	airport's 	added 	cost 	of
maintenance, 	operations, 	and 	administration 	was 	approximately $12 	for 	an
air-carrier operation, but not significantly greater than zero 	for 	a 	general-
aviation 	operation. 	The 	author 	cited 	a 	study 	of 	British 	airports 	that
obtained similar results.
IS Off-peak 	operations 	incur 	delay 	costs 	that 	are 	significantly 	lower
than 	the 	costs 	incurred 	by 	peak-period 	opera tions. 	Dela y 	costs 	ha ve 	been
estimated 	by 	A. 	Carlin 	and 	R. 	Park, 	"Marginal 	Cost 	Pricing 	of 	Airport
Runway 	Capacity," 	American 	Economic 	Review, 	June 	1970, 	pp. 	310-319; J.
Likens, 	"The 	Welfare Costs 	of 	Nonoptimal 	Airport 	Utilization," 	Journal 	of
Public Economics, 1976, pp. 81-102; S. Morrison (1983); and J. Yance, "Airline
Demand for Use of An Airport and Airport Rents," Transportation Research,
December 	1971, 	pp. 	267-281. 	Using 	the 	delay 	estimates 	of 	Flight
Transportation 	Associates 	(PACE, 	p. 	15) 	and 	estimates 	of 	the 	value 	of
passengers' 	time, 	it 	may 	be 	possible 	to 	obtain 	estimates 	of 	marginal 	delay
cost and -the optimal congestion charge. 	Marginal delay cost is the cost of
added delay imposed on passengers by an added flight.
16 The 	current fee 	structure at Logan 	does 	not 	take 	into 	account 	the
delay 	costs 	that a 	peak-period aircraft operation 	imposes on 	other 	aircraft
operators and their passengers. 	Hence, too many flights are offered during
peak periods, and delays are too long.
7
r

aircraft impose delay costs w~en they land during a peak period. 	Although
these costs may be related 	to aircraft weight, current weight-based landing
fees overstate that relationship, charging large aircraft too much and small
aircraft too little. IT 	Efficient landing fees include the delay costs that each
aircraft imposes, a portion of which is unrelated to aircraft weight. IS
Charging 	higher 	peak-period 	landing 	fees 	will 	cause 	some 	aircraft
operators 	to 	shift 	peak-period 	flights 	to 	other 	times 	or 	other 	airports,
thereby reducing the length of delays experienced by passengers and aircraft
operators who continue to land during peak periods. 	Those passengers and
operators 	who 	are 	willing 	to 	pay 	the 	higher 	fees 	are 	the 	ones 	to 	whom a
landing at such times 	is worth the most. 	Also, charging lower fees during
off-peak periods will benefit those operators and passengers who are flexible
I
enough to shift landings to off-peak periods.
As 	noted 	above, 	two 	of 	Massport's 	alternative 	proposals 	would 	vary
landing fees by time of day. 	Such variation attempts to reflect the higher
IT Walters 	(1978, 	p. 	133) 	notes 	that 	larger 	aircraft 	require 	greater
spacing because of the increased turbulence that they generate. 	The author
also 	notes, 	however, 	that 	existing 	weight-based 	landing 	fees 	overstate 	the
resulting added costs of handling such aircraft.
18 Landing 	fees 	that 	attempt 	to 	recover 	the 	historical 	costs 	of 	the
airport 	can 	lead 	to 	distortions 	that 	increase 	delays. 	First, 	the 	historical
costs of older airport facilities tend to be lower because of depreciation and
because 	price 	levels 	were 	lower 	when 	those 	facilities 	were 	constructed.
Thus, landing fees 	based on historical costs tend to be lower than efficient
fees. 	Second, for any given level of historical costs, congested airports have
more 	landings 	across 	which 	to 	spread 	those 	costs 	than 	do 	uncongested
airports. 	Hence, landing fees based on historical costs tend to be lower at
congested 	airports 	than 	at 	uncongested 	airports 	(Walters, 	1978, 	p. 	134).
Finally, some airports, regardless of how busy, underprice 	landings 	because
historical costs are covered 	primarily by 	revenue from 	concessions, such as
restaurants and car-rental agencies 	(Office of Technology Assessment, 1984,
p. 131).
8

delay costs of providing a 	peak-period landing 	than an 	off-peak landing. 19
Like other services offered at different times of the day with different costs
to 	consumers, 	peak 	and 	off-peak 	landings 	are 	different 	services, 	and
economic efficiency would require that consumers be charged different prices
for such services. 2o
3) Noise Costs to Nearby Residents
Aircraft 	noise 	imposed 	on 	residents 	located 	near 	an 	airport 	such 	as
Logan 	makes 	their 	houses 	less 	desirable. 	Massport 	is 	committed 	to 	keep
cumulative 	noise 	impact 	at 	or 	below 	1984 	levels 	(PACE, 	p. 	37). 	To
implement 	this 	commitment 	efficiently, 	Massport 	could 	vary 	landing 	fees
based on the noisiness of aircraft. 	Thus, noisy aircraft would pay a higher
noise charge than would quiet aircraft. 	The charges, which would encourage
the substitution of quieter 'for noisier aircraft, could be set so that Massport
achieved its target of 1984 noise levels. 21
19 Likens' (1976) results indicate that takeoffs also impose delay costs
that 	vary 	by 	time 	of 	day. 	This 	result 	suggests 	that 	Massport 	consider
charging 	takeoff 	fees 	that 	vary 	by 	time 	of 	day, 	provided 	that 	the 	added
administrative costs do not outweigh the added benefits.
20 	A 	peak-period 	landing is 	higher cost 	than an 	off-peak 	landing 	in
the same way that that an overnight stay in a hotel room is higher cost on
a business day than on a weekend, and that a telephone call during business
hours is higher cost than a call at night.
To 	reflect 	higher 	congestion 	costs, 	hotels 	charge 	higher 	prices 	for
rooms on business days 	than on weekends; and 	telephone companies charge
higher 	rates 	for 	calls 	during 	business 	hours 	than 	for 	nighttime 	calls.
Similarly, efficiency requires that airports charge more for 	landings during
congested peak periods than durin.g_Iess-congested off-peak periods.
21 We make no judgment regarding the efficiency of these levels. 	The
calculation 	of 	efficient 	noise charges 	is discussed 	in 	Walters 	(1978) and 	S.
Borins, "Pricing and 	Investment in 	A Transportation 	Network: The Case of
Toronto Airport," Canadian Journal of Economics, 1978, pp. 680·700.
9
.. ~

B. Implementation of Efficiency Approach
Efficient 	landing 	fees 	should 	reflect 	all 	the 	costs 	of 	providing 	a
landing, 	including costs imposed 	on 	the airport, on 	passengers and aircraft
operators, and on nearby residents. 	Efficient peak-period landing fees would
be 	higher 	than 	efficient 	off-peak 	fees 	primarily 	because 	the 	added 	delay
costs to passengers and aircraft operators are higher during peak periods.
Some airports have taken steps toward efficient landing fees. 	London's
Heathrow Airport has imposed peak-hour surcharges on all aircraft. 	Airports
in 	the 	New 	York 	area 	have 	imposed 	such 	surcharges 	on 	general 	aviation.
Massport imposed a minimum daily use fee for all aircraft, which was later
discounted 	during 	off-peak 	periods. 	More 	recently, 	Massport 	discontinued
these fees (PACE, pp. 25-26). 	Although the impact of these surcharges has
.not been 	rigorously estimated, the surcharges appear to have shifted flights
to 	off-peak 	hours. 22 	Such 	shifting 	may 	help 	a void 	the 	need 	to 	expand
airport 	capacity, 'at great 	expense, 	which 	might 	then 	be 	idle 	much 	of 	the
day.
Efficient 	landing 	fees 	have 	been 	estimated 	for 	several 	airports,
including 	Logan.23 	These 	estimates 	suggest 	that 	the 	peak-hour 	surcharges
that 	have 	been imposed 	were smaller than efficient pricing 	would 	require.
Nevertheless, 	the success of 	peak-hour surcharges indicates 	that Logan 	and
other airports can proceed further toward efficient pricing, as have firms in
22 See Office of Technology 	Assessment, 	Airport Svstem 	Development,
August 	1984, pp. 	118-119 and 	131-132); and 	PACE, p. 25-26. 	Al though tM - -
number 	of 	peak-period 	flights 	increase<r at- Heathrow 	two 	years 	after 	the
peak-hour surcharges were imposed, this may simply reflect demand growth,
which would increase the efficient surcharge.
23 See, 	for 	example, 	Borins 	(1978) 	and 	S. 	Morrison, 	"Optimal 	Pricing
and Investment Policies for Airport Landing Areas," Working Paper No. SL·
7907, University of California, Berkeley, November 1979.
10

other 	industries, 	such 	as 	telecommunications, 	natural 	gas 	transmission 	and
distribution, and electrical distribution.H
Efficient landing fees can also provide much needed information on the
timing and location of airport expansion. 	As noted above, efficient landing
fees 	include delay costs. 	The need for airport expansion can be judged by
comparing the delay costs with existing capacity to the costs of building new
capacity. 	Expansion should occur 	when 	the delay costs over the 	remaining
life of existing capacity are expected 	to exceed 	the added cost of building
new 	capacity. 	In 	that 	case, 	passengers and 	aircraft 	operators 	would 	save
more 	in 	delay 	costs 	than 	the 	cost 	to 	society 	of 	the 	added 	capacity. 	By
contrast, 	expansion 	would 	not 	be 	justified 	when 	the 	added 	capacity 	cost
exceeds the expected delay costS. 25
.
III. Benefit and Cost Estima tes
To estimate the net 	benefits of Massport's 	proposed 	pricing change 	to
591 	+ $0.45/1,000 Ibs., Massport should consider the 	approach 	we detail 	in
the Appendix. 	Based on data provided in PACE by Massport, the proposed
pricing 	change 	could 	save 	nearly 	70,000 	passengers 	more 	than 	2,500 	hours
2,( Efficient 	fees 	are 	unlikely 	to 	shift 	the 	peak 	to 	a 	different 	time
period because of the strong preference of business travelers for peak-period
flights'.
25 The costs of added capacity might be met from the revenues raised
by 	the landing 	fees, 	which 	in 	turn 	might be 	reflected 	in 	higher 	fares 	for
peak-period 	flights. 	It 	should 	be 	noted 	that 	PACE 	does 	not 	address 	the
financing of added capacity.
Logan now has four 	major runways (PACE, p. 	12). 	However, we lack
information on the current optimal number of runways at the airport. 	For
the 	mid-1970's, 	Morrison 	(1983) 	estimated 	that 	the 	optimal 	number 	of
runways 	at 	Logan 	was 	between 	three 	and 	five, 	depending 	on 	the 	interest
rate and the value of passengers' time.
I I
••.;0: '.'~..,.' -.;::

per day. 	To achieve these savings. approximately 67 general-aviation flights
would be eliminated by the increase in landing fees.
Passengers 	on 	flights 	that 	continue 	to 	land 	at 	Logan 	after 	the 	price
change 	receive 	benefits 	In 	the 	form 	of 	reduced 	delay 	time. 	Estimates 	of
benefits should 	include 	this saving only 	when 	instrument-flight-rule (IFR-}
or 	worse) conditions 	prevai1.26 	The value of 	the savings should 	equal 	the
number of passengers on the flights that continue to land, multiplied by the
amount of time saved per passenger and by a conservative estimate of those
passengers' value of time.
Cost 	estimates should 	consist 	of 	the 	losses 	to 	passengers 	on 	cancelled
general-aviation 	flights. 	flights 	whose 	operators 	are 	unwilling 	to 	pay 	the
increased 	landing 	fees. 21 	These 	are 	the 	losses 	stemming 	from 	cancelling
these 	landings; 	that 	is. 	the 	amount 	that 	the 	operators 	would 	be 	willing 	to
pay for the landin.ss. 	The smallest amount that an operator is willing to pay
is the original landing fee. 	The biggest amount that an operator is willing
to 	pay 	is 	slightly 	less 	than 	the 	new 	landing 	fee. 	The 	total 	value 	of 	the
26 We lack information on the extent of delays under VFR conditions.
If 	delays 	are 	also 	reduced 	under 	these 	conditions. 	then 	the 	estimated
benefits 	may 	be 	increased. 	However, 	it 	should 	be 	noted 	that 	total
elimination 	of 	delays 	would 	not 	be 	efficient. 	To 	do 	so 	would 	eliminate
landings that are worth more than the added time sa ved.
21 This 	estimate 	is 	a 	substantial 	overstatement 	of 	the 	loss 	given 	the
information 	provided 	by 	Massport. 	It 	should 	be 	noted. 	however. 	that
Massport's 	estimates 	of 	the 	number 	of 	flights 	eliminated. 	and 	thus 	our
estimates of the loss. do not take into account any effects of the proposed
$65 	fee 	for 	use 	of 	the 	general-aviation 	terminal. 	This 	fee 	is 	likely 	to
eliminate 	more 	general-aviation 	flights 	and 	thus 	to 	increase 	both 	the 	loss
and 	the 	benefits because delays 	would 	be 	further 	reduced. 	With available
information, however, we are unable to estimate the net effect of this fee.
Given 	that 	Massport 	does 	nOt 	propose 	to 	increase 	total 	revenues
generated 	by 	landing 	fees, 	our 	cost 	estimates 	equal 	the 	willingness 	of
aircraft 	operators 	and 	passengers 	to 	pay 	for 	the 	foregone 	landings, 	minus
what 	they actually 	pay, 	plus the 	increased 	landing 	fees 	that 	the remaining
flights pay.
12

eliminated flights consists of the average of these two values, multiplied by
the number of flights eliminated.
IV. Landing Fees that Vary By Time of nay
As an alternative to the landing fees considered above, which would be
uniform 	across 	the 	hours 	of 	the 	day. 	Massport 	proposes 	to 	charge 	$91 	+
$0.72/1,000 	Ibs. during 	peak 	periods and $50 + $0.72/1,000 Ibs. 	during off-
peak periods. 	This alternative landing-fee schedule would raise the price of
an off-peak landing less than the price of a peak-period landing. 	Although
we 	would 	need 	additional 	information 	to estimate 	the 	net 	benefits of 	this
pricing schedule, it may offer even greater net benefits than a uniform price	,
during 	every 	hour 	of 	the 	day.28 	By eliminating 	fewer 	off-peak 	landings,
time-variable pricing would impose smaller costs than would the proposal to
.
charge 	$91 	+ $0.45/1,000 	Ibs. 	at 	all 	hours 	of 	the 	day.29 	The 	alternative
proposal 	would 	also 	increase the likelihood 	that, 	in response to the 	pricing
change, some general-a via tion opera tors would shift their 	flights from 	peak
to 	off-peak 	times 	at 	Logan 	instead 	of 	cancelling 	their 	flights 	or. shifting
them to another airport. so
28 In order to be most effective, peak hour prICIng would also need to
consider the 	time of 	takeoff. 	It 	is apparent 	that some of 	the 	benefits of
peak hour pricing would be lost if planes landed during off-peak hours but
departed d1tring peak periods.
29 A similar argument would apply to a takeoff fee.
30 To 	the 	extent 	that 	the 	rescheduled 	flights 	increased 	delays 	during
off-peak periods, then the estimated benefits of time-variable fees would be
reduced somewhat compared to the uniform fee.
13

V. Priority and Postponable Senice 31
We offer for discussion a system of priority service at a 	higher price
for Logan's IFR-I capacity of 80 operations/hour, and postponable service at
a lower price for the remaining capacity of 30 operations/hour under visual-
flight-rule (VFR) conditions.32 	On IFR-l days, Logan would first attempt to
handle 	the 	priority operations at 	their scheduled 	times. 33 	By contrast, 	the
postponable flights would be delayed until an off-peak time when they could
be accommodated without delaying a priority flight. 3(
A system 	with 	both 	priority 	and 	postponable service 	appears 	to 	have
significant 	advantages 	over 	the 	current 	system 	in 	which 	all 	flights 	are
postponable. 	Now, in the event of weather problems or mechanical problems	,
31 We 	recognize 	that 	such 	a 	system 	would 	have 	to 	be 	implemented
jointly by the FAA, Logan, and other airports. 	Nevertheless, we discuss the
system 	in 	these 	c;omments 	because 	of 	its 	relevance 	to 	the 	net 	benefits 	of
Massport's pricing proposals.
32 In 	fact, 	under 	IFR 	conditions, 	Logan's 	capacity 	can 	vary
substantially, depending on the runway configuration that can be used, from
56 operations per hour to 98 operations per hour (telephone interview with
Massport staff). 	This variation suggests the plan might be modified to offer
different degrees of priority for the different levels of capacity. 	This would
clearly increase the complexity and administrative cost of the plan. 	Whether
the added benefits would exceed the costs would 	have to be determined by
careful analysis.
33 We discuss priority service only with regard to weather conditions at
Logan. 	We recognize, however, the relevance of weather conditions facing
the 	air 	traffic 	control 	system 	and 	other 	airports. 	For 	example, 	a 	flight
given priority at Logan could still be delayed because of weather elsewhere.
Also, aircraft whose landings were postponed at Logan would have to remain
at their origin airports, perhaps increasing congestion aHhose airports.
3( Alternatively, 	operators 	whose 	flights 	were 	postponed 	in, 	for
example, the first afternoon peak hour might purchase priority service in the
second 	afternoon 	peak 	hour. 	This 	variant 	of 	our 	suggestion 	would 	make
postponable service more attractive, but would also add to the complexity of
the system. 	Whether the added benefits exceed 	the added costs would have
to be determined by a careful examination of this and other alternative systems.
14

to one flight, all flights 	risk delay. 	There is no market in which 	aircraft
operators 	can 	purchase 	priority 	service 	from 	airports 	and 	the 	air 	traffic
control system. 	Both airlines and owners of general-aviation aircraft may be
better off 	if they 	were 	offered a choice 	between such 	service at a 	higher
price and postponable service at a lower price. 3s
A 	system 	of 	priority 	and 	postponable service 	would 	eliminate 	flights
only under IFR-I 	conditions or worse at 	Logan, which occur an average of
one day in five. 	If such a system were added to the proposal to charge $91
+ 	$0.45/1,000 	lbs. 	at 	all 	hours 	of 	the 	day, 	the 	estimated 	cost 	would 	be
reduced 	to 	one-fifth 	of 	the 	estimated 	cost 	of 	the 	price 	change 	by 	itself,
which would eliminate flights every day.36
Implementing such a pl,an might cause aircraft operators to change their
ticketing policies. 	For example, operators could respond to the higher price
of priority service ?y charging higher fares, or by offering fewer discounted
.
seats 	on 	priority 	flights. 	Operators 	could 	buy 	priority 	service 	for 	flights
that 	carry 	passengers 	who 	place 	a 	high 	value 	on 	on-time service, 	such 	as
business travelers, 	and 	postponable service for 	flights 	that carry passengers
who 	place 	a 	lower 	value 	on 	on-time service, such as 	students. 	The 	latter
3S Neither the current system nor any contemplated system can ensure
that 	a 	flight 	will 	never 	be 	delayed. 	A 	system 	of 	priority 	service 	might,
however, permit some flights to reduce the risk of a delay.
36 If 	the 	postponable 	flights 	that 	were 	not 	cancelled 	caused 	the
reduction in priority-flight delays to be somewhat smaller, then the benefits
would 	also 	be 	somewhat 	reduced. 	With 	available 	information, 	we 	cannot
evaluate this possibility.
IS

might be willing to risk delay when the weather is bad to save money every
time they fly.sf
VI. Priority Peak Hour Fee Procedure
One of the alternative plans in PACE would group aircraft by size and
give large aircraft priority for 85 percent of the flights during peak periods.
Rights 	to 	the 	remaining 	flights 	would 	be 	auctioned 	to 	small 	aircraft 	and
essential service (PACE, Exhibit M).38
Although 	we 	do 	not 	have 	sufficient 	information 	at 	this 	time 	to
calculate the net benefits of such a plan, there is evidence that, in general,
the use of auctions to allo~ate landings can yield substantial 	net benefits. 39
However, the benefits of this plan might be 	reduced 	because of the use of
the 8S-percent cutoff point. 	Net benefits may be increased if the auction is
.
applied to all aircraft sizes.
Although aircraft size tends to be positively related to the net benefits
of a flight to society, decreasing the number of small-aircraft landings by a
31 Although we do not know whether airlines would find it profitable
to 	offer 	both 	priority 	and 	postponable 	service 	to 	their 	passengers, 	we
suggest 	that 	such 	a 	system 	merits 	consideration 	as 	a 	possible 	solution 	to
weather-related 	delay 	problems. 	FAA 	data 	indicate 	that 	weather 	causes
approximately 70 percent of total delays (see Ogur, Wagner, and Vita, 1988,
p. 34). 	One issue that would have to be carefully addressed in implementing
such a scheme is how to handle flights that are en route when a change in
the 	weather 	reduces 	the 	capacity 	of 	the 	system. 	In 	this 	regard, 	it 	is
important to recall that all flights are postponable under the current system.
38 PACE seeks to increase the average size of the aircraft that land at
Logan in order to maximize passenger throughput (PACE, p. 20), subject to
the constraint that revenues equal costs (PACE, p. 22).
39 Studies of airport slot auctions 	are cited 	in 	Koran and 	Ogur, 	1983,
p.6.
16

system that assigns them lower pri~rity may lead to economically inefficient
results, because some flights in small aircraft may have higher net benefits
than 	some flights 	in 	larger aircraft. 	For example, 	the only 	morning-peak-
period 	flight 	in a commuter aircraft 	between a small city and 	Logan 	may
provide 	more 	net 	benefits 	to 	society 	than 	the 	fifth 	morning-peak-period
flight, 	in a 	larger aircraft, between another, 	larger city and 	Logan. 	If so,
the operator of 	the commuter aircraft may be willing to pay a high peak-
period landing fee, but the operator of the larger aircraft may not.
Exclusion 	of 	a 	more 	beneficial 	flight 	may 	occur 	when 	the 	larger
aircraft receives priority, but is carrying fewer passengers than the smaller
aircraft, 	which 	does 	not 	receive 	priority 	(see 	PACE, 	Exhibit 	M). 	For
example, some 	flights 	in 	a 	Beech 	99 	with 	fewer 	than 	30 seats 	may 	carry
more passengers and be wi'uing to pay more for landings than some flights in
a DC-3 with 30 seats (PACE, Exhibit H).40
VII. Conclusions
PACE 	appears 	to 	be 	a 	step 	toward 	economically 	efficient 	pricing 	of
airport services that would provide significant benefits to aircraft operators
and passengers. 	Based on Massport's assumptions regarding cost and demand
conditions at Logan, it appears that the proposal 	to charge landing fees of
$91 	+ $0.45/1,000 	lbs. 	would 	yield 	benefits 	that 	are 	greater 	than 	its costs.
40 An 	addit-ional 	problem with 	this 	scheme 	is 	that 	it 	would 	limit 	the
total 	number of 	operations 	to Logan's 	IFR-I 	capacity of 80 operations 	per
hour, 	even 	under 	VFR 	conditions 	(PACE, 	Exhibit 	M). 	This 	problem 	was
examined 	above 	in 	our 	consideration 	of 	a 	scheme 	for 	priority 	and
postponable service.
17

The alternative proposal to charge S91 + $0.72/1,000 Ibs. on-peak and S50 +
$0.72/1,000 	Ibs. 	off-peak 	may 	increase 	net 	benefits 	even 	further. 	Such 	a
plan would probably eliminate fewer off-peak landings and would encourage
peak-period 	general-aviation 	flights 	to 	shift 	to 	off-peak 	times 	instead 	of
being cancelled or shifting to less-convenient airports!l 	Net benefits might
be further increased if fees were set equal to the added cost of providing a
landing. 	Finally, 	a 	system 	of 	priority and 	postponable 	service 	might 	also
increase net benefits .
.u We 	lack 	sufficient 	information 	to 	reach 	any 	conclusions 	regarding
the effect on net benefits of the proposal to charge a $65 fee for the use of
the general-aviation terminal and 	the proposal 	to charge S300 + S0.45/1,000
Ibs. on the peak and only S0.45/1,000 Ibs. off the peak.
18
r

APPENDIX
BENEFIT AND COST ESTIMATES: ASSUMPTIONS AND CALCULATIONS
Assumptions
With 	the 	information 	available 	in 	PACE, 	we 	can 	perform 	a 	sample
estimation of the benefits and costs of Massport's proposal to change landing
fees 	from 	$1.31/1,000 	Ibs., 	with 	a 	minimum 	of $25 	per 	landing, 	to 	$91 	+
$0.45/1,000 	Ibs. 	To 	do 	this 	estimation, 	we 	will 	take 	as 	given 	Massport's
description of cost and demand conditions at Logan.
On the cost side, the number of takeoffs and landings (operations) that
Logan 	can 	supply 	depends 	on 	the 	weather 	at 	the 	airport. 42 	Logan 	can
supply 110 operations/hour under VFR conditions and ideal wind conditions.
I
.Logan 	can 	supply an 	average 	of 80 operations/hour 	under 	IFR 	conditions.
Reduced capacity due to impaired 	visibility -- IFR-l 	or worse -- occurs 20
percent 	of 	the 	time, 	on 	average. 	Unfavorable 	wind 	conditions 	cause
additional capacity reductions (PACE, pp. 12-13).43
The demand for operations at Logan is subject to hourly and seasonal
peaks 	and 	is 	projected 	to 	grow 	substantially. 	An 	hourly 	peak 	demand 	of
approximately 	100 operations/hour occurs at Logan during the morning and
afternoon peak periods (PACE, pp. 13-14 and Exhibit E). 	August is the peak
month 	for 	scheduled 	operations 	at 	Logan; 	March 	is 	the 	peak 	month 	for
general 	aviation 	operations (PACE, 	Exhibit 	M). 	Projected 	increases 	in 	the
42 Massport's 	estimates 	of 	Logan's 	ability 	to 	supply 	operations
(capacity) 	assume 	that 	delays 	do 	not 	exceed 	the 	FAA's 	standard 	of 	15
minutes per operation (telephone interview with Massport staff).
43 Such unfavorable wind conditions occur an average of 2-3 percent of
the time under VFR conditions (telephone interview with Massport staff).
19

demand 	for 	operations at 	Logan 	will cause 	peak-period 	demand 	to 	exceed
VFR capacity within the next decade (PACE, pp. 14-15).
Thus, at prevailing fares and landing fees, peak-period demand exceeds
Logan's capacity by approximately 20 operations/hour under IFR-I conditions.
As 	a 	result, 	approximately 	7000 	passengers/hour 	experience 	delays 	of
approximately 20 minutes during the morning peak and 60 minutes during the
afternoon 	peak. 	Massport 	considers 	dela ys 	grea ter 	than 	15 	mi n utes 	per
opera tion 	to 	be 	unaccepta ble 	(PACE, 	pp. 	15-16). 	Over 	the 	entire 	day,
operations 	incur 	a 	total 	of 	34,204 	minutes 	of 	delay, 	or 	an 	average 	of 	26
minutes per operation (PACE, Exhibit L).H
At 	Logan, 	air 	carriers 	conduct 	approximately 	60 	percent 	of 	the
operations, 	carry 	94 	percent 	of 	the 	passengers, 	and 	use 	aircraft 	with 	an
average capacity of 165 seats. 	Commuter airlines conduct 30 percent of the
operations, 	carry 	5 	percent 	of 	the 	passengers, 	and 	use 	aircraft 	with 	an
average 	capacity 'between 	20 	and 	25 	seats. 	General 	aviation 	conducts 	10
percent 	of 	the 	operations 	and 	carries 	less 	than 	one 	percent 	of 	the
passengers (PACE, pp. 18-19).
To estimate benefits and costs, we make use of Massport's assumptions
and 	estimates. 	We 	adopt 	the assumption 	that 	the 	proposed 	pricing change
will eliminate five 	percent of the 	operations in each 	hour, and 	that all of
the 	eliminated 	operations 	will 	be 	by 	general-aviation 	aircraft 	(PACE, 	p.
29).46 	Thus, we assume that air carriers and commuters will not respond to
H 	Exhibit L pres-ents average delay per aircraft, which is equivalent to
average delay per operation (telephone interview with Massport staff).
45 We assume that the proposed pricing change will raise the minimum
landing 	fee 	that 	is 	actually 	paid 	from 	$25 	to 	approximately 	SIOO. 	This
assumption 	is 	based on 	Massport's 	projection 	that 	the 	proposed change 	will
increase 	the 	landing 	fees 	for 	a 	Beech 99 	and 	a 	Beech 	1900 	to 	$95.54 	and
20

the proposed change in 	landing fees 	by changing the number of operations
demanded." 	We also use Massport's estimate that the proposed change will
reduce total delay time to 18,735 minutes, or an average of 	IS 	minutes per
operation, 	the 	level 	that 	Massport 	considers acceptable 	(PACE, 	Exhibit 	L).
Thus, the proposed change will decrease the average delay by II 	minutes per
operation.
The estimated benefits of the 	proposed pricing change equal the value
of the time saved by passengers on the remaining flights. 	We make several
assumptions 	that 	reduce 	these 	estimated 	benefits.
S98.48, respectively.
First, 	we 	assume 	that
46 Air-carrier operati~ns are 	likely 	to 	be less 	responsive 	to changes in
landing fees 	than are general-aviation operations. 	First, landing fees 	are a
relatively 	small 	fraction 	of 	the 	total 	operating 	costs 	of 	air-carrier 	flights.
Also, 	air-carrier 	passengers 	to 	Boston 	appear 	to 	have 	no 	equally 	good
alternative airports to Logan. 	Thus, in response to the decrease in landing
fees that large aircraft would pay (see PACE, Exhibit J), it is reasonable as
a 	first 	approximation 	to 	assume 	that 	air 	carriers 	will 	not 	increase 	their
number of operations.
By contrast, 	landing 	fees 	are 	a 	larger 	fraction 	of 	the 	total 	operating
costs of general-a via tion flights. 	Also, general-a via tion passengers appear to
have 	both 	good 	alternative 	airports 	-- 	for 	example, 	Hanscom 	Field 	-- 	and
good alternative carriers, the scheduled carriers (PACE, pp. 19-20). 	Thus, in
response 	to 	the 	higher 	landing 	fees 	that 	small 	aircraft 	would 	pay, 	it 	is
reasonable 	to 	assume 	that 	general 	aviation 	will 	reduce 	the 	number 	of 	its
operations. 	Although 	we 	have less information on commuter carriers, their
responsiveness 	probably 	lies 	somewhere 	between 	those 	of 	the 	other 	two
groups.
At 	present 	we 	lack 	the 	information 	needed 	to 	estimate 	benefits 	and
costs 	under 	the 	assumption 	that 	air 	carriers 	and 	commuters 	change 	the
number 	of 	their 	operations 	in 	response 	to 	changes 	in 	landing 	fees. 	This
assumption would probably be needed to estimate the benefits and costs of
Massport's alternative proposal to charge $300 + $0.45/1,000 Ibs. during peak
periods and $0.45/1,000 Ibs. during nonpeak periods. 	The additional needed
information 	would 	include 	the number of 	flights 	and 	the distance of each
flight by each aircraft type flown by air carriers and commuters. 	With this
information, 	it 	might 	be 	possible 	to 	use estimates of 	the 	responsiveness of
the carriers to landing fee changes to predict the change in the number of operations.
21

both air carriers and commuters operate with 	load 	factors of SO percent..n
Second, we assume that the general-aviation flights that remain despite the
price 	increase 	carry 	an 	average 	of 	two 	passengers 	per 	fIigh t. 4S 	Both 	of
these 	assumptions 	probably lead 	us 	to understate 	the 	number of 	passengers
whose delay time is reduced. 	Third, we value passengers' saved 	time at $S
per 	hour.'HI 	Finally, 	we 	omit any 	benefits 	to 	aircraft 	operators, 	who 	may
benefit 	from 	decreased 	fuel 	costs 	and 	crew 	salaries 	because 	of 	reduced
delays.
The 	estimated costs of 	the 	proposed 	pricing change equal 	the 	amount
that 	operators 	are 	willing 	to 	pay 	for 	foregone 	landings, 	rather 	than
cancelling 	these 	flights 	or 	diverting 	them 	to 	other 	times 	or 	other airports,
minus 	the 	resources 	that 	would 	be 	used 	to 	provide 	the 	landings. 	We
I
I
increase the estimated costs by assuming that no resources would be used to
provide 	the 	landings. 5o 	We 	also 	assume 	that 	the 	operators' 	willingness 	to
41 In its calculations, Massport assumes a load factor of 70 percent for
selected aircraft sizes (see PACE, Exhibit H).
4S According 	to 	Massport, 	almost 	7000 	passengers 	per 	peak 	hour 	use
Logan during peak periods, of which less than one percent are served by the
10 general-aviation flights per peak hour. 	These figures imply an average of
fewer than seven passengers per general-aviation flight.
49 A. DeVany (liThe Revealed Value of Time in Air Travel," Review of
Economics 	and 	Statistics, 	1974, 	pp. 	77-82) 	estimated 	the 	value 	of 	air
passengers' 	time 	at 	$7.28 	per 	hour 	in 	1968. 	Adjusted 	for 	inflation, 	this
would equal $20.37 in 	1986 dollars. 	Also for 	1968, Carlin and Park (1970)
assumed values of $6 per hour for air-carrier passengers, and $12 per hour
for 	general-aviation 	passengers. 	Morrison (1983) 	used 	values of time equal
to 	$5, 	$10, 	and 	$15 	per 	hour. 	By 	lowering 	fares 	and 	making 	air 	travel
affordable 	to 	lower-income 	consumers, 	deregulation 	may 	have 	lowered 	air
passengers' 	average 	value 	of 	time. 	Nevertheless, 	our 	$5 	assumption 	is
probably lower than the actual value of air travelers' time.
50 For 	the 	mid-1970s, 	Morrison 	(1983) 	estimated 	that 	the 	typical
airport's 	marginal 	cost 	of 	maintenance, 	operations, and 	administration 	was
approximately $12 for an air carrier, but not significantly greater than zero
for general aviation.
22

. ,
pay for landings is equal to the value of the flights to passengers minus the
operating costs of the flights. 51
Calculatlons 52
Number of operations per day at current landing fees:
Total 	number 	of 	operations 	.. 	total 	minutes 	of 	delay 	divided 	by 	average
minutes of delay per operation
= (34,204 minutes)/(26 minutes/operation)
.. 1,316 operations
Number of air-carrier operations .. (0.6)(1,316 operations)
- 789 operations
Number of commuter operations - (0.3)(1,316 operations)
.. 395 operations
.,
Number of general-aviation operations .. 1,316 operations - 789 operations
- 395 operations
51 For each flight, the area below the demand curve for the flight and
above 	the 	flight's 	operating cost is a 	measure 	of 	the 	economic concept 	of
the 	consumer 	surplus 	that 	passengers 	receive 	from 	the 	flight. 	Consumer
surplus 	equals 	the 	amount 	that 	consumers 	are 	willing 	to 	pay, 	rather 	than
forgo the flight, minus the amount that they actually pay, which we assume
equals 	the 	operating 	costs 	of 	the 	flight. 	For 	general-aviation 	flights, 	the
operator's willingness 	to 	pay for a 	landing is a 	good approximation of 	the
passengers' willingness to pay for the flight minus the operating costs of the
flight. 	In the case of business, family, or single-person flights, there is only
one 	decision 	maker, 	or decision-making 	group, 	whose 	willingness 	to 	pay 	is
relevant. 	In 	the 	case 	of 	a 	group 	of 	unrelated 	individuals, small 	numbers
facilitate a reasonably accurate determination of the combined willingness to
pay. 	These conditions do not 	hold 	for 	the 	larger 	groups of 	passengers on
commuter 	and 	air-carrier 	flights. 	See 	Yance 	(1971) 	and 	Koran 	and 	Ogur
(1983) for an analysis of air carrier's demand for landings.
The measure that we use equals the entire area under t-he-demand curve
for 	landings, 	between 	the 	original 	quantity 	of 	landings arid 	the 	reduced
quantity after 	the price change (see 	R. Turvey, "How to Judge 	When 	Price
Changes 	Will 	Improve 	Resource 	Allocation," 	Economic 	Journal, 	December
1974, pp. 825-832).
52 The 	results 	presented 	in 	this 	section 	are 	affected 	slightly 	by
rounding errors.
23

• 132 opera tions
Number of operations per day at proposed landing fees:
Total 	number 	of 	operations 	= total 	minutes 	of 	delay 	divided 	by 	average
minutes of delay per operation
= (18,735 minutes)/(15 minutes/operation)
:0 1,249 operations
Number 	of 	general-aviation 	operations = 	1,249 operations - 789 operations-
395 opera tions
'" 65 operations 53
Number of passengers per day at proposed landing fees:
Number of air-carrier 	passengers - 	number of operations 	times 	number of
seats per operation times load factor
= (789 operations)(l65 seats/operation)
(0.5 passengers/seat)
==65,119
Number 	of 	commuter 	passengers 	= number 	of 	operations 	times 	number 	of
seats per operation times load factor
= (395 operations)(22.5 seats/operation)
(0.5 passengers/sea t)
= 4,440
Number 	of 	general-a viation 	passengers 	= 	(65 	operations)(2
passengers/ opera tion)
= 130
Total number of passengers = 65,119 + 4,440 + 130
53 By assumption the number of air-carrier and commuter operations is
unaffected by the change in landing fees.
24

= 69,689
Hours saved, benefits, and costs per d ay 64:
Hours saved - 	frequency of IFR days times number of passengers times hours
sa ved per passenger
= (0.2)(69,689 passengers)(J 1/60 hours/passenger)
= 2,555 hours
Benefits = hours sa ved times value of time per hour
= (2,555 hours)($5/hour)
= $12,776
Costs 55 = number of landings per operation times number of general-aviation
operations eliminated times average value per operation
= (0.5 landings/opera tion)( 67 opera tions){[($1 00+25)(0.5)]/opera tion}	.
= $2,079
Benefits/Costs - $12,776/$2,079
.
= 6.14
Benefits - Costs = $10,697
Hours saved and net benefits per year: 56 .
50{ Time saved and 	benefits are 	multiplied 	by 0.2 	because they accrue
only under IFR-l or worse conditions, which occur an average of one day in
five, or 20 percent of the time.
55 The 	number of general-aviation 	operations eliminated 	was obtained
by subtracting the total number of operations at the proposed landing fees,
1,249, from the total number of operations at the current landing fees, 1,316.
The number of landings eliminated is obtained by multiplying the number of
operations by (0.5) because there 	is one takeoff for each landing, and thus
half as many landings as operations. 	Because landings are eliminated every
day, costs are-not multiplied by (0.2). 	As we noted above, Massport charges
fees per landing, but not per takeoff.
56 For 	these 	calculations, 	we 	assume 	that 	each 	flight 	is 	made 	on 	the
250 weekdays of the year. 	Seasonal variation in the extent of delays, due to
weather and seasonal demand peaking, may affect net benefits. 	We lack the
information needed to examine these possible effects.
25

Hours saved - (2,555 hours/day)(250 days/year)
- 638,817 hours
Benefits - Costs • (SI0,697/day)(250 days/year)
: 	$2,674,253
Present discounted value of benefits· costs (PDV)57:
POY = ($2,674,253)/(0.10)
= $26,742,533
57 For 	this 	calculation, 	we 	assume 	that 	the 	benefits 	accrue 	in
perpetuity and are discounted at 10 percent per year.
26