Commercial Aviation: Air Service Trends At Small Communities	 
Since October 2000 (29-MAR-02, GAO-02-432).			 
								 
Most major U.S. airlines began realizing net operating losses	 
early in the 2001, and all of the major U.S. passenger carriers  
except Southwest Airlines reported losses for the year. Travelers
throughout the nation shared in the difficulties. In October	 
2000, the typical or median small community that GAO analyzed had
service from two airlines, with a total of nine daily departing  
flights. Forty-one percent of the communities were served by only
one airline with size being the most obvious factor for service  
limitations. However, the level of service also varied by the	 
level of local economic activity.  The total number of daily	 
departures from these small communities declined by 19 percent	 
between October 2000 and October 2001. Although carriers had	 
reduced total departure levels at small communities before	 
September 11th, airlines made even more reductions after that	 
date.  Because profitability is so critical to airline decisions 
about what markets to serve and how to serve them, the changes in
service levels in small communities can be traced to economic	 
factors. Two such factors--the economic decline that began in	 
early 2001 and the collapse of airline passenger traffic after	 
September 11--are widely acknowledged as the main contributors to
declining profitability in the industry.			 
-------------------------Indexing Terms------------------------- 
REPORTNUM:   GAO-02-432 					        
    ACCNO:   A02940						        
  TITLE:     Commercial Aviation: Air Service Trends At Small	      
Communities Since October 2000					 
     DATE:   03/29/2002 
  SUBJECT:   Airline industry					 
	     Travel						 
	     Terrorism						 
	     Losses						 

******************************************************************
** This file contains an ASCII representation of the text of a  **
** GAO Product.                                                 **
**                                                              **
** No attempt has been made to display graphic images, although **
** figure captions are reproduced.  Tables are included, but    **
** may not resemble those in the printed version.               **
**                                                              **
** Please see the PDF (Portable Document Format) file, when     **
** available, for a complete electronic file of the printed     **
** document's contents.                                         **
**                                                              **
******************************************************************
GAO-02-432
     
Report to Congressional Requesters

United States General Accounting Office

GAO

March 2002 COMMERCIAL AVIATION

Air Service Trends At Small Communities Since October 2000

GAO- 02- 432

Page i GAO- 02- 432 Small Community Air Service Letter 1

Results in Brief 3 Background 5 Most Small Communities Had Limited Air
Service in October 2000 7 Airline Service at Small Communities Declined
Between October

2000 and October 2001 17 National Economic Decline and September 11 Attacks,
Along with

Airlines? Strategic Decisions, Underlie Shifts in Air Service 22 Concluding
Observations 28 Agency Comments 29

Appendix I Objectives, Scope, and Methodology 31 Internet Survey of Airport
Directors 35 Other Data Limitations to Our Work 37

Appendix II Small Hub Airports Tend to Have More Commercial Air Service and
Were Less Affected by Airline Service Reductions 38

Appendix III Change in Air Service at 202 U. S. Nonhub Airports, October
2000- October 2001 42

Appendix IV Changes in Air Service at Small Communities in Alaska and Hawaii
54

Appendix V Economic Factors Affect Air Service in Small Communities 56

Regression Model 56 Economic Principles 56 Data 57 Variables Used in the
Model 57 Model Specification 58 Results 58 Contents

Page ii GAO- 02- 432 Small Community Air Service Appendix VI Changes in Air
Service at Small Communities

Made By Express Airlines I and Mesaba Between 2000 and 2001 60

Appendix VII Changes in Air Service at Small Communities Made By Great Lakes
Aviation and Air Wisconsin Between 2000 and 2001 62

Appendix VIII GAO Contacts and Staff Acknowledgments 64 GAO Contacts 64
Acknowledgments 64

Related GAO Products 65

Tables

Table 1: Differences Among Categories of Commercial Airports 7 Table 2:
Summary of Air Service at Small Communities, October

2000 9 Table 3: Range of Service Levels Among Small Community

Airports, October 2000- October 2001 19 Table 4: Number of Nonhub Airports
Eliminated by Various

Selection Criteria 33 Table 5: Characteristics of Air Service at Nonhub and
Small Hub

Airports (October 2000) 38 Table 6: Median Airline Service Levels at Nonhub
and Small Hub

Airports, October 2000 39 Table 7: Changes in the Number of Carriers Serving
Small

Community Airports and Nonstop Destinations 42 Table 8: Changes in the
Average Number of Daily Turboprop and

Jet Departures at Small Communities 48 Table 9: Median Number of Carriers,
Nonstop Destinations Served,

Daily Turboprop and Jet Departures for Nonhub Airports in Alaska and Hawaii
54

Page iii GAO- 02- 432 Small Community Air Service

Table 10: Median Number of Carriers Serving, Nonstop Destinations Served,
Daily Turboprop and Jet Departures for Small Hub Airports in Alaska and
Hawaii 55 Table 11: Characteristics of 202 Small Communities Included in

Economic Analysis 57 Table 12: Nonstop Air Service Changes for Communities
Served by

Mesaba and Express Airlines I Between October 2000 and October 2001 61 Table
13: Nonstop Air Service Changes for Communities Served by

Great Lakes and Air Wisconsin Between October 2000 and October 2001 63

Figures

Figure 1: Location and Estimated Catchment Areas of 202 Small Community
Airports in the Continental United States 10 Figure 2: Comparison of Air
Service Indicators Across Small

Communities by Population Category, October 2000 (Median Service Levels) 12
Figure 3: Proximity of Small Community Airports to Other Airports

Either Served by a Low- fare Airline or Serving as a Major Airline?s Hub 15
Figure 4: Change in Total Daily Departures for Small Communities,

October 2000- October 2001 18 Figure 5: Change in the Number of Carriers
Serving Small

Communities, October 2000- October 2001 20 Figure 6: Change in Total U. S.
Passenger Enplanements and

Revenue, September 2000- August 2001 23 Figure 7: Passenger Traffic and
Revenues Before and After

September 2001 24 Figure 8: Change in the Number of Airlines Serving Small

Communities with Nonhub Airports, October 2000- October 2001 40 Figure 9:
Change in the Number of Airlines Serving Communities

with Small Hub Airports, October 2000- October 2001 41

Page iv GAO- 02- 432 Small Community Air Service Abbreviations

BTS Bureau of Transportation Statistics DOT Department of Transportation EAS
Essential Air Service FAA Federal Aviation Administration GAO General
Accounting Office

Page 1 GAO- 02- 432 Small Community Air Service

March 29, 2002 Congressional Requesters: For the airline industry and air
travelers, 2001 was a difficult year. Obviously, the terrorist attacks of
September 11 exerted unprecedented effects on the airlines and their
passengers. But, the industry?s problems preceded those tragedies. Most
major U. S. airlines began realizing net operating losses early in the year,
and all of the major U. S. passenger carriers except Southwest Airlines
reported losses for 2001. United Airlines, American Airlines, and Delta Air
Lines each reported losses of more than $1 billion. Delta Air Lines and its
passengers experienced a difficult period when the pilots at its regional
affiliate Comair went on strike for 89 days, grounding Comair?s 815 daily
flights. Sun Country Airlines suspended scheduled operations. Travelers
throughout the nation shared in the difficulties- passengers at the nation?s
largest airports did so, as did those at airports serving the nation?s small
communities. 1

In recent years, we have reported on the effects of changes in the airline
industry on service, including service at small community airports. 2
Concerned about air service to small communities, especially in light of
recent events, you asked us to

 describe the overall level of air service at the nation?s small
communities in 2000 and the main factors that contributed to that service
level;  examine how the nature and extent of air service changed among the

nation?s small communities in 2001, including a specific accounting for how
service changed after the September 11 terrorist attacks; and  identify key
factors that have influenced these changes in air service.

1 In this report, ?small communities? refers to a selected sample of
communities served by airports that are defined as ?nonhubs? by statute (49
U. S. C. 41731). These small communities include such locations as Altoona,
Pennsylvania; Hattiesburg, Mississippi; Salina, Kansas; and Eureka/ Arcata,
California- communities from which relatively few passengers travel compared
with other U. S. airports. The background section of this report discusses
the small communities included in this study in more detail.

2 See, for example, U. S. General Accounting Office, Airline Deregulation:
Changes in Airfares, Service Quality, and Barriers to Entry, GAO/ RCED- 99-
92 (Washington, D. C.: Mar. 4, 1999) and Aviation Competition: Regional Jet
Service Yet to Reach Many Small Communities, GAO- 01- 344 (Washington, D.
C.: Feb. 14, 2001). A more complete list of Related GAO Products follows
appendix VIII.

United States General Accounting Office Washington, DC 20548

Page 2 GAO- 02- 432 Small Community Air Service

Our analysis of air service at these communities is based in part on
published airline service schedules for 202 small communities in the
continental United States. 3 To assess changes in air service over time, we
compared the schedules that the airlines developed for the week of October
15- 21, 2000, against two different schedules that the airlines developed
for the week of October 15- 21, 2001. 4 We assessed general changes in
scheduled air service by comparing the October 2000 schedules against
schedules for the week of October 15- 21, 2001, that the airlines developed
before the September 11 attacks. To account more specifically for how
service changed in small communities after the September 11 terrorist
attacks, we compared the changes in airline schedules against a second set
of schedules for the week of October 15- 21, 2001, which the airlines
developed after the September 11 attacks. We refer to the first as the
?original? schedules and to the second as the ?revised? schedules. 5 As with
our previous reports on changes in airfares and service, we analyzed various
changes in service that occurred in these communities such as changes in the
number of turboprop and jet departures, changes in the number of carriers
providing service, and changes in the number of destinations served with
nonstop flights. We recognize that airlines make frequent changes to their
service schedules, and that service at these communities may have changed
since mid- October, 2001. In addition, we developed an Internet- based
questionnaire for managers of small airports we included in the study. This
questionnaire allowed the managers to comment on the level of service they
were receiving and to provide information to assess factors that might
affect schedule changes. We supplemented this work with reviews of other
information and interviews with various industry experts and officials.
Additional information on the scope and methodology of this report can be
found in appendix I.

3 In addition to the 202 communities we analyzed within the 48 contiguous
states, we also analyzed air service at 65 small communities in Alaska and
Hawaii. These two states have markedly different patterns of commercial air
service from the rest of the United States. To avoid skewing the results, we
analyzed them separately. See appendix IV for this discussion. See
appendixes I and III for additional information on the communities we
studied.

4 We obtained air service schedule data from the Kiehl Hendrickson Group.
The first October 2001 schedule dataset reflected the schedule as of August
30, 2001, and, therefore, does not reflect the airline industry?s reaction
to the events of September 11. The second schedule dataset for October 15-
21 reflected the schedule as of October 12, 2001.

5 Because our analyses involved a review of ?snapshots? in time and the
nature of the service may have been fluid, we could not fully measure
interim service changes that may have occurred (e. g., carriers initiating
service at a community but discontinuing it soon afterwards) within a single
time period.

Page 3 GAO- 02- 432 Small Community Air Service

In October 2000, the typical or median small community in our analysis had
service from two airlines, with a total of nine daily departing flights.
Forty- one percent of the communities were served by only one airline.
Airlines performed the vast majority of these flights- over 80 percent- with
turboprop aircraft. The most obvious factor that accounts for the relatively
limited level of service at these communities is their size. The median
community population was about 120,000, and the median number of daily
passenger enplanements in 1999 was about 150. Within the group of 202
communities we studied, those with the smallest populations generally had
the lowest levels of service. However, the level of service also varied for
two other main reasons. One is the level of local economic activity, as
measured by such indicators as employment or per capita income. Small
community airports with the highest levels of service tended to be those
serving communities with more economic activity, such as Northwest Arkansas
Regional Airport (located near the headquarters for Wal- Mart Stores, Inc.)
or resort destinations like Key West, Florida. Such communities may have 60
or more scheduled commercial departures per day, while communities at the
other end of the scale have only 1 or 2. Another factor that contributes to
limited service for many of these communities is their proximity to larger
airports. Nearly half of the 202 communities we analyzed are within 100
miles of an airport that serves as a major airline?s hub or that is served
by a low- fare carrier. This is particularly the case in the eastern half of
the United States. Airport directors at many small communities reported that
passengers opt to drive to other larger airports for different connections
or lower fares from a major airline.

Between October 2000 and October 2001 (revised), the total number of daily
departures from these small communities declined by 19 percent. Although
carriers had clearly reduced total departure levels at small communities
before the September 11 terrorist attacks, airlines made more of the total
reduction in departures after September 11. Analyses of industry service
levels show that communities of all sizes shared in service reductions. At
the typical small community, the number of departures dropped by three
flights per day, from nine to six. To the extent that these communities had
jet service, it was largely unaffected; nearly all of the decline came in
turboprop flights. Service indicators other than the number of departures-
such as the number of airlines providing service- also showed general
declines, both before and after September 11. For example, the percentage of
small communities served by only one airline increased from 41 percent in
October 2000 to 47 percent by October 2001, with slightly more of the
increase coming before September 11. When one or more carriers pulled out of
a community, passengers often lost Results in Brief

Page 4 GAO- 02- 432 Small Community Air Service

connecting service to other destinations. However, while service reductions
predominated, airlines initiated service at 14 of the 202 communities.
Virtually all of those increases in service had been planned or put in place
before September 11.

Because profitability is so critical to the decisions airlines make about
what markets they serve and how they serve them, the changes in service
levels in small communities can be traced to economic factors. Two such
factors- the economic decline that began in early 2001 (according to the
National Bureau of Economic Research) and the collapse of airline passenger
traffic after September 11- are widely acknowledged as the main contributors
to declining profitability in the industry. As the nation?s economic
performance declined, fewer passengers- especially business passengers who
generally pay higher fares- opted to fly, and airline revenues dropped.
Carriers sought ways to control costs and did so, in part, by reducing
scheduled service to many locations, including small communities. After
September 11, passenger traffic and revenues plummeted, exacerbating the
situation. As at larger airports, passenger levels at small community
airports dropped substantially, according to our survey of small community
airport directors. Airlines responded by cutting capacity, partly by
reducing the number of flights. Where airlines decided to withdraw
altogether from a small community, they most often did so in communities
where they were competing with other airlines and had a limited market
share. Such competition was more likely to be going on in small communities
with populations greater than 100,000; communities with even smaller
populations generally already received service from only one airline. As a
result, communities with populations above 100,000 tended to see greater
service reductions than their smaller counterparts. Besides the economic
slowdown and September 11 attacks, airlines? decisions about the makeup and
deployment of their fleet also influenced service levels at small
communities. For example, some communities lost service when carriers
retired certain small types of aircraft.

We provided a copy of the draft report to the Department of Transportation
(DOT) for review and formal comment. We also provided sections of our draft
report for technical comment to Northwest Airlines, United Airlines, Great
Lakes Aviation, and Air Wisconsin. Officials with DOT and the airlines
offered only technical comments, which we incorporated into the report as
appropriate.

Page 5 GAO- 02- 432 Small Community Air Service

The U. S. air transportation structure is dominated by ?hub- and- spoke?
networks and by agreements between major airlines and their regional
affiliates. Since the deregulation of U. S. commercial aviation in 1978,
most major airlines have developed hub- and- spoke systems. 6 For example,
Northwest Airlines (Northwest) has hubs in Minneapolis, Detroit, and
Memphis. United Airlines (United) has hubs in Chicago (at O?Hare
International Airport), Denver, Los Angeles, San Francisco, and Washington,
D. C. (at Dulles International Airport). Major airlines provide nonstop
service to many ?spoke? cities from their hubs, ranging from large cities
like Portland, Oregon, to smaller communities such as Des Moines, Iowa, and
Lincoln, Nebraska. Depending on the size of those markets (i. e., the number
of passengers flying nonstop between the hub and the ?spoke? community), the
major airlines may operate their own large jets on those routes or use
regional affiliates to provide service to other communities, usually with
regional jet 7 and turboprop aircraft.

The airports in small ?spoke? communities are the smallest in the nation?s
commercial air system. The Federal Aviation Administration (FAA) and federal
law categorize the nation?s commercial airports into four main groups based
on the number of passenger enplanements- large hubs, medium hubs, small
hubs, and nonhubs. 8 Generally, airports in major

6 For the purposes of this report, major airlines include Alaska Airlines,
America West, American Airlines, American Trans Air, Continental Airlines,
Delta Air Lines, Northwest Airlines, Southwest Airlines, United Airlines,
and US Airways. Seven of these 10 airlines operate hub- and- spoke systems;
Southwest, Alaska, and American Trans Air do not. With a hub- and- spoke
network, carriers can combine ?local? passengers (those originating at or
destined to the hub) with ?connecting? passengers (those not originating at
or destined to the hub but traveling via the hub) on the same flight.

7 Although regional jets typically seat 32 to 70 passengers, there is no
uniformly accepted definition of a regional jet either in the industry or in
federal laws and regulations. For example, the Wendell H. Ford Aviation
Investment and Reform Act for the 21st Century, P. L. 106- 181, variously
defines a regional jet as having a maximum seating capacity of ?not less
than 30 nor more than 75? (sec. 210) or ?of less than 71? (sec. 231). Within
the industry, ?regional jet? is sometimes used to describe larger aircraft,
such as the Fokker F- 100 (107 seats) and Boeing 717 (106 seats), and older-
technology aircraft, such as the Fokker F- 28 (69 seats) and BAe 146- 100
(70- 82 seats).

8 The categories are based on the number of passengers boarding an aircraft
(enplaned) for all operations of U. S. carriers in the United States. A
large hub enplanes at least 1 percent of all passengers, a medium hub 0.25
to 0.99 percent, a small hub 0.05 to 0.249 percent, and a nonhub less than
0.05 percent. Nonhubs and small hubs are defined in 49 U. S. C. 41731;
medium hubs are defined in 49 U. S. C. 41714; and large hubs are defined in
49 U. S. C. 47134. A passenger flying from Baltimore to San Francisco who
connects to a different flight in Cincinnati counts as two passenger
enplanements. Another passenger who might fly nonstop between Baltimore and
San Francisco would represent only a single enplanement. Background

Page 6 GAO- 02- 432 Small Community Air Service

metropolitan areas- such as Chicago, New York, Tampa, and Los Angeles- are
?large hubs.? There are 31 large hubs, and they serve the majority of
airline traffic, accounting for nearly 70 percent of U. S. passenger
enplanements in 1999. At the opposite end of the spectrum are the nonhub
airports- the airports for the communities that are the focus of this study.
In all, 404 airports were categorized as nonhubs in 1999. As a group, they
enplaned only about 3 percent of passengers in 1999. Of those airports, we
analyzed 267, which are generally the largest airports within this group. 9
These included 202 small community airports in the continental United States
and 65 in Alaska and Hawaii. Table 1 provides more information about the
four airport categories, along with an illustration of each type.

9 We specifically excluded certain nonhub airport communities from our
analysis. These included airports where fewer than 2,500 passengers boarded
in 1999 (7 per day), along with those that are receiving service subsidized
in part by the Department of Transportation?s Essential Air Service (EAS)
program. For additional information on the EAS program, see U. S. General
Accounting Office, Essential Air Service: Changes in Subsidy Levels, Air
Carrier Costs, and Passenger Traffic, GAO/ RCED- 00- 34 (Washington, D. C.:
Apr. 14, 2000). We also excluded nonhub airports located in U. S.
territories and those in metropolitan areas with populations of 1 million or
more. For additional information on the scope of communities we reviewed,
see appendix I.

Page 7 GAO- 02- 432 Small Community Air Service

Table 1: Differences Among Categories of Commercial Airports Hub category
Number of

airports Percent of total

U. S. passengers enplaned 1999

Median number of passenger enplanements, 1999 Example a

Large 31 69.7 14,026,868 Lambert- St. Louis International Airport, St.
Louis, Mo. 2000 population - 2.6 million b 1999 annual enplanements - 15,
075, 992 number of carriers serving - 18 average daily flights - 560 Medium
37 19.3 3, 305,073 Kansas City International, Kansas City, Mo.

2000 population - 1.8 million b 1999 annual enplanements - 5,760,037 number
of carriers serving - 18 average daily flights - 235 Small 74 7. 8 534,218
Springfield- Branson Regional Airport, Springfield, Mo.

2000 population - 326,000 b 1999 annual enplanements - 349,320 number of
carriers serving - 5 average daily flights - 28 Non 404 3.3 20,379 Joplin
Regional Airport, Joplin, Mo.

2000 population - 157,000 b 1999 annual enplanements - 28, 877 number of
carriers serving - 2 average daily flights - 5 a Enplanement data are from
1999. Flight schedule data are for October 15- 21, 2001 (revised).

b Populations shown are for the metropolitan area. Source: GAO analysis of
enplanement data from FAA, airline schedule data from the Kiehl Hendrickson
Group, and population data from the U. S. Census Bureau.

The typical 10 community of the 202 small communities in our analysis had
nine departing flights per day in October 2000, and most had no jet service.
They were typically served by two airlines, though 41 percent had service
from only one airline. Individually, however, they varied considerably, from
having no more than 1 or 2 daily departures to having more than 60. As a
group, their limited level of service is related to their small populations.
As individual communities, the varied levels of service reflect differences
in other factors such as the level of local economic activity and proximity
to nearby airports.

10 In this report, by ?typical? community, we mean the ?median? community
where 50 percent of the communities fall above and 50 percent fall below
this midpoint community. Most Small

Communities Had Limited Air Service in October 2000

Page 8 GAO- 02- 432 Small Community Air Service

In October 2000, the typical small community among the 202 we analyzed had
the following levels of service:

 Service from two different airlines or their regional affiliates, each
providing service to a different hub where passengers could make connections
to other flights in the airline?s hub- and- spoke system. However, a
substantial minority of the communities- 41 percent- had service from only
one airline.  Nine departing flights a day, most if not all of them
turboprops rather than

jets. In all, only 67 of the 202 communities had any jet service. 11 The
level of service varied significantly from community to community. At the
higher end were airports serving resort destinations like Key West, Florida,
where five different carriers operated 44 average daily departures to six
nonstop destinations, and communities such as Fayetteville and Bentonville,
Arkansas (the headquarters for Wal- Mart Stores, Inc.), near the Northwest
Arkansas Regional Airport, where five air carriers scheduled 42 average
daily jet and turboprop departures to seven nonstop destinations. The
highest number of daily departures- 62- was for Nantucket, Massachusetts, a
resort community served by turboprop and even smaller piston aircraft. At
the other extreme were communities such as Hattiesburg, Mississippi, and
Thief River Falls, Minnesota, with an average of 3 and 1 daily departures,
respectively. In total, the 10 small communities with the most air service
typically had more than 38 scheduled departures per day, while the 10 small
communities with the least air service typically had fewer than 3 scheduled
departures per day. 12 Table 2 summarizes the range of air service that was
available at the 202 small community airports in October 2000.

11 In this report, we commonly discuss air service at small communities in
terms of average daily departures. We calculated those averages by
determining the total number of departures scheduled for the entire week and
dividing by seven. We stated daily averages then by rounding to the nearest
integer.

12 Two communities had no service scheduled during the week of October 15-
21, 2000. For purposes of discussing the range of scheduled service in this
paragraph, we excluded those communities. Most Small Communities

Had Few Flights and Few Air Carriers, Though Service Levels Varied
Considerably

Page 9 GAO- 02- 432 Small Community Air Service

Table 2: Summary of Air Service at Small Communities, October 2000 Air
service dimension Median Range a

Number of carriers 2 0- 6 Number of daily nonstop destinations 2 0- 12
Number of daily turboprop flights b 8 0- 62 Number of daily jet flights c 0
0- 26 Total number of daily flights 9 0- 62 a The lower end of the range for
each air service dimension was 0 because two communities- Marquette,
Michigan, and West Yellowstone, Montana- had no service for the week of
October 15- 21, 2000. b Includes flights by some piston aircraft at some
communities.

c Of the 202 small communities, 135 had no jet service in October 2000,
which is why the median, or midpoint, community had 0 daily jet departures.
Of those 67 small communities that had jet service, the median community had
six daily jet departures. This explains in part why the median for total
daily departures does not simply equal the sum of the daily turboprop and
jet departure median values.

Source: GAO analysis of data from the Kiehl Hendrickson Group.

For purposes of comparison, appendix II provides additional information on
key differences in the scope of air service that airlines scheduled at
nonhub and small hub communities. Small hub airports tend to serve somewhat
larger communities and have significantly more commercial air service than
do the 202 nonhub airports in the continental United States.

The most obvious reason for the generally limited level of service at these
small communities is their small size. As a whole, the 202 airports served a
small portion of the U. S. population and geographic territory. In 2000, the
median population of the 202 nonhub airport communities in our analysis was
about 120,000, and the median number of daily passenger enplanements in 1999
was about 150. 13 However, airports typically serve populations and
businesses in a larger surrounding area- typically referred to as a
?catchment area.? An airport?s catchment area is the potential geographic
area for drawing passengers. The geographic size of a catchment area varies
from airport to airport depending on such factors as how close an airport is
to other airports and whether the airport is served by a low- fare airline
(and, therefore, attractive to passengers from farther away). Catchment area
size estimates provided by the airport directors we surveyed showed that
these airports potentially serve a total population of

13 1999 is the most recent year for which enplanement data was available.
Overall Limited Level of

Service Reflects Communities? Small Size

Page 10 GAO- 02- 432 Small Community Air Service

about 35 million (about 12 percent of the continental U. S. population). 14
Figure 1 shows the location and catchments areas of the 202 airports in our
analysis as estimated by airport directors responding to our survey and our
estimates of catchment areas for those who did not respond to the survey.

Figure 1: Location and Estimated Catchment Areas of 202 Small Community
Airports in the Continental United States

Source: GAO analysis of information from the FAA, Census Bureau, and survey
of small community airport directors.

14 Our representation of airports? catchment area sizes and populations are
based on survey respondents? estimates of their own catchment areas. We then
used data from the U. S. Bureau of the Census to estimate the population for
that particular geographic territory. For airports that did not respond to
the survey, we estimated the geographic size of their catchment areas using
the mean distance for small community airports in the same geographic region
as small community airports that did respond. We then calculated the
populations for these areas using the same procedures noted above.

Page 11 GAO- 02- 432 Small Community Air Service

The small size of these markets greatly affects the level of service because
airlines? decisions about offering service are motivated primarily by the
need to maintain profitability. An airline?s profitability generally depends
on its ability to generate revenues in a given market while managing its
costs. The airline industry is capital- and labor- intensive and is largely
dependent on passenger traffic for its revenues. The airlines use
sophisticated computer models to help them identify whether certain markets
can be served profitably. 15 The limited amount of passenger traffic from
many of these communities limits the number of flights airlines provide. It
is also not surprising that turboprops have typically provided most service
to these communities, because turboprop aircraft are generally the least
expensive type of aircraft to buy and operate. 16

The role of size in limiting a small community?s service can be seen by
stratifying the small communities into population groups. As part of our
analysis, we separated the 202 small communities into three groups- those
smaller than 100,000, those with populations of 100,000 to 249,999, and
those with populations of 250, 000 and greater. As figure 2 shows, the
?smallest of the small? typically had lower median levels of service as
measured by such indicators as number of daily departures (both turboprop
and jet) and service from more than one airline.

15 These proprietary models take into account such considerations as the
carrier?s operating costs, estimated passenger traffic, and competition in
the market (including the type of aircraft competitors used, the number of
daily flights they scheduled, and the fares they charged).

16 We reported in February 2001 that of the 157 U. S. cities that air
carriers served with regional jets in October 2000, only 13 (8 percent) were
communities with populations under 100,000. Because regional jets are
generally more expensive to purchase and operate than turboprop aircraft,
and because air carriers? regional jet deployment strategies are profit
driven, the amount of high- yield business traffic is a key factor in those
deployment decisions. See U. S. General Accounting Office, Aviation
Competition: Regional Jet Service Yet to Reach Many Small Communities, GAO-
01- 344 (Washington, D. C.: Feb. 14, 2001).

Page 12 GAO- 02- 432 Small Community Air Service

Figure 2: Comparison of Air Service Indicators Across Small Communities by
Population Category, October 2000 (Median Service Levels)

Source: GAO analysis of airline schedule data from the Kiehl Hendrickson
Group.

Besides population, a variety of other factors may influence how much
service an individual community receives. In our analysis, two such factors
stood out. One of these was the level of economic activity. The airline
industry is highly sensitive to the business cycle, and its economic
performance is strongly correlated with fluctuations in personal disposable
income and gross domestic product. When the economy is growing, the demand
for air transportation grows, allowing carriers to raise yields (prices) and
profitability. When the economy falls into recession, unemployment grows,
individuals postpone discretionary Variation in Air Service Is

Related to Level of Local Economic Activity

Page 13 GAO- 02- 432 Small Community Air Service

travel, and airline yields and profitability decline. 17 In particular, a
key element to the profitability of an airline?s operation in a given
location is the availability of high- yield passenger traffic- that is,
business travelers who are more likely to pay higher airfares than leisure
travelers. Communities with greater amounts of local business activity may
have more (and different) air service than communities with less economic
activity. Of course, the reverse may also be true- that local economic
activity cannot improve without enhanced air service. Thus, the presence or
absence of air service may also positively or negatively affect local
economic activity, rather than local economic activity dictating the amount
and type of air service.

Our analysis showed a statistically valid relationship between the economic
characteristics of small communities and the amount of air service that they
received. Economic principles led us to expect that passenger demand for air
service would be greater in communities with more jobs and higher incomes.
Our results were consistent with these expectations. Larger communities with
more income and ?regional product? 18 had service from more major carriers
and had more weekly departures. 19 For example, for every additional 25,000
jobs in a county, a community received 4.3 more jet departures per week and
4.8 more turboprop departures per week. Similarly, for every additional
$5,000 in per capita income, a community received 3.3 more jet departures
per week and 12.7 more turboprop departures per week. 20 In other words, if
two small communities, A and B, were similar except that Community A had

17 For an overview of the relationship between overall economic activity and
the demand for air transportation, see, for example, Paul Stephen Dempsey
and Laurence E. Gesell,

Airline Management: Strategies for the 21st Century, Coast Aire
Publications, 1997. 18 Regional product is a concept similar to gross
domestic product (i. e., the output of goods and services produced by labor
and property located in the United States), only measured at the regional
level.

19 We obtained the data used to measure the characteristics of a small
community from the Regional Economic Information System database, which is
produced by the Bureau of Economic Analysis. These data are somewhat
imperfect measures of economic activity at individual communities, in part
because they are available only by county, not city. We defined a
community?s size in terms of both total full- and part- time employment and
population, and we defined a community?s wealth as per capita income. We
used manufacturing earnings as a proxy measure of regional product. We
defined the level of service received by small communities as both the
number of departures per week (jet and turboprop aircraft) and as the number
of major carriers serving a community. See appendix V for a more detailed
discussion of this analysis.

20 These results are statistically significant at the 95- percent level.

Page 14 GAO- 02- 432 Small Community Air Service

$5,000 more in income per capita than Community B, Community A would have
had 16 more departures per week than Community B. 21

A third main factor that stood out as helping to explain the variation in
service levels between small communities, in addition to relative size and
economic activity, was the community?s relative proximity to larger
airports. If a small community is located within relatively close driving
distance of another commercial airport, passengers may drive to the other
airport, rather than fly to or from the local community airport. This
tendency to lose passengers to other airports is referred to as ?leakage.?
Of the 202 small communities in our study, 94 (47 percent) are within 100
miles of an airport that is served by a low- fare airline 22 or that serves
as a hub for a major airline. 23 Figure 3 shows circles with 100- mile
radiuses around those hub or low- fare carrier airports, and thus the number
of small communities that are within 100 miles of those alternative
airports. As figure 3 also shows, the concentration of nonhub airports that
are close to larger airports is much greater east of the Mississippi River
than west of the river. Over 70 percent of the eastern small communities are
within 100 miles of a hub or low- fare airport, compared with 26 percent of
the western small communities.

21 The regression model holds other factors constant between the
hypothetical communities A and B: population, manufacturing earnings, and
distance from an airport served by a low- fare carrier. See appendix V for
additional information.

22 We adopted the Department of Transportation?s (DOT) definition of a low-
fare airline and included AirTran, American Trans Air, Frontier, JetBlue,
Southwest, Spirit, and Vanguard. 23 Distance between airports was measured
in statute (air) miles. Driving distances are thus greater. We used a 100-
mile radius to approximate the distance that passengers may travel to reach
an alternative airport because officials from one large low- fare carrier
have reported drawing passengers from this distance. We also examined the
number of small community airports within 50 statute miles of one of these
alternative airports. We used 50 statute miles to approximate 70 highway
miles, which was the statutory minimum distance used in the EAS program.
Small community airports within 70 highway miles from the nearest medium or
large hub airport were traditionally ineligible for subsidized air service.
Of the 202 small community airports, 10 percent are within 50 statute miles
of an airline hub or airport served by a low- fare carrier. Variation Is
Also Related to

Proximity to Larger Airports

Page 15 GAO- 02- 432 Small Community Air Service

Figure 3: Proximity of Small Community Airports to Other Airports Either
Served by a Low- fare Airline or Serving as a Major Airline?s Hub

Source: GAO analysis.

Our survey of small community airport officials confirmed the likely effect
of being close to alternative airports. When asked whether they believed
local residents drove to another airport for airline service (prior to
September 11), over half of them said that they believed this occurred to a
great or very great extent. Eighty- one percent of them attributed the
leakage to the availability of lower fares from a major airline at the
alternative airport. According to the results of our survey of airport
directors, more small community airports that are closer to other larger

Page 16 GAO- 02- 432 Small Community Air Service

airports experience a greater extent of passenger leakage than small
community airports that are farther away from other larger airports. 24

We were not able to obtain and analyze certain key data that might explain
in detail why passengers might opt to use different airports rather than
their local facility. In particular, we were not able to obtain information
on the differences in airfares among competing airports. 25 However, prior
GAO reports indicate that fares at small community airports tend to be
higher than fares at larger airports. 26 While choosing to drive to other
airports in the vicinity that offer service from other airlines may allow
passengers to gain the flight options and fares they want or need- a clear
benefit to the individual traveler- it likely affects the local community?s
ability to attract or retain other competitive air service. In addition,
there may be other factors that influence the amount and type of service
that air carriers can provide at small communities. As agreed, we intend to
examine possible approaches to enhancing air service to these communities in
a subsequent report.

24 Of airport managers at communities more than 100 miles from another
airport, 68 percent report that passengers drive to the other airport. Of
airport managers at communities less than 100 miles from another airport, 76
percent report that passengers drive to the other airport.

25 Data on airfares for the quarter including October 2001 would not be
available from the Bureau of Transportation Statistics until late February
or early March 2002. Additionally, there is a lack of complete and
representative fare data for small communities, especially for local
passengers who do not connect to large carrier services. This is because
public data on airfares is developed from a 10- percent sample of tickets
collected from large air carriers, which comprises DOT?s ?Passenger Origin-
Destination Survey? (O& D Survey). Small certificated air carriers and
commuter carriers do not participate in the O& D Survey. Thus, there are
inherent statistical sampling limitations in the O& D Survey data.

26 See, for example, U. S. General Accounting Office, Airline Deregulation:
Changes in Airfares, Service Quality, and Barriers to Entry, GAO/ RCED- 99-
92 (Washington, D. C.: Mar. 4, 1999) and Department of Transportation,
Domestic Airline Fares Consumer Report: Third Quarter 1998 (Special Feature:
Fare Premiums by City), (Washington, D. C.: April 1999).

Page 17 GAO- 02- 432 Small Community Air Service

Between October 2000 and October 2001 (revised), the number of total daily
departures in small communities dropped by 19 percent. Airlines planned part
of these decreases before September 11 (a 6 percent reduction) but made even
steeper reductions (13 percent) afterward. In 36 communities, at least one
of the airlines providing service withdrew entirely from the market, with
most of these withdrawals coming before September 11. The number of
communities with service from only one airline grew by 12, raising the
percentage of communities with one- airline service to 47 percent. While
many communities lost service, carriers initiated service at 14 communities.
Nearly all of these gains occurred prior to September 11.

Airlines substantially reduced total scheduled departures at the 202 small
communities we reviewed between October 2000 and October 2001. As figure 4
shows, airlines scheduled an average of 2,406 departures daily during the
week of October 15- 21, 2000. In their original schedules for the week of
October 15- 21, 2001 (that is, the schedules prepared before September 11),
airlines had planned to operate an average of 2,257 departures per day, a
reduction of 6 percent from the October 2000 level. Airlines made further-
and sharper- service reductions following September 11. According to our
analysis of the airlines? revised schedules for October 2001, the average
number of scheduled daily departures from smaller communities dropped to
1,937, or about 320 (13 percent) fewer departures than originally planned.
Combined, these schedule changes amounted to a total reduction of about 19
percent from October 2000?s flight schedule. The median community in our
group of 202 had nine daily departures in October 2000. After the combined
drop that was planned both before and after September 11, the median
community had six daily departures in October 2001. 27

27 To determine if the November 2001 schedule showed any significant change
from the revised October 2001 level, we analyzed the airlines? schedules for
that month as well. Our analysis showed that the total number of daily
departures for November 2001 was essentially the same as that scheduled for
October 2001. We recognize that the airlines make frequent adjustments to
their schedules and that overall service levels may have changed since then.
Airline Service at

Small Communities Declined Between October 2000 and October 2001

Air Service Levels in Small Communities Declined Both Before and After
September 11 Attacks

Page 18 GAO- 02- 432 Small Community Air Service

Figure 4: Change in Total Daily Departures for Small Communities, October
2000- October 2001

Source: GAO analysis of airline schedule data from the Kiehl Hendrickson
Group.

Other industry data regarding service decreases was consistent with the
decreases identified in our 202 small communities. According to one industry
analysis, the changes in daily scheduled seats from U. S. airports were
generally comparable across airports of all sizes. Small hubs experienced a
greater relative decrease in service (- 15. 5 percent) compared to nonhubs
(- 13.5 percent). Large hubs had the greatest relative decrease in total
available seats (- 16.3 percent), and medium hubs had a smaller decrease (-
12. 4 percent).

The 19- percent drop in average daily departures came almost exclusively on
turboprop flights. In October 2000, 67 of the 202 communities had some jet
service. Airlines tended not to reduce jet flights in those locations where
they already were in place. Overall, there were slightly more jet departures
in October 2001 than in October 2000. As table 3 shows, median daily
turboprop departures dropped from 8 to 6 before September 11 and from 6 to 5
afterward. 28 On other service measures- number of airlines

28 According to industry sources, some of the decline in turboprop flights
and gain in jet flights can be attributed to strategies that carriers
adopted in recent years to phase out turboprop aircraft and replace them
with regional jets. For example, Atlantic Coast Airlines, Inc., which
operates as a United Express and as a Delta Connection carrier, is planning
to become an ?all- jet? carrier by the end of 2003. More specific
information about this strategy is discussed later in this report.

0 500

1,000 1,500

2,000 2,500

October 2000 October 2001 (original) October 2001 (revised) Total departures

Page 19 GAO- 02- 432 Small Community Air Service

providing service and number of nonstop departures- the median for these
communities remained the same.

Table 3: Range of Service Levels Among Small Community Airports, October
2000- October 2001 October 2000 October 2001 (original) October 2001
(revised) Air service dimension Median Range a Median Range a Median Range a

Number of carriers 2 0- 6 2 0- 6 2 0- 5 Number of daily nonstop destinations
2 0- 12 2 0- 12 2 0- 12 Number of daily turboprop departures 8 0- 62 6 0- 66
5 0- 62 Number of daily jet departures b 0 0- 26 0 0- 32 0 0- 29 Total
number of daily departures 9 0- 62 8 0- 66 6 0- 62

a Some communities had no service scheduled for the week we analyzed; thus,
all of the low- range numbers are zeros. b Of the 202 small communities, 135
had no jet service in October 2000, which is why the median, or midpoint,
for daily jet departures is 0. Of those 67 small communities that had jet
service, the median value for the number of daily jet departures was 6. This
explains, in part, why the median for total daily departures does not simply
equal the sum of the daily turboprop and jet departure median values.

Source: GAO analysis of airline schedule data from the Kiehl Hendrickson
Group.

While the typical small community had the same number of airlines- two-
providing service both in October 2000 and October 2001, a number of
communities gained or lost a carrier. In the aggregate, the movement was
downward, with 36 communities experiencing a net decline in the number of
airlines providing service and 14 communities experiencing a net increase.
For the 36 communities that lost airlines, most lost them as a result of
airline decisions made prior to September 11. Likewise, communities that
experienced net gains in the number of carriers did so primarily as a result
of airline decisions made before September 11. Among communities that lost
airlines, two (Cumberland, Maryland, and Rockford, Illinois) lost service
altogether. 29

The overall effect of these gains and losses was a decrease in the number of
communities served by four, three, and two airlines and an increase in the
number of communities served by only one airline (see fig. 5). In all, the
number of communities served by only one airline increased from 83 (41
percent) to 95 (47 percent) of the 202 communities in our review.

29 Cumberland, Maryland, has since received service from a different carrier
under a statefunded subsidy program. Communities Saw Net

Decline in Number of Airlines Providing Service

Page 20 GAO- 02- 432 Small Community Air Service

Figure 5: Change in the Number of Carriers Serving Small Communities,
October 2000- October 2001

Source: GAO analysis of airline schedule data from the Kiehl Hendrickson
Group.

At a minimum, communities that lost an airline were at risk of losing
connecting service to some destinations. Of the 36 communities that lost
service from an airline, 5 lost the services of a carrier but did not lose
access to other destinations, but 31 lost connecting service to other
destinations. For example, when Abilene, Texas, lost its service from one of
the two airlines that had been providing service, it lost one- stop
connections to 14 destinations. Excluding the 2 communities that lost all
service, the other 34 communities lost an average of 12 one- stop
connections when one of the carriers discontinued flight operations there.

Page 21 GAO- 02- 432 Small Community Air Service

Service changes at Lake Charles, Louisiana, illustrate what happens when an
airline withdraws from a market. In October 2000, Continental Express and
American Eagle both served Lake Charles Regional Airport. Continental
Express was the dominant carrier, providing 40 weekly flights (57 percent of
the total capacity, as measured by the number of available seats on
departing flights). After September 11, American Eagle discontinued its 27
weekly flights from Lake Charles to Dallas- Fort Worth. Continental Express
continued to fly, offering 38 weekly flights (2 fewer) to Houston. The loss
of American Eagle?s service also meant that Lake Charles?s passengers could
no longer reach 13 other destinations via onestop connections at Dallas-
destinations that Continental did not serve.

It is difficult to assess the effect of losing a carrier on competition at a
particular community. For one thing, the number of carriers providing
service to a small community is an imperfect measure of competition. In the
airline industry, competition is normally defined in terms of the number of
different carriers serving the same city- pair market- that is, the route
between the same two cities. Most small communities that received service
from two or more carriers had nonstop flights to two or more airlines? hubs.
In the nonstop markets between the small community and those hubs, there was
probably little direct competition initially; passengers, therefore,
experienced little if any loss of competition on those routes if one of the
carriers discontinued service. 30 However, if the passenger?s final
destination was not an airline hub city, then different airlines may compete
directly in offering connecting service to the same destination city but
through their different respective hubs. In such cases, the loss of an
airline?s service at a small community means the loss of a competitive
choice. Where competition is lost, the risk that consumers may be subject to
higher airfares increases. 31

30 An airline may be able to carry passengers to another airline?s hubs with
connecting service over the first airline?s hubs. That connecting service
may exert some competitive restraint on prices. However, connecting service
does not always represent an adequate substitute for nonstop service,
especially for business travelers.

31 A number of studies- including our own- have shown that markets with
fewer competitors, especially those dominated by a single carrier, have
higher fares. See, for example, Steven A. Morrison, ?New Entrants, Dominated
Hubs, and Predatory Behavior, ? Statement before the Subcommittee on
Antitrust, Business Rights, and Competition, Committee on the Judiciary,
United States Senate (Apr. 1, 1998). See also U. S. General Accounting
Office, Airline Competition: Higher Fares and Less Competition Continue at
Concentrated Airports, GAO/ RCED- 93- 171 (Washington, D. C.: July 15,
1993). And as noted earlier, prior GAO reports indicate that fares at small
community airports tend to be higher than fares at larger airports.

Page 22 GAO- 02- 432 Small Community Air Service

Two primary external events that occurred since October 2000- the economic
decline that began in early 2001 and the collapse of airline passenger
traffic after September 11- significantly affected carriers? financial
conditions and thus influenced decisions about service throughout their
networks, including service to small communities. As the nation?s economic
performance declined, fewer passengers opted to fly. Consequently, airline
revenues dropped, and airlines sought ways to control costs. They did so, in
part, by reducing scheduled operations. In many small communities, they
reduced the number of flights they were providing, and in some communities
where they had a small portion of the market, they pulled out altogether.
After September 11, passenger traffic and revenue plummeted, exacerbating
the situation. Beyond their reactions to the economic slowdown and the
events of September 11, airlines also made some changes on the basis of
long- range decisions about the composition and deployment of their fleets-
decisions, generally, to reduce turboprop operations and increase regional
jet service. Some communities lost service as carriers retired certain types
of small aircraft.

Nationally, the U. S. economy slowed down during 2001 and moved into its
first recession (as defined by the National Bureau of Economic Research)
since 1991. This change in the national economy is reflected in airline
passenger and revenue data. In the latter parts of 2000, monthly airline
passenger traffic and revenue were still growing compared with the same
periods in 1999. But beginning in February 2001, passenger traffic generally
declined. Additionally, reflecting a drop in high- yield business traffic,
total passenger revenues decreased at a steeper rate than passenger traffic,
as shown in figure 6. For the U. S. airline industry as a whole, data from
the Bureau of Transportation Statistics (BTS) indicate that airlines? net
income turned negative in the second quarter of 2001. Simultaneously across
the industry, airline costs were rising. Carriers began efforts to control
costs, in part by reducing service. As the economy slowed down, industry
analysts projected that U. S. commercial airlines would lose over $2 billion
in 2001. National Economic

Decline and September 11 Attacks, Along with Airlines? Strategic Decisions,
Underlie Shifts in Air Service

Broad National Economic Decline and September 11 Attacks ad Significant
Effects on Airline Service Levels Nationwide

Page 23 GAO- 02- 432 Small Community Air Service

Figure 6: Change in Total U. S. Passenger Enplanements and Revenue,
September 2000- August 2001

Source: GAO presentation of data from the Air Transport Association.

The events of September 11 accelerated and aggravated negative financial
trends already evident in the airline industry. In response to significant
losses experienced by the carriers stemming from the temporary shutdown of
the nation?s airspace and the drop in passenger traffic, the president
signed the Air Transportation Safety and Stabilization Act, which provided
up to $4. 5 billion in emergency assistance to compensate the nation?s
passenger air carriers for these losses. 32 The change in the airlines?
financial condition may be attributable to both the continued deterioration
of passenger revenues and the inability of airlines to cut their expenses
proportionately. Figure 7 shows the significant drop in passenger traffic
after September 11. Data from BTS indicate that passenger enplanements
between September 2000 and September 2001 on large air carriers dropped by
over 34 percent nationally.

32 P. L. 107- 42. The act also provided up to $500 million in grants to all-
cargo airlines.

Page 24 GAO- 02- 432 Small Community Air Service

Figure 7: Passenger Traffic and Revenues Before and After September 2001

Source: GAO presentation of data from the Air Transport Association.

As passenger traffic and revenues plummeted, carriers? efforts to control
costs included significant reductions in total capacity- in other words,
service reductions. These reductions were dramatic. According to data from
BTS, carriers flew 20 percent fewer departures in September 2001 than in
September 2000. Different airlines approached such cost- cutting in
different ways. For example, US Airways retired 111 older aircraft from its
fleet, eliminating its Boeing 737- 200s, MD- 80s, and Fokker F- 100s. Some

Page 25 GAO- 02- 432 Small Community Air Service

carriers also replaced service from their large mainline jets with smaller
aircraft operated by regional affiliates to better match capacity with
passenger demand, as United did in some markets. In addition, United reduced
the total number of departures in its system from about 2,400 before
September 11 to 1,654 by October 31, in part by reducing early morning and
late evening flights. Service to smaller communities was affected as part of
the overall decrease in operations.

These two factors- the economic downturn and aftermath of September 11-
played out in small communities as well as in larger markets. As with the
nation as a whole, small communities saw dramatic decreases in passenger
traffic. According to our survey of airport officials, passenger traffic at
small communities fell by 32 percent between September 2000 and September
2001- about the same percentage that, according to BTS data, passenger
traffic decreased throughout the country. Over 80 percent of the airport
managers we surveyed reported that passenger fear (that is, general
apprehension related to the events of September 11, 2001) was a key factor
in decreased enplanements at their airport since September 11. Airport
directors also reported that passenger enplanements dropped because of air
carrier service changes (e. g., fewer departures, smaller aircraft, or fewer
carriers). In addition, managers indicated that basic economic conditions
and post- September 11 airport security requirements reduced enplanements.
Thus, the general reductions in service that occurred at small communities
can be seen as reflecting airlines? overall response to these factors.

Another way that these factors can be seen at work in small communities is
in the decisions airlines made to withdraw from a community. In most cases,
when an airline withdrew entirely from a community, it was a community in
which the airline was competing with other airlines and had only a limited
market share. More specifically, of the 36 small communities that lost a
carrier between October 2000 and October 2001, there were only six instances
in which the carrier that discontinued operations was the largest service
provider at the community.

The effect of these decisions to withdraw from multiple- carrier markets can
be seen in one characteristic we observed in the airline schedule data we
analyzed: Among the 202 communities we analyzed, service reductions tended
to be greater in those communities with populations above 100,000 than in
communities with populations below 100,000. This was true across several
types of service indicators, such as number of carriers, total number of
daily departures, and number of nonstop flights to more than Service
Decisions for

Small Communities Reflect Airlines? Attempts to Adjust to the National
Downturn

Page 26 GAO- 02- 432 Small Community Air Service

one destination. Across all these indicators, communities with populations
below 100,000 typically had lower levels of service than their larger
counterparts both in October 2000 and October 2001, but compared with these
larger communities, they lost less of that service during the 1- year period
we measured. One reason may be that over half of small communities with
populations less than 100,000 were served by only one airline, both in
October 2000 and in October 2001. Thus, airlines? decisions to withdraw from
multiple- carrier markets had little effect on them.

While the economic downturn and the events of September 11 were potent
factors in shaping airline service to small communities, some of the changes
that were occurring reflected airline efforts on other fronts. The number of
departures or available seating capacity at some small community airports
changed when some major airlines directed their regional affiliates to shift
some of their aircraft fleets to operate at different hubs in their systems.
33 Similarly, changes in the number of departures or available seating
capacity at some small community airports reflected strategic decisions that
carriers had made about the composition and deployment of their fleets-
decisions to replace their turboprop aircraft with regional jet aircraft.
These decisions were made with the concurrence of the regional carriers?
mainline partners. Three examples illustrate how such restructurings often
affected service to some small communities.

According to a Northwest official, the carrier began restructuring parts of
its Northwest Airlink regional fleet in 2002. Northwest began retiring
turboprops at its wholly- owned affiliate, Express Airlines I, while
increasing the number of regional jets in that carrier?s fleet and deploying
them at all three of its hubs- Detroit, Minneapolis/ St. Paul, and Memphis.
Northwest decided that its other regional carrier, Mesaba Airlines, would
become the sole operator of turboprops at its hubs beginning in February
2002. Mesaba also operates 69- seat regional jets.

33 The relationship between the mainline carrier and regional affiliate can
take several forms, varying from outright ownership by the major airline, to
partial ownership by the major airline, to a marketing alliance devoid of
any ownership by the major airline. For example, American Eagle airlines is
a wholly owned regional carrier subsidiary of AMR Corp. (the parent company
of American Airlines), Mesaba Airlines is partially owned by Northwest, and
United contracts with several airlines to operate as its regional carrier,
United Express, in certain markets. Internal Airline Strategies

Dictated Other Schedule Changes

Northwest?s Restructuring with Regional Affiliates

Page 27 GAO- 02- 432 Small Community Air Service

According to our analysis, between October 2000 and October 2001, Express
Airlines I and Mesaba altered service at 60 small communities. Overall, more
small communities lost service than gained service from these carriers
during this period. A total of 49 small communities lost some capacity (e.
g., through a reduction in flight frequency or use of smaller aircraft) from
these carriers, with four of them losing service from Express Airlines I and
Mesaba entirely because, according to a Northwest official, they were no
longer profitable. 34 On the other hand, 11 communities gained service- 9 of
them gaining additional flights or extra capacity through larger aircraft,
and 2 gaining start- up service from the two airlines. Appendix VI provides
more information about the small community service changes made by Express
Airlines I and Mesaba between 2000 and 2001.

Service changed at some communities when United renegotiated the contract
with one of its regional carriers- Great Lakes Aviation. In 2000, both Great
Lakes and Air Wisconsin served as United Express carriers operating between
United?s Chicago and Denver hubs. However, beginning in May 2001 under a
revised contract, Great Lakes no longer operated as a United Express carrier
and instead continued in a ?codesharing? 35 relationship with United. Under
this new arrangement, Great Lakes could decide which markets it served, but
United was free to decide whether or not to codeshare on those routes.
Furthermore, United expanded the amount of service Air Wisconsin (as a
United Express carrier) provided to many of these communities.

Of the 202 communities in our study, 16 were served by Great Lakes, 15 by
Air Wisconsin, and 9 by both. Between October 2000 and October 2001,
United?s changes altered air service between 39 of the 40 communities served
by one of these carriers. Of the 39 communities with service changes, Great
Lakes pulled out completely from 4. Either Great Lakes or Air Wisconsin
decreased capacity at another 30 communities. Five

34 Of the four communities that lost service from Express Airlines I and
Mesaba, Northwest continued to serve one (Bismarck, North Dakota) with
mainline operations. 35 Codesharing allows an airline to sell seats on its
partner?s plane as if they were its own, enabling the airline to expand its
route network without adding any planes. For example, if United and Great
Lakes have a codesharing agreement and United flies from Chicago to Denver
(and Great Lakes does not), and Great Lakes flies from Denver to Cody,
Wyoming (and United does not), then United could sell tickets from Chicago
to Cody as its own flight, and the computer reservation system would
indicate that United provides seamless (? on- line?) service to these
cities. United?s Contract

Renegotiation with One of Its Regional Affiliates

Page 28 GAO- 02- 432 Small Community Air Service

communities gained new service or additional capacity. Appendix VII provides
more information about the communities and how they were affected.

According to industry sources, some of the decline in turboprop flights and
gain in jet flights can be attributed to strategies that some carriers
adopted in recent years to phase out turboprop aircraft and replace them
with regional jets.

For example, Atlantic Coast Airlines, Inc., which operates as a United
Express and as a Delta Connection carrier, is planning to become an ?all-
jet? carrier by the end of 2003. In October 2000, Atlantic Coast operated 87
aircraft, including 34 regional jets and 53 turboprops, to 51 destinations
from Washington Dulles and Chicago O?Hare. Of those 51 markets, 21 were
served exclusively with turboprops. In October 2001, Atlantic Coast operated
117 aircraft, including 81 regional jets and 36 turboprops, to 60
destinations. Of those 60 markets, 15 were served exclusively with
turboprops. By December 2001, Atlantic Coast had retired all of its 19- seat
turboprop aircraft and ended service to two small communities- Lynchburg and
Shenandoah Valley, Virginia- when it did so. Other regional carriers, such
as American Eagle and Continental Express, have also decided to become ?all-
jet? carriers.

It is not surprising that most small communities have fewer carrier options
and less competition than larger communities. The economics of airline
operations- that is, the need to cover the cost of operating turboprop or
jet service with sufficient passenger revenue- mean that small communities
that generate relatively little passenger traffic make profitable operations
difficult. Because small communities generate relatively little passenger
traffic (especially high- fare business traffic), they tend to have more
limited air service than larger communities. As a result, passengers who use
these communities? airports often have less service: fewer nonstop flights
to fewer destinations.

The declines in air service at small communities in 2001 generally
paralleled declines at larger airports. However, because small community
airports had much more limited service initially, such decreases may subject
passengers to or from those communities to significant effects. For example,
when small communities lose a competitive air carrier choice, they may lose
access to many destinations through one- stop connecting service. Similarly,
although we were unable to analyze how airfares changed when the number of
carriers serving a community changed, Some Regional Carriers Adopt

?All- Jet? Strategies Concluding Observations

Page 29 GAO- 02- 432 Small Community Air Service

travelers to or from those communities that lost service from one or more
carriers may be more vulnerable to noncompetitive pricing and service
patterns.

The number of communities subject to this vulnerability increased during
2001. Because of the relationship between economic activity and air service,
airlines may restore some air service at small communities when local
economic conditions improve. However, trends in the industry- such as the
replacement of some turboprop aircraft with regional jets- may make it
increasingly difficult for air carriers to operate competitive and
profitable air service to some small communities.

We provided a copy of the draft report to DOT for review and formal comment.
We also provided sections of our draft report for technical comment to
Northwest Airlines, United Airlines, Great Lakes Aviation, and Air
Wisconsin. Officials with DOT and the airlines offered only technical
comments, which we incorporated into the report, as appropriate.

We are sending copies of this report to the Honorable Norman Y. Mineta,
secretary of transportation; United Airlines; Northwest Airlines; the
Regional Airline Association; and other interested parties. We will also
send copies to others upon request.

If you or your staffs have any questions about this report, please contact
me, HeckerJ@ gao. gov, or Steve Martin at (202) 512- 2834, MartinS@ gao.
gov. Other key contributors to this report are listed in appendix VIII.

JayEtta Z. Hecker Director, Physical Infrastructure Issues Agency Comments

Page 30 GAO- 02- 432 Small Community Air Service

List of Congressional Requesters The Honorable John Rockefeller, IV
Chairman, Subcommittee on Aviation Committee on Commerce, Science, and
Transportation United States Senate

The Honorable Don Young Chairman The Honorable James Oberstar Ranking
Minority Member Committee on Transportation and Infrastructure House of
Representatives

The Honorable John Mica Chairman The Honorable William Lipinski Ranking
Minority Member Subcommittee on Aviation Committee on Transportation and
Infrastructure House of Representatives

The Honorable Olympia Snowe The Honorable Ron Wyden United States Senate

The Honorable John Peterson House of Representatives

Appendix I: Objectives, Scope, and Methodology

Page 31 GAO- 02- 432 Small Community Air Service

This report examines the changing air service conditions in small
communities. Our work focused on three objectives: (1) describing the
overall level of air service at the nation?s small communities in 2000 and
the main factors that contributed to that service level; (2) examining how
the nature and extent of air service changed among the nation?s small
communities in 2001, including a specific accounting for how service changed
after the September 11 terrorist attacks; and (3) identifying key factors
that have influenced these air service changes.

To analyze the overall level of service in 2000 and how the nature and
extent of air service at small communities changed in 2001, we first defined
the universe of small communities. We began by including all nonhub and
small hub airports, which various statutes define as small communities. 36
We then narrowed that definition by including only those nonhub and small
hub airports included on the Air Carrier Activity Information System (ACAIS)
that supports the Federal Aviation Administration?s (FAA) Airport
Improvement Program (AIP) entitlement activities. The ACAIS database
contains data on cargo volume and passenger enplanements submitted by air
carriers to the Department of Transportation (DOT).

The ACAIS database categorizes airports by the number of annual
enplanements. According to a DOT official, there are three categories:

 Primary: Public airports with scheduled, commercial air service with at
least 10,000 annual enplanements. These airports are eligible for a minimum
entitlement AIP funding of between $650,000 and $1 million. 37  Nonprimary:
Public airports with scheduled, commercial air service with

annual enplanements between 2,500 and 9,999. These airports are not 36 For
example, the Wendell H. Ford Aviation Investment and Reform Act for the 21st
Century (AIR- 21), P. L. 106- 181, defines small communities as including
both nonhub and small hub community airports. The categories of airports-
large hub, medium hub, small hub, and nonhub- are defined by statute.
Nonhubs and small hubs are defined in 49 U. S. C. 41731; medium hubs are
defined in 49 U. S. C. 41714; and large hubs are defined in 49 U. S. C.
47134. The categories are based on the number of passengers boarding an
aircraft (enplaned) for all operations of U. S. carriers in the United
States. A large hub enplanes at least 1 percent of all passengers, a medium
hub 0.25 to 0.99 percent, a small hub 0.05 to 0.249 percent, and a nonhub
less than 0.05 percent. In 2000, there were a total of 546 airports: 31
large hubs, 37 medium hubs, 74 small hubs, and 404 nonhubs. The Federal
Aviation Administration (FAA) sometimes defines hubs as geographic areas
rather than as airports. In this report, however, when we discuss hubs, we
are referring to airports.

37 49 U. S. C. 47114( c)( 1)( B). Appendix I: Objectives, Scope, and

Methodology

Appendix I: Objectives, Scope, and Methodology

Page 32 GAO- 02- 432 Small Community Air Service

eligible for AIP entitlement funds, but are eligible for ?commercial service
funds,? which are discretionary AIP funds.  Other: Airports that have
scheduled service, but not necessarily

commercial service and have less than 2,500 enplanements. To limit the scope
of our research, we included only those airports that had more than 2, 500
annual enplanements (approximately 7 passengers enplaning per day) in 1999.
From this list, we eliminated airports that were located in territories,
those at which commercial service was subsidized through DOT?s Essential Air
Service (EAS) program as of July, 2001, 38 those for which our data
indicated that carriers had scheduled no service at any time between June
2001 and July 2002, and those nonhub airports that were located in
metropolitan areas with populations of one million or greater (e. g., Meigs
Field in Chicago). We eliminated the latter group of airports because
travelers in those metropolitan areas are not limited to air service from
the small airport; rather, they have a choice of other larger airports in
the immediate area.

We then compared various aspects of air service at the nonhub and small hub
airports to see if there was a significant difference between the two. Based
on that analysis and agreement with the requesters? staffs, we defined small
communities as those served by nonhub airports that met the above- mentioned
conditions. Table 4 summarizes the number of nonhub airports affected by
each of these filters.

38 We agreed with requesting members? staffs to exclude communities that
received subsidized service from the Essential Air Service (EAS) program
because of the inherently different situation with commercial air service in
those locations. The Airline Deregulation Act of 1978 guaranteed that
communities served by air carriers before deregulation would continue to
receive a certain level of scheduled air service. Under the EAS program, the
federal government subsidizes air carriers to provide service to certain
small communities. For additional information on the EAS program, see U. S.
General Accounting Office,

Essential Air Service: Changes in Subsidy Levels, Air Carrier Costs, and
Passenger Traffic, GAO/ RCED- 00- 34 (Washington, D. C.: Apr. 14, 2000). In
addition, GAO is initiating a separate review of the EAS program in early
2002.

Appendix I: Objectives, Scope, and Methodology

Page 33 GAO- 02- 432 Small Community Air Service

Table 4: Number of Nonhub Airports Eliminated by Various Selection Criteria
Selection criteria Nonhub airports

Initial number of airports 404 Airports in U. S. territories (12) Airports
had no scheduled service from June 2001 to July 2002 (41) Airports receiving
EAS- subsidized service a (61) Airports in metropolitan areas of 1 million
or more a (23)

Total airports analyzed 267

a Some airports fell into multiple categories and thus may have already been
excluded from an earlier filter. We did not double count airports. Source:
GAO?s analysis.

As part of our analysis, we also grouped the nonhub airports based on the
size of the surrounding areas? populations. Because many of these airports
are within metropolitan statistical areas (MSAs), we used those population
totals. 39 If an airport was not located within an MSA, we used the county
population.

To determine what overall level of service airlines provided at the nation?s
small communities in 2000, we examined air service schedules published by
the airlines for the week of October 15- 21, 2000. As with our previous
reports on changes in air fares and service, 40 the types of service we
focused our analysis on were: the number of carriers serving the airport, if
the airport was dominated by a single carrier, 41 the number of nonstop
destinations served out of the airport, the number of hubs served out of the
airport, the number of turboprop and jet departures per week out of the
airport, and the types of aircraft serving the airport. We determined these
air service dimensions using airline flight schedule information submitted
by all U. S. airlines that we purchased from the Kiehl

39 An MSA is a geographic entity designated by the federal Office of
Management and Budget for use by federal statistical agencies. In general,
an MSA is a metropolitan area with a population of 100,000 or more, often
defined in terms of counties, except in New England, where MSAs are defined
in terms of county subdivisions (primarily cities and towns).

40 See, for example, U. S. General Accounting Office, Airline Deregulation:
Changes in Airfares, Service Quality, and Barriers to Entry, GAO/ RCED- 99-
92 (Washington, D. C.: Mar. 4, 1999). A more complete list of related GAO
Products follows appendix VIII.

41 A dominated airport was one in which a single carrier provided more than
50 percent of the total capacity at the airport (measured in terms of seats
available for sale).

Appendix I: Objectives, Scope, and Methodology

Page 34 GAO- 02- 432 Small Community Air Service

Hendrickson Group, an aviation consulting firm. We did not independently
assess the reliability of the Kiehl Hendrickson Group?s data, which it
purchases from another vendor, Innovata, LLC. According to the Kiehl
Hendrickson Group, Innovata employs numerous proprietary quality assurance
edit checks to ensure data integrity.

To determine factors associated with those service levels, we reviewed
available literature on air service and local economic development, and we
interviewed industry officials, consultants, academic experts, and airport
officials. Based on that information, we identified a number of factors that
relate air service levels with various aspects of small communities. Among
the elements identified were the population of those small communities and
the proximity of small community airports to other larger airports, many of
which served either as a hub for a major airline or which was served by a
low- fare carrier. 42 We obtained community population data from the U. S.
Bureau of the Census. In addition, we asked the airport directors at the
small community airports to estimate the size of their airport?s ?catchment
area.? An airport?s catchment area is the geographic area from which it
draws passengers. For those who did not respond to the survey, we estimated
the size of their catchment areas based on the average size of the catchment
area for other small community airports in the same geographic region. We
then calculated the total population living within the catchment areas using
2000 census tract population data. For each small community airport, we also
identified the nearest major airline hub facility and nearest airport served
by a low- fare carrier and determined the distance between those airports to
the small community airport. We statistically analyzed the extent to which
some of the identified factors contributed to overall service levels. That
analysis is described in greater detail in appendix V.

To determine how air service has changed at small communities over time, we
analyzed changes in scheduled air service for different time periods. We
used our analysis of air service for the week of October 15- 21, 2000 as a
baseline for comparison. To minimize the possible effects of seasonality in
air service, we then examined air service schedules for the week of

42 We adopted DOT?s definition of a low- fare airline and included AirTran,
American Trans Air, Frontier, JetBlue, Southwest, Spirit, and Vanguard.

Appendix I: Objectives, Scope, and Methodology

Page 35 GAO- 02- 432 Small Community Air Service

October 15- 21, 2001. 43 To identify the service changes at small community
airports that might be separately attributable to the 2001 economic downturn
and the September 11 terrorist attacks, we examined two different sets of
airline schedule data for October 15- 21, 2001: those that airlines had
published prior to September 11, 2001, and those published by the airlines
following September 11, 2001. The first October 2001 schedule dataset
reflected the schedule as of August 30, 2001, and is, therefore, not
reflective of the airline industry?s reaction to the events of September 11.
The second schedule dataset for October 15- 21 reflected the schedule as of
October 12, 2001. Finally, to determine if airlines continued to make
substantial changes to their scheduled service, we also analyzed their
schedules for the week of November 1- 7, 2001. We recognize that airlines
make frequent changes to their service schedules, and that service at these
communities may have changed since then. We analyzed the same service
elements as for the week of October 15- 21, 2000.

To determine factors associated with the changes in service at small
community airports, we surveyed airport directors at nonhub and small hub
airports. We also interviewed officials from major and commuter airlines,
FAA and DOT, and industry experts. The survey responses helped us to
identify the individual airport perspectives on how their service has
changed and the impact of those changes, as well as the major factors
affecting the service changes. We interviewed airline officials to
understand how and why the major airlines were reducing and/ or transferring
small community airport routes to commuter carriers and how the different
types of contractual relationships affect the route changes. In addition,
airline officials described why many airlines are moving away from turboprop
to regional jets. FAA and DOT officials, and industry experts provided
further information on the state of the airline industry, particularly the
vulnerability of small community airports.

To collect information on the operational activities of small and nonhub
airports, and the opinions of their managers on a variety of issues, we
conducted a Web questionnaire survey of 280 U. S. airports from December 10,
2001, through January 29, 2002.

43 Analyses of airline service often uses data on airline schedules or
operations during the months of either May or September, because those
months are considered to be free of the effects of summer vacation travel or
lessened travel during winter months. We could not use September data
because of the effects of September 11, and we considered data for May 2001
to be too dated for our purposes. Internet Survey of

Airport Directors

Appendix I: Objectives, Scope, and Methodology

Page 36 GAO- 02- 432 Small Community Air Service

Using data from the FAA, Kiehl Hendrickson Group, American Association of
Airport Executives (AAAE), and the State of Alaska, we developed a sample of
280 small and nonhub airports. We did not survey the airport directors of
all nonhub and small hub airports. Because of the special circumstances (e.
g., unique remoteness) of smaller Alaska airports, we included another
criterion for incorporating them into our database: We only included Alaska
small and nonhubs that, in addition to meeting the prior criteria, were Part
139 certified. 44 This additional criterion resulted in a total of 20 Alaska
airports being included in our survey.

We developed our survey instrument in consultation with AAAE officials, who
reviewed our draft questionnaire and made suggestions about content and
format. We also pretested the draft questionnaire at four airports in our
study population. These airports were Bellingham International and Spokane
International (Washington) and Hagerstown, Maryland, and Richmond, Virginia.
We chose these airports because they represented- in terms of annual
enplanements, location, and airport type- the kinds of airports that would
be asked to complete our final questionnaire. We incorporated changes into
our survey instrument as a result of these pretests. The final questionnaire
was posted on the World Wide Web for respondents to complete. A complete
reproduction of the Web survey can be viewed in Adobe Acrobat pdf format at
www. gao. gov/ special. pubs/ d02432sv. pdf.

We sent e- mail messages or otherwise contacted airports in our survey
database in late November 2001 to notify them of our survey. We then sent
each airport representative a unique username and password and invited them
to fill out an automated questionnaire posted on the World Wide Web in early
December of 2001. About 12 percent of the airport representatives completed
a paper version of the questionnaire in lieu of completing the survey on-
line.

During the survey fieldwork period, we made at least three follow- up
contacts with each airport that had not yet responded to ask them to
participate. We used all completed responses received by January 29, 2002,
in the analysis for this report. We received responses from 207 airports in
which the respondent had indicated that they had completed their

44 These regulations prescribe rules governing the certification and
operation of land airports which serve any scheduled or unscheduled
passenger operation of an air carrier that is conducted with an aircraft
having a seating capacity of more than 30 passengers.

Appendix I: Objectives, Scope, and Methodology

Page 37 GAO- 02- 432 Small Community Air Service

questionnaire and that GAO could use the data (a 74 percent response rate).
Response rates did not vary appreciably across small hub and nonhub airports
and results of the follow- up efforts showed no evidence that our survey
results were not representative of the actual study population. Some
questions in our survey instrument were not answered by all of the airports
completing a useable questionnaire, but this rate of item nonresponse was
generally low.

In addition to any systematic bias or random error in our survey results
that may have been caused by our inability to obtain answers from all of the
airports in the population on all of our questions (nonresponse error),
estimates from questionnaire surveys may be subject to other sources of
error. We took steps to limit these errors. We checked our sample list of
airports against other sources to help ensure its completeness, we pretested
our questionnaire and had experts review it, and we checked our analysis for
programming errors. We did not, however, verify the answers reported by
airport directors.

Other important issues may be relevant to an analysis of the service changes
at small community airports. However, a lack of detailed information on
these factors limited the scope of this review. For example, we were not
able to obtain information on the differences in airfares at small
communities and at competing airports. Airfare data for the quarter
including October 2001 would not be available from the Bureau of
Transportation Statistics until late February or early March 2002.
Additionally, there is a lack of complete and representative fare data for
small communities, especially for local passengers who do not connect to
large carrier services. This is because public data on airfares is developed
from a 10 percent sample of tickets collected from large air carriers, which
comprises DOT?s ?Passenger Origin- Destination Survey? (O& D Survey). Small
certificated air carriers and commuter carriers do not participate in the O&
D Survey. Thus, there are inherent statistical sampling limitations in the
O& D Survey data. In addition, airlines? decisions about profitability of
operations in certain markets are proprietary confidential.

We conducted our work from April 2001 through March 2002 in accordance with
generally accepted government auditing standards. Other Data

Limitations to Our Work

Appendix II: Small Hub Airports Tend to Have More Commercial Air Service and
Were Less Affected by Airline Service Reductions

Page 38 GAO- 02- 432 Small Community Air Service

Small hub airports, the closest point of comparison to nonhubs, tend to
serve somewhat larger communities and have significantly more commercial air
service than do the 202 nonhub airports in the continental United States
included in our analysis. The median population for small hub communities
included in our analysis was about 417,000. As table 5 shows, only about 2
percent of small hub communities had service from two or fewer major
carriers; the other 98 percent had service from more carriers. Additionally,
only about 2 percent of small hub communities had service to three or fewer
nonstop destinations; the other 98 percent had nonstop service to more
locations. In addition, almost two- thirds of their nonstop destinations
were into major airline hubs, and the majority of their flights were on jet
aircraft (as opposed to turboprop or piston aircraft).

Table 5: Characteristics of Air Service at Nonhub and Small Hub Airports
(October 2000)

Airports with: Nonhubs Small hubs

Service from two or fewer major carriers 72.8% 1.9% Dominated by one carrier
84.2% 37.7% Service to three or fewer nonstop destinations 70.7% 1.9%
Percent of total nonstop destinations that are to major airline hubs 58.7%
64.4% Percent of daily flights that are on turboprop or piston aircraft
81.4% 37.5%

Source: GAO analysis of data from the Kiehl Hendrickson Group.

Compared to small communities with nonhub airports, the communities with
small hub airports had much greater daily air service. As table 6 shows, on
average, small hubs had significantly more carriers, more jet and turboprop
flights, and more nonstop destination options. For example, the median small
hub airport community was served by six airlines, compared with two airlines
for the small communities in our analysis, and the median number of daily
departures was 45, compared with 9 for small communities. Table 6 provides
additional information regarding small hubs and compares key differences in
the scope of air service that airlines scheduled at these two airport
categories. Appendix II: Small Hub Airports Tend to

Have More Commercial Air Service and Were Less Affected by Airline Service
Reductions

Appendix II: Small Hub Airports Tend to Have More Commercial Air Service and
Were Less Affected by Airline Service Reductions

Page 39 GAO- 02- 432 Small Community Air Service

Table 6: Median Airline Service Levels at Nonhub and Small Hub Airports,
October 2000

Air service dimension Small

communities (nonhub airports)

Communities with small hub

airports

Airlines providing commercial service 2 6 Daily nonstop destinations 2 10
Daily nonstop destinations that are airline hubs 1 7 Daily turboprop
departures 8 16 Daily jet departures 0 25 Total daily departures 9 a 45 a
The median number of total daily departures does not necessarily equal the
sum of the median number of daily turboprop departures and the median number
of daily jet departures because the median values are calculated separately
for each air service dimension, and because the distribution of values in
each category may not be identical.

Source: GAO analysis of data from the Kiehl Hendrickson Group.

Compared to the experience of small communities, small hub airports saw
relatively little change in their airline schedules during the period we
analyzed. For example, there was little change in the number of small hubs
that had service from three or more carriers (see figs. 8 and 9). In
addition, the number of small hubs with service to more than two nonstop
destinations did not change, and the number of small hubs dominated by a
single carrier declined slightly.

Appendix II: Small Hub Airports Tend to Have More Commercial Air Service and
Were Less Affected by Airline Service Reductions

Page 40 GAO- 02- 432 Small Community Air Service

Figure 8: Change in the Number of Airlines Serving Small Communities with
Nonhub Airports, October 2000- October 2001

Source: GAO analysis of data from the Kiehl Hendrickson Group.

Appendix II: Small Hub Airports Tend to Have More Commercial Air Service and
Were Less Affected by Airline Service Reductions

Page 41 GAO- 02- 432 Small Community Air Service

Figure 9: Change in the Number of Airlines Serving Communities with Small
Hub Airports, October 2000- October 2001

Source: GAO analysis of data from the Kiehl Hendrickson Group.

According to other data on changes in daily scheduled seats from U. S.
airports, airports of all sizes experienced generally comparable decreases
in total service. Small hubs experienced a greater relative decrease in
service (- 15.5 percent) compared to nonhubs (- 13.5 percent). Large hubs
had the greatest relative decrease in total available seats (- 16.3
percent), and medium hubs had a smaller decrease (- 12.4 percent).

Appendix III: Change in Air Service at 202 U. S. Nonhub Airports, October
2000- October 2001

Page 42 GAO- 02- 432 Small Community Air Service

Table 7: Changes in the Number of Carriers Serving Small Community Airports
and Nonstop Destinations Major Carriers Number of Nonstop Destinations State
Community Population

(2000) Enplanements (1999) Oct

-00 Oct- 01

(original) Oct- 01 (revised) Total

Change Oct00 Oct- 01

(original) Oct- 01 (revised) Total

Change

Dothan 137,916 66,025 2 1 1 -1 3 11- 2 AL Montgomery 333,055 231,061 4 3 3
-1 5 44- 1

Fort Smith 207,290 102,583 3 2 2 -1 2 220 Texarkana 129,749 46,049 2 2 2 0 1
110 AR Fayetteville 311,121 320,438 5 5 5 0 7 992 Fort Huachuca Sierra
117,755 7, 005 1 1 1 0 1 110 Flagstaff 122,366 33,978 1 1 1 0 1 110 AZ

Yuma 160,026 64,078 2 2 2 0 2 220 Arcata/ Eureka 126,518 111,071 2 2 2 0 5
550 Bakersfield 661,645 147,142 3 3 3 0 4 33- 1

Chico 203,171 30,004 1 1 1 0 1 110 Imperial 142,361 24,834 1 1 1 0 1 110
Inyokern 661,645 9, 089 1 1 1 0 1 110 Modesto 446,997 28,314 1 1 1 0 1 110
Monterey 401,762 258,605 3 3 3 0 3 330 Redding 163,256 74,606 2 3 3 1 3 441
San Luis Obispo 246,681 147,028 3 3 3 0 3 330 Santa Maria 399,347 44,591 1 1
1 0 1 110 CA

Visalia 368,021 10,255 1 1 1 0 2 11- 1

Aspen 14,872 215,685 2 2 2 0 2 220 Durango 43,941 96,647 3 3 3 0 3 341 Eagle
41,659 175,457 1 1 1 0 1 110 Grand Junction 116,255 137,793 2 3 3 1 3 330
Gunnison 13,956 57,953 1 1 1 0 1 110 Hayden 19,690 108,797 1 1 1 0 1 110
Montrose 33,432 70,799 2 3 3 1 2 331 CO

Telluride 6,594 22,483 2 1 1 -1 2 11- 1

Naples 251,377 54,494 3 2 2 -1 4 32- 2

Daytona Beach 443,343 275,231 2 2 1 -1 3 32- 1

Key West 79,589 275,909 5 5 5 0 6 660 Gainesville 217,955 152,087 2 2 2 0 2
220 Melbourne 476,230 273,813 2 3 1 -1 2 620 FL

Panama City 148,217 164,426 3 3 3 0 5 661

Appendix III: Change in Air Service at 202 U. S. Nonhub Airports, October
2000- October 2001

Appendix III: Change in Air Service at 202 U. S. Nonhub Airports, October
2000- October 2001

Page 43 GAO- 02- 432 Small Community Air Service

Major Carriers Number of Nonstop Destinations State Community Population

(2000) Enplanements (1999) Oct

-00 Oct- 01

(original) Oct- 01 (revised) Total

Change Oct00 Oct- 01

(original) Oct- 01 (revised) Total

Change

Albany 120,822 44,339 1 1 1 0 1 110 Augusta 477,441 215,556 2 2 2 0 2 220
Athens 153,444 11,234 1 1 1 0 1 110 Brunswick 67,568 24,492 1 1 1 0 1 110
Columbus 274,624 93,512 3 2 2 -1 4 22- 2

Macon 322,549 30,207 1 1 1 0 2 220 GA

Valdosta 92,115 32,695 1 1 1 0 1 110 Waterloo 128,012 58,904 3 3 3 0 4 43- 1

Dubuque 89,143 55,555 3 2 2 -1 4 22- 2

Fort Dodge 40,235 11,801 1 1 1 0 1 110 Mason City 46,447 13,477 1 1 1 0 2
220 IA

Sioux City 124,130 89,563 2 2 2 0 2 220 Idaho Falls 82,522 120,699 2 2 2 0 2
331 Lewiston 37,410 67,041 1 1 1 0 2 331 Pocatello 7, 538 46,679 2 2 2 0 2
331 Hailey 18,991 67,632 2 2 2 0 2 220 ID

Twin Falls 64,284 36,425 1 1 1 0 1 110 Bloomington/ Normal 150,433 217,596 6
5 5 -1 6 55- 1

Champaign/ Urbana 179,669 133,845 3 3 3 0 4 33- 1

Decatur 114,706 24,989 1 1 1 0 1 110 Peoria 347,387 219,791 4 5 5 1 8 880
Rockford 371,236 32,608 1 0 0 -1 1 00- 1

Springfield 201,437 80,755 2 4 4 2 3 441 IL

Quincy 68,277 11,415 2 2 2 0 2 220 Evansville 171,922 257,966 5 5 5 0 9 87-
2 IN Lafayette 182,821 19,228 2 1 1 -1 3 11- 2

Manhattan 62,843 19,908 1 1 1 0 2 220 KS Salina 53,597 15,978 2 1 1 -1 3 22-
1

KY Paducah 65,514 26,300 2 2 2 0 4 32- 2

Alexandria 126,337 116,006 4 3 3 -1 5 44- 1

Lake Charles 183,577 76,263 2 2 1 -1 2 21- 1

Lafayette 385,647 189,772 4 4 4 0 4 440 LA

Monroe 147,250 122,412 3 3 3 0 8 66- 2

Nantucket 9, 520 289,655 3 3 3 0 6 660 Hyannis 162,582 208,508 3 3 3 0 4 340
Vineyard Haven 14,987 73,461 2 2 2 0 6 65- 1

MA Provincetown 162,582 15,925 1 1 1 0 1 110

Appendix III: Change in Air Service at 202 U. S. Nonhub Airports, October
2000- October 2001

Page 44 GAO- 02- 432 Small Community Air Service

Major Carriers Number of Nonstop Destinations State Community Population

(2000) Enplanements (1999) Oct

-00 Oct- 01

(original) Oct- 01 (revised) Total

Change Oct00 Oct- 01

(original) Oct- 01 (revised) Total

Change

Cumberland 27,078 6, 142 1 0 0 -1 2 00- 2

Hagerstown 131,923 27,050 1 1 1 0 2 11- 1 MD Salisbury 86,644 73,124 1 1 1 0
3 32- 1

Alpena 31,314 10,263 1 1 1 0 2 220 Kalamazoo 452,851 278,212 6 6 5 -1 6 65-
1

Sault Ste. Marie 38,543 14,937 1 1 1 0 1 110 Hancock 36,016 27,998 1 1 1 0 3
11- 2

Escanaba 38,520 20,550 2 2 2 0 5 33- 2

Saginaw 403,070 294,483 4 3 3 -1 5 44- 1

Pellston 31,437 31,977 1 1 1 0 3 330 Marquette 64,634 43,200 0 1 1 1 0 111
MI

Traverse City 77,654 189,809 4 4 3 -1 7 74- 3

Bemidji 39,650 29,457 2 2 2 0 3 330 Brainerd 55,099 19,190 1 1 1 0 2 220
Duluth 200,528 140,835 2 2 2 0 2 231 Grand Rapids 43,992 10,367 1 1 1 0 1
210 Hibbing 200,528 15,709 1 1 1 0 1 210 International Falls 14,355 22,460 1
1 1 0 1 121 Rochester 124,277 152,492 2 3 2 0 2 320 MN

Thief River Falls 13,584 8, 854 1 1 1 0 1 110 Columbia 135,454 26,268 2 1 1
-1 3 11- 2 MO Joplin 157,322 28,877 3 2 2 -1 4 22- 2

Greenville 62,977 13,265 1 1 1 0 2 220 Columbus 61,586 44,976 2 2 2 0 4 43-
1

Meridian 78,161 30,991 2 2 2 0 4 440 Hattiesburg/ Laurel 64,958 12,331 1 1 1
0 2 21- 1

MS Tupelo 75,755 15,494 1 1 1 0 2 220 Billings 129,352 338,769 5 5 5 0 12 12
12 0

Butte 34,606 47,963 2 2 2 0 3 330 Bozeman 67,831 223,006 4 4 4 0 6 55- 1

Kalispell 74,471 146,942 4 4 4 0 6 65- 1

Great Falls 80,357 138,705 4 4 4 0 7 770 Helena 55,716 79,166 3 3 3 0 5 550
Missoula 95,802 221,292 4 5 5 1 7 781 MT

West Yellowstone 67,831 4, 998 0 1 0 0 0 200

Appendix III: Change in Air Service at 202 U. S. Nonhub Airports, October
2000- October 2001

Page 45 GAO- 02- 432 Small Community Air Service

Major Carriers Number of Nonstop Destinations State Community Population

(2000) Enplanements (1999) Oct

-00 Oct- 01

(original) Oct- 01 (revised) Total

Change Oct00 Oct- 01

(original) Oct- 01 (revised) Total

Change

Asheville 225,965 283,144 2 3 3 1 5 661 New Bern 91,436 73,882 2 2 1 -1 2
21- 1

Fayetteville 302,963 157,906 2 2 2 0 2 220 Hickory 341,851 21,532 1 1 1 0 1
110 Wilmington 233,450 246,790 3 3 2 -1 4 32- 2

Jacksonville 150,355 54,722 1 1 1 0 2 220 Greenville 133,798 43,756 1 1 1 0
2 11- 1

NC Pinehurst 74,769 20,238 1 1 1 0 1 110 Bismarck 94,719 129,327 3 3 3 0 3
441 Fargo 174,367 226,385 2 2 2 0 3 330 Grand Forks 97,478 88,281 1 1 1 0 1
110 Williston 19,761 5, 613 1 2 2 1 1 332 ND

Minot 58,795 74,333 1 1 1 0 1 110 Scottsbluff 36,951 12,219 1 1 1 0 1 110
Grand Island 53,354 13,063 1 2 2 1 5 44- 1 NE Lincoln 250,291 281,169 4 4 4
0 5 550 NH Lebanon 81,743 20,152 1 1 1 0 2 331

Carlsbad 51,658 7, 787 2 2 2 0 3 330 Farmington 113,801 53,538 3 4 4 1 3 341
Hobbs 55,511 2, 512 2 2 2 0 3 552 Roswell 61,382 18,832 2 2 2 0 4 440 NM

Santa Fe 147,635 26,178 2 1 2 0 2 131 NV Elko 45,291 119,295 1 1 1 0 2 220

Binghamton 252,320 136,305 3 3 3 0 8 55- 3

Elmira/ Corning 91,070 108,124 2 2 2 0 6 63- 3

Ithaca 96,501 101,945 1 1 1 0 6 54- 2

NY Jamestown 139,750 20,827 1 1 1 0 1 110 Toledo 618,203 248,017 6 5 5 -1 6
55- 1 OH Youngstown/ Warren 594,746 40,274 2 2 1 -1 4 43- 1

OK Lawton 114,996 62,335 2 1 1 -1 1 110 Klamath Falls 63,775 33,729 1 1 1 0
1 110 Medford 181,269 224,699 2 2 2 0 4 440 North Bend 62,779 29,886 1 1 1 0
1 110 Pendleton 70,548 14,019 1 1 1 0 2 220 OR

Redmond 115,367 140,915 2 2 2 0 3 330

Appendix III: Change in Air Service at 202 U. S. Nonhub Airports, October
2000- October 2001

Page 46 GAO- 02- 432 Small Community Air Service

Major Carriers Number of Nonstop Destinations State Community Population

(2000) Enplanements (1999) Oct

-00 Oct- 01

(original) Oct- 01 (revised) Total

Change Oct00 Oct- 01

(original) Oct- 01 (revised) Total

Change

Altoona 129,144 16,969 1 1 1 0 3 21- 2

Wilkes- Barre/ Scranton 624,776 234,292 4 3 3 -1 9 64- 5

Bradford 45,936 13,131 1 1 1 0 1 110 Du Bois 45,932 17,355 1 1 1 0 2 21- 1

Erie 280,843 167,507 3 2 2 -1 3 22- 1

Williamsport 120,044 46,519 1 1 1 0 2 220 Johnstown 232,621 20,899 1 1 1 0 2
120 Lancaster 470,658 19,342 1 1 1 0 1 110 Reading 373,638 52,519 1 1 1 0 2
220 PA

State College 135,758 126,945 3 3 3 0 5 550 Block Island 293,566 11,190 1 1
1 0 1 110 RI Westerly 293,566 12,142 1 1 1 0 1 110 Florence 125,761 57,123 2
2 2 0 2 220 SC Hilton Head Island 120,937 100,194 1 1 1 0 2 11- 1

Aberdeen 35,460 25,365 1 1 1 0 3 330 Watertown 25,897 9, 324 1 1 1 0 1 221
Pierre 16,481 18,228 2 2 2 0 2 331 SD

Rapid City 88,565 195,209 3 3 3 0 3 330 Chattanooga 465,161 303,689 4 4 4 0
5 650 TN Bristol/ Johnson/ King 480,091 221,228 4 4 4 0 6 660 Abilene
126,555 47,984 2 2 1 -1 3 21- 2

Waco 213,517 67,045 2 2 1 -1 4 21- 3

Beaumont/ Port Arthur 385,090 97,537 3 2 2 -1 2 220 Brownsville 335,227
71,949 1 1 1 0 1 110 College Station 152,415 93,005 2 2 2 0 2 220 Longview
208,780 30,497 1 1 1 0 1 110 Killeen 312,952 90,418 3 3 3 0 3 22- 1

Laredo 193,117 87,739 2 2 2 0 2 220 Mc Allen 569,463 311,237 3 3 2 -1 3 32-
1

San Angelo 104,010 39,411 2 2 1 -1 2 21- 1

Wichita Falls 140,518 55,903 2 1 1 -1 1 110 Tyler 174,706 74,233 3 2 1 -2 3
21- 2

TX Victoria 84,088 20,016 1 1 1 0 1 110 UT St. George 90,354 33,707 2 2 2 0
2 331

Charlottesville 159,576 171,150 3 4 4 1 6 771 Lynchburg 214,911 85,822 3 3 3
0 4 440 VA

Staunton/ Harrisburg 65,615 16,494 2 2 2 0 2 220

Appendix III: Change in Air Service at 202 U. S. Nonhub Airports, October
2000- October 2001

Page 47 GAO- 02- 432 Small Community Air Service

Major Carriers Number of Nonstop Destinations State Community Population

(2000) Enplanements (1999) Oct

-00 Oct- 01

(original) Oct- 01 (revised) Total

Change Oct00 Oct- 01

(original) Oct- 01 (revised) Total

Change

Walla Walla 55,180 31,166 1 1 1 0 1 110 Bellingham 166,814 97,406 2 2 2 0 1
110 Port Angeles 64,525 28,201 1 1 1 0 2 11- 1

Wenatchee 32,603 52,855 1 1 1 0 2 11- 1

Pasco 191,822 206,105 3 3 3 0 6 44- 2

Pullman/ Moscow, ID 40,740 34,887 1 1 1 0 2 220 WA

Yakima 222,581 89,569 2 2 2 0 1 110 Appleton 358,365 266,629 4 4 4 0 8 66- 2

Mosinee 125,834 142,980 3 3 3 0 6 660 Eau Claire 148,337 20,611 1 1 1 0 2
220 La Crosse 126,838 113,640 3 3 3 0 4 440 WI

Rhinelander 36,776 38,651 2 1 1 -1 5 22- 3

Clarksburg 68,652 16,276 1 2 2 1 2 431 Charleston 251,662 266,679 5 5 5 0 11
11 10 -1

Huntington 42,903 62,609 2 2 2 0 3 330 Lewisburg 34,453 12,771 1 2 2 1 4 43-
1

Morgantown 81,866 21,561 1 1 1 0 3 32- 1

WV Parkersburg 151,237 25,677 1 1 1 0 1 110 Cody 25,786 28,326 2 2 2 0 2 220
Casper 66,533 66,184 2 2 2 0 2 220 Cheyenne 81,607 20,520 1 1 1 0 1 110
Gillette 33,698 15,356 1 1 1 0 1 110 Jackson 18,251 165,595 2 2 2 0 2 220
Riverton 35,804 13,327 1 1 1 0 2 220 WY

Sheridan 26,560 15,052 1 1 1 0 1 110 Source: GAO analysis of data from U. S.
Census Bureau, Federal Aviation Administration, and Kiehl Hendrickson Group.

Appendix III: Change in Air Service at 202 U. S. Nonhub Airports, October
2000- October 2001

Page 48 GAO- 02- 432 Small Community Air Service

Table 8: Changes in the Average Number of Daily Turboprop and Jet Departures
at Small Communities Daily turboprop departures Daily jet departures

State Community Oct- 00 Oct- 01 (original) Oct- 01

(revised) Total

turboprop change Oct- 00 Oct- 01

(original) Oct- 01 (revised) Total jet

change

AL Dothan 12 3 7 -5 0 400 Montgomery 13 10 10 -3 6 1093 Fort S mith 13 13
11- 2 2 11- 1

Texarkana 10 10 9 -1 1 110 AR Fayetteville 13 13 12- 1 13 13 13 0

Fort Huachuca Sierra 4 4 3- 1 0 000 Flagstaff 8 6 5 -3 0 000 AZ

Yuma 13 13 10 -3 0 000 Arcata/ Eureka 13 13 13 0 0 000 Bakersfield 13 13 130
4 32- 2

Chico 6 6 5- 1 0 000 Imperial 6 6 6 0 0 000 Inyokern 4 4 4 0 0 000 Modesto 6
6 5 -1 0 000 Monterey 13 13 13 0 4 43- 1

Redding 12 13 12 0 0 000 San Luis Obispo 13 13 13 0 0 000 Santa Maria 10 8 8
-2 0 000 CA

Visalia 5 4 3- 2 0 000 Aspen 3 2 2 -1 6 75- 1

Durango 12 12 12 0 3 11- 2

Eagle 7 1 1 -6 0 222 Grand Junction 13 13 13 0 0 111 Gunnison 7 4 4 -3 0 000
Hayden 7 7 6- 1 0 000 Montrose 8 8 6 -2 0 111 CO

Telluride 5 5 5 0 0 000 Naples 12 10 8 -4 0 000 Daytona Beach 1 1 1 0 8 991
Key West 13 13 13 0 0 000 Gainesville 7 7 7 0 5 550 Melbourne 1 1 1 0 9 1390
FL

Panama City 13 13 13 0 1 110

Appendix III: Change in Air Service at 202 U. S. Nonhub Airports, October
2000- October 2001

Page 49 GAO- 02- 432 Small Community Air Service

Daily turboprop departures Daily jet departures State Community Oct- 00 Oct-
01

(original) Oct- 01 (revised)

Total turboprop

change Oct- 00 Oct- 01 (original) Oct- 01

(revised) Total jet change

Albany 8 1 6 -2 0 400 Augusta 13 13 12 -1 3 441 Athens 4 4 4 0 0 000
Brunswick 5 5 5 0 0 000 Columbus 13 12 12 -1 0 000 Macon 7 7 70 1 110 GA

Valdosta 7 7 70 0 000 Waterloo 13 13 12 -1 0 000 Dubuque 12 3 4 -8 0 444
Fort Dodge 5 5 3- 2 0 000 Mason City 9 9 5 -4 0 000 IA

Sioux City 12 11 9 -3 2 331 Idaho Falls 10 5 5 -5 5 10105 Lewiston 10 10 8
-2 0 111 Pocatello 10 6 6- 4 0 433 Hailey 10 11 111 0 000 ID

Twin Falls 6 6 6 0 0 000 Bloomington/ Normal 13 13 13 0 8 991 Champaign/
Urbana 13 11 8 -5 4 551 Decatur 5 4 3- 2 0 000 Peoria 13 13 13 0 11 12 11 0

Rockford 5 1 1 -4 0 000 Springfield 13 13 13 0 0 000 IL

Quincy 7 7 6- 1 0 000 Evansville 13 13 13 0 11 13 13 2

IN Lafayette 8 3 3 -5 0 000 Manhattan 8 8 6 -2 0 000 KS Salina 7 5 4 -3 0
000 KY Paducah 9 8 7 -2 0 000

Alexandria 13 13 13 0 0 000 Lake Charles 11 10 7 -4 0 000 Lafayette 13 13 13
0 2 320 LA

Monroe 10 9 8 -2 6 660 Nantucket 13 13 13 0 0 000 Hyannis 13 13 13 0 0 000
Vineyard Haven 13 13 13 0 0 000 MA

Provincetown 6 7 71 0 000 Cumberland 5 1 1 -4 0 000 Hagerstown 7 6 5 -2 0
000 MD Salisbury 13 12 11- 2 0 000

Appendix III: Change in Air Service at 202 U. S. Nonhub Airports, October
2000- October 2001

Page 50 GAO- 02- 432 Small Community Air Service

Daily turboprop departures Daily jet departures State Community Oct- 00 Oct-
01

(original) Oct- 01 (revised)

Total turboprop

change Oct- 00 Oct- 01 (original) Oct- 01

(revised) Total jet change

Alpena 7 9 5 -2 0 000 Kalamazoo 13 13 12- 1 13 13 13 0

Sault Ste. Marie 4 5 3 -1 0 000 Hancock 8 5 5 -3 0 000 Escanaba 9 7 6 -3 0
000 Saginaw 9 5 4 -5 13 13 13 0

Pellston 9 9 7 -2 0 000 Marquette 1 1 1 0 0 111 MI

Traverse City 13 13 9 -4 6 1093 Bemidji 10 10 8 -2 0 000 Brainerd 9 7 5 -4 0
000 Duluth 3 3 41 10 10 8 -2

Grand Rapids 4 4 3 -1 0 000 Hibbing 4 3 3 -1 0 000 International Falls 3 3 4
1 0 000 Rochester 4 4 3- 1 10 13 9 -1

MN Thief River Falls 2 2 2 0 0 000 Columbia 7 6 4 -3 5 00- 5

MO Joplin 7 7 6 -1 4 00- 4

Greenville 5 5 4 -1 0 000 Columbus 13 12 11 -2 1 221 Meridian 13 13 12 -1 0
000 Hattiesburg/ Laurel 4 4 3 -1 0 000 MS

Tupelo 6 7 7 1 0 000 Billings 13 13 13 0 13 13 13 0

Butte 1 1 1 0 7 770 Bozeman 3 2 2- 1 10 12 12 2

Kalispell 6 4 4 -2 7 770 Great Falls 6 5 5 -1 11 10 11 0

Helena 7 7 6 -1 6 65- 1

Missoula 5 2 2 -3 12 13 13 1

MT West Yellowstone 1 1 1 0 0 100 Asheville 13 13 130 8 10102 New Bern 13 13
7 -6 0 000 Fayetteville 9 13 13 4 6 32- 4

Hickory 7 6 5 -2 0 000 Wilmington 13 13 11 -2 8 75- 3

Jacksonville 6 7 6 0 0 000 Greenville 9 8 6 -3 0 000 NC

Pinehurst 6 6 4 -2 0 000

Appendix III: Change in Air Service at 202 U. S. Nonhub Airports, October
2000- October 2001

Page 51 GAO- 02- 432 Small Community Air Service

Daily turboprop departures Daily jet departures State Community Oct- 00 Oct-
01

(original) Oct- 01 (revised)

Total turboprop

change Oct- 00 Oct- 01 (original) Oct- 01

(revised) Total jet change

Bismarck 7 8 70 4 43- 1

Fargo 2 2 20 13 13 12 -1

Grand Forks 4 4 3 -1 4 43- 1

Williston 3 5 5 2 0 000 ND

Minot 1 1 1 0 4 33- 1

Scottsbluff 5 5 4 -1 0 000 Grand Island 8 7 7 -1 0 000 NE

Lincoln 10 7 7 -3 8 13124 NH Lebanon 8 8 6 -2 0 000

Carlsbad 5 5 50 0 000 Farmington 13 13 13 0 0 000 Hobbs 4 5 5 1 0 000
Roswell 9 8 8- 1 0 000 NM

Santa Fe 13 11 10 -3 0 000 NV Elko 12 11 11 -1 0 000

Binghamton 13 13 13 0 2 331 Elmira/ Corning 13 13 9 -4 4 440 Ithaca 13 13
12- 1 4 33- 1

NY Jamestown 7 6 4 -3 0 000 Toledo 13 10 8 -5 13 13 13 0

OH Youngstown/ Warren 9 5 4 -5 0 000 OK Lawton 12 8 7 -5 0 000

Klamath Falls 5 5 4 -1 0 000 Medford 13 13 13 0 4 440 North B end 5 5 4- 1 0
000 Pendleton 6 6 5 -1 0 000 OR

Redmond 13 13 13 0 0 000 Altoona 7 6 4 -3 0 000 WilkesBarre/ Scranton 13 9
7- 6 11 10 10 -1

Bradford 6 5 4 -2 0 000 Du B ois 6 6 5- 1 0 000 Erie 13 7 5- 8 4 651
Williamsport 11 11 7 -4 0 000 Johnstown 6 6 5 -1 0 000 Lancaster 5 5 4 -1 0
000 Reading 11 11 6 -5 0 000 PA

State College 13 13 13 0 0 000 Block I sland 9 9 90 0 000 RI Westerly 9 9 90
0 000 Florence 12 13 11 -1 0 000 SC Hilton Head Island 12 12 12 0 0 000

Appendix III: Change in Air Service at 202 U. S. Nonhub Airports, October
2000- October 2001

Page 52 GAO- 02- 432 Small Community Air Service

Daily turboprop departures Daily jet departures State Community Oct- 00 Oct-
01

(original) Oct- 01 (revised)

Total turboprop

change Oct- 00 Oct- 01 (original) Oct- 01

(revised) Total jet change

Aberdeen 13 9 8 -5 0 000 Watertown 4 4 40 0 000 Pierre 7 6 5 -2 0 000 SD

Rapid City 10 6 4 -6 8 1091 Chattanooga 13 13 12 -1 13 13 13 0 TN Bristol/
Johnson/ King 13 13 13 0 7 44- 3

Abilene 11 10 6 -5 1 221 Waco 12 12 8 -4 0 000 Beaumont/ Port Arthur 13 13
13 0 0 000 Brownsville 4 3 4 0 3 430 College Station 13 13 13 0 0 000
Longview 6 6 5 -1 0 000 Killeen 13 13 13 0 0 000 Laredo 11 11 10 -1 0 000 Mc
Allen 2 1 1 -1 10 9 8 -2

San Angelo 11 10 6 -5 0 000 Wichita Falls 11 7 6 -5 0 000 Tyler 13 13 9 -4 0
000 TX

Victoria 5 5 4- 1 0 000 UT St. George 8 10 10 2 0 000

Charlottesville 13 13 13 0 2 331 Lynchburg 13 13 13 0 0 000 VA

Staunton/ Harrisburg 10 9 6 -4 0 000 Walla Walla 6 5 4 -2 0 000 Bellingham
13 13 12 -1 0 000 Port Angeles 9 5 4 -5 0 000 Wenatchee 9 8 6 -3 0 000 Pasco
13 13 130 5 550 Pullman/ Moscow, ID 6 6 5 -1 0 111 WA

Yakima 13 13 11 -2 0 000 Appleton 13 11 8 -5 13 13 13 0

Mosinee 13 13 130 1 221 Eau Claire 9 9 7 -2 0 000 La Crosse 9 9 7 -2 8 86- 2

WI Rhinelander 11 5 5 -6 0 000

Appendix III: Change in Air Service at 202 U. S. Nonhub Airports, October
2000- October 2001

Page 53 GAO- 02- 432 Small Community Air Service

Daily turboprop departures Daily jet departures State Community Oct- 00 Oct-
01

(original) Oct- 01 (revised)

Total turboprop

change Oct- 00 Oct- 01 (original) Oct- 01

(revised) Total jet change

WV Clarksburg 8 8 5 -3 0 111 Charleston 13 13 13 0 13 13 13 0

Huntington 13 13 12 -1 0 000 Lewisburg 4 3 3 -1 1 221 Morgantown 12 10 7 -5
0 000 Parkersburg 8 8 6 -2 0 000 WY Cody 6 7 7 1 0 000

Casper 12 12 9 -3 1 110 Cheyenne 10 7 6 -4 0 000 Gillette 6 5 4 -2 0 000
Jackson 11 11 11 0 2 11- 1

Riverton 8 7 7- 1 0 000 Sheridan 5 4 4 -1 0 000

Note: a Average daily departures calculated by dividing total weekly
departures by 7 and rounding result to nearest integer.

Source: GAO analysis of data from the Kiehl Hendrickson Group.

Appendix IV: Changes in Air Service at Small Communities in Alaska and
Hawaii

Page 54 GAO- 02- 432 Small Community Air Service

Unique conditions affecting air service in Alaska and Hawaii required us to
look at these two states separately from the rest of the United States. Both
states have distinctive geographies: they are both located outside the
continental United States and both have unique topographies that require air
service to be used as a major source of intrastate travel.

We examined air service at 63 nonhub airports in Alaska and 2 in Hawaii. All
of the Alaska airports were located in communities with less than 100,000
population; the median population was 7,208. The median passenger
enplanements in 1999 was 5,176 (about 14 per day). The two Hawaii airports
were located in larger communities; the average population was 128,094.
Their median enplanements in 1999 were 108,258 (about 297 per day).

There was little change in air service at the small community airports in
Alaska and Hawaii between October 2000 and October 2001 (revised), as the
median level of service represented by the indicators below show (see table
9). None of these airports had nonstop service to a major airline network?s
hub. (The major U. S. airlines do not operate hubs in either state.) From
October 2000 to October 2001 (revised), the number of these airports that
were dominated by a single airline increased slightly, from 65 percent to 69
percent.

Table 9: Median Number of Carriers, Nonstop Destinations Served, Daily
Turboprop and Jet Departures for Nonhub Airports in Alaska and Hawaii

Time frame Number of carriers

Nonstop destinations

served Daily

turboprop departures Daily jet

departures

October 2000 3 3 6 0 October 2001 (original) 3 3 6 0 October 2001 (revised)
3 3 6 0

Source: GAO analysis of airline schedule data from the Kiehl Hendrickson
Group.

There are two communities that are categorized as small hubs in Alaska-
Juneau and Fairbanks- and three in Hawaii- Hilo, Kailua/ Kona, and Lihue.
The Alaska airports were all located in communities with populations less
than 100,000 and had no service to an airline hub. Hawaii?s small hubs were
in communities with populations of less than 250,000. Two of those
communities had service to two airline hubs (San Francisco International and
Los Angeles International). Appendix IV: Changes in Air Service at Small

Communities in Alaska and Hawaii

Appendix IV: Changes in Air Service at Small Communities in Alaska and
Hawaii

Page 55 GAO- 02- 432 Small Community Air Service

Small hub airports in Alaska and Hawaii have notably more passenger traffic
and air service than the states? nonhubs. The median passenger enplanements
in the two Alaska airports in 1999 were 385,470 (about 1,056 per day), and
the median passenger enplanements in Hawaii were 1,271,744 (about 3,484 per
day). Typically, Alaska and Hawaii small hubs received service from four
major or independent carriers with service to seven nonstop destinations. In
addition, small hubs typically had 213 jet departures per week (30 jet
departures per day) in October 2000.

Generally, the overall amount of service for these small hub airports
declined between October 2000 and October 2001. Specifically, airlines
scheduled 50 fewer weekly jet departures (eight per day), and added 8
additional weekly turboprop departures (one per day). See table 10.

Table 10: Median Number of Carriers Serving, Nonstop Destinations Served,
Daily Turboprop and Jet Departures for Small Hub Airports in Alaska and
Hawaii

Time frame Number of carriers

Nonstop destinations

served Daily

turboprop departures Daily jet

departures

October 2000 5 7 2 30 October 2001 (original) 4 6 0 27 October 2001
(revised) 4 7 3 22

Note: Numbers are rounded to the nearest integer. Source: GAO analysis of
airline schedule data from the Kiehl Hendrickson Group.

Appendix V: Economic Factors Affect Air Service in Small Communities

Page 56 GAO- 02- 432 Small Community Air Service

To examine the factors associated with air service in small communities in
October 2000, we statistically analyzed certain economic characteristics of
these communities. Our process and the outcomes of our analysis are outlined
below.

For this study, we used regression analysis to explore which factors, called
independent variables, explain differences in the level of service, called
the dependent variable, in small communities in October 2000. A regression
model is a statistical tool that enables researchers to investigate
relationships between the dependent variable and the independent variables.

To examine the factors associated with the level of air service provided to
small communities in October 2000, we used an ordinary least squares
regression model. We developed several models, looking at the contribution
of each independent variable to the predictive ability of the models, and
the overall explanatory power of the models as measured by the coefficient
of determination, or r- squared. R- squared is a measure of the proportion
of the total variation in the dependent variable that can be explained by
the independent variables in that particular model.

Economic principles indicate that as income, market population, and the
price of substitute service increase, demand for a service will increase.
Under these conditions, within a competitive marketplace, as passenger
demand increases, the supply of air service will increase to meet that
demand. We, therefore, expect that communities with greater levels of income
and gross regional product 45 and larger populations and employment levels
will experience more substantial air service. Likewise, we expect that
communities that are farther from an airport with a lowfare carrier will
realize better service. 46

45 Regional product is a concept similar to gross domestic product (i. e.,
the output of goods and services produced by labor and property located in
the United States), only measured at the regional level.

46 We were unable to obtain data on airfares and the cost of alternative
modes of air travel. However, as the distance to an alternative carrier
increases, the cost of gaining access to the carrier also increases. In our
analysis, we include the minimum distance to a low- fare carrier as a
substitute measure for the price of alternative travel. Appendix V: Economic
Factors Affect Air

Service in Small Communities Regression Model

Economic Principles

Appendix V: Economic Factors Affect Air Service in Small Communities

Page 57 GAO- 02- 432 Small Community Air Service

We obtained the economic data used in the regression analysis from the
Regional Economic Information System database produced by the Bureau of
Economic Analysis. The data were collected for October 1999 at the county
level. 47 We then created a dataset containing variables for each county,
including population, total employment, manufacturing earnings, and per
capita income. We merged this dataset with the data on air service and the
distance between airports to create a final working dataset for this
analysis. Table 11 summarizes the descriptive statistics of the economic
variables and other factors for the 202 small communities in our analysis.

Table 11: Characteristics of 202 Small Communities Included in Economic
Analysis Variable Mean Minimum Maximum

Employment (full- and part- time jobs) 77,558 5, 397 309,598 Manufacturing
earnings $344,579 $3,403 $2,573,390 Minimum distance to a low- fare carrier
airport (miles) 128 0 546 Per capita income $25,147 $13,339 $65,573
Population 126,813 5, 464 642,495 Number of jet departures per week (October
2000) 15.50 0 183 Number of turboprop departures per week (October 2000)
67.88 0 437 Number of major carriers serving communities (October 2000) 2.01
0 6 Number of nonstop destinations served (October 2000) 2.94 0 12

Source: GAO analysis of data from the Bureau of Economic Analysis.

We used employment and population to represent the size of a community and
per capita income as a measure of income. We expect that a community with a
larger manufacturing sector will have a greater demand for business travel.
However, data on business travel and regional exports were unavailable for
this study. In addition, it is difficult to obtain data on gross regional
product (a measure similar to gross domestic product that is applied at the
regional level). Therefore, for the purposes of our analysis, we used
manufacturing earnings to represent the level of export activity from a
region and, hence, as an indicator of the possible demand for business
travel.

47 1999 was the most recent year for which regional economic data on
earnings and per capita income were available. Data

Variables Used in the Model

Appendix V: Economic Factors Affect Air Service in Small Communities

Page 58 GAO- 02- 432 Small Community Air Service

Using the regression to explain variation in air service, we focused
primarily on modeling the number of weekly departures (jet and turboprop)
from a small community. 48 Multiple univariate and multivariate models of
jet and turboprop departures were specified as a function of the independent
variables to examine the consistency and robustness of the findings. The
results of a final model are discussed below, in which jet and turboprop
departures are specified as a function of employment (or population),
manufacturing earnings, minimum distance to a low- fare carrier, and per
capita income. 49

The results of our regression models indicate that, as expected, employment
(or population), manufacturing earnings, minimum distance from a low- fare
carrier, and per capita income had a positive effect on the level of air
service received by a small community. Below are quantitative statistics
from specific models. 50

 After controlling for distance to a low- fare carrier, manufacturing
earnings, and population, we found that for every additional $5,000 in per
capita income, a community received 3.3 and 12.7 more jet and turboprop 51

48 Models using the number of major carriers and the percentage of total
seats available on jets as dependent variables were also estimated. In
general, we found that employment, per capita income, population, and
distance from a low- fare carrier positively affected the number of major
carriers serving a community and the percentage of total seats served by
jets.

49 The employment and population variables are highly collinear. Therefore,
to avoid biased parameter estimates due to multicollinearity, ordinary least
squares regression models include only one of the two factors as independent
variables.

50 The quantitative data discussed are all statistically significant at (at
least) the 95 percent level. Only 200 observations were included in the
regression analysis because two small communities did not have service in
October 2000.

51 Two separate regressions were estimated using jet and turboprop
departures as the dependent variables. Results are similar when the
employment variable is substituted for the population variable - 2.6 more
jet and 12. 2 turboprop departures per week, respectively. These results are
statistically significant at least at the 90 percent level. Model
Specification

Results

Appendix V: Economic Factors Affect Air Service in Small Communities

Page 59 GAO- 02- 432 Small Community Air Service

departures per week, respectively. 52 In other words, if two small
communities, A and B, were identical in every way except that Community A
had $5,000 more in per capita income than Community B, then Community A had
roughly 16 more total departures per week than Community B. This difference
in the number of total departures was attributable to the difference in per
capita income.  After controlling for distance to a low- fare carrier,
manufacturing

earnings, and per capita income, we found that a community received 4.3 and
4.8 more jet and turboprop departures per week respectively for every
additional 25,000 jobs in the community.  After controlling for distance to
a low- fare carrier, population, and per

capita income, we found that a community with $250,000 more in manufacturing
earnings received 4.8 more jet departures per week than an otherwise similar
community.  After controlling for manufacturing earnings, per capita
income, and

employment, we found that a community received 4.7 more jet departures per
week for every additional 50 miles separating the airport from a lowfare
carrier. 53

52 The value of the coefficients of determination in all regression models
(R- squared) were consistently below 0.20, denoting that the independent
variables explain less than 20 percent of the variation in air service to
the communities. As noted previously, airlines decide to provide service to
these small communities on a market- by- market basis using individual
market data not available for our study. Through our use of regression
analysis, we are attempting to describe service generally across the
different communities using limited aggregate data. While the data restrict
our analysis to modeling broadly defined differences in air service, we do
not believe that those data limitations detract from the findings.

53 Our findings also revealed that a community that is 50 miles from an
airport with a lowfare carrier loses 5.0 turboprop departures per week-
resulting in a net loss of departures.

Appendix VI: Changes in Air Service at Small Communities Made By Express
Airlines I and Mesaba Between 2000 and 2001

Page 60 GAO- 02- 432 Small Community Air Service

According to our analysis, between October 2000 and October 2001, Express
Airlines I and Mesaba altered service at 60 small communities. Overall, more
small communities lost service than gained service from these carriers
during this period. A total of 49 small communities lost service, 4 of which
(Bismarck, North Dakota; Columbus, Georgia; Dothan, Alabama; and Rockford,
Illinois) lost all nonstop service from Express Airlines I and Mesaba. 54 On
the other hand, 11 communities gained service. Nine gained additional
flights or extra capacity (i. e., number of seats available for purchase)
through larger aircraft, and two (Charlottesville, Virginia, and
Springfield, Illinois) gained start- up service from the two airlines.

 Express Airlines I made service changes at 27 small communities between
2000 and 2001. Of these 27 communities, Express Airlines I reduced service
at 13, increased service at 13, and took mixed actions at 1 other (reducing
the number of daily departures but adding more available seating capacity by
using larger aircraft).  Mesaba altered its weekly service at 44 small
communities between 2000

and 2001. Mesaba ended all service to 2 communities. At 37 other
communities, Mesaba?s service reductions averaged two departures per day per
community. On the other hand, Mesaba increased service at 3 small
communities and launched new service at another. At Sioux City, Iowa, Mesaba
decreased average daily departures but increased total seating capacity by
substituting larger aircraft.  Eleven communities were served by both-
Mesaba and Express Airlines I.

Service reductions that Mesaba made at 8 of the 11 were offset by service
additions from Express Airlines I, often with new regional jet service.

Table 12 summarizes the small community service changes made by Express
Airlines I and Mesaba between October 2000 and October 2001.

54 Northwest maintains mainline service at Bismarck. Appendix VI: Changes in
Air Service at Small

Communities Made By Express Airlines I and Mesaba Between 2000 and 2001

Appendix VI: Changes in Air Service at Small Communities Made By Express
Airlines I and Mesaba Between 2000 and 2001

Page 61 GAO- 02- 432 Small Community Air Service

Table 12: Nonstop Air Service Changes for Communities Served by Mesaba and
Express Airlines I Between October 2000 and October 2001

Lost all service (4) Lost capacity (45)

Gained capacity (9) Gained new

service (2)

Columbus, GA (Express Airlines I) Appleton, WI

(Mesaba) a Alpena, MI (Mesaba, Express Airlines I) Hibbing, MN

(Mesaba) Mosinee, WI (Express Airlines I) a Springfield, IL

(Express Airlines I) Dothan, AL (Express Airlines I) Kalamazoo, MI

(Mesaba) a Binghamton, NY (Mesaba) La Crosse, WI

(Mesaba) Lincoln, NE (Express Airlines I) a Charlottesville, VA

(Mesaba) Bismarck, ND (Mesaba) Bloomington, IL

(Mesaba) a Bemidji, MN (Mesaba) Mason City, IA (Mesaba)

Saginaw, MI (Express Airlines I/ Mesaba) b Rockford, IL (Mesaba, Express
Airlines I)

Alexandria, LA (Express Airlines I) Brainerd, MN

(Mesaba) Peoria, IL (Mesaba) Sioux City, IA (Express Airlines I/

Mesaba) d Chattanooga, TN (Express Airlines I) Sault Ste. Marie, MI

(Mesaba) Pierre, SD (Mesaba) Joplin, MO (Express Airlines I) Evansville, IN
(Mesaba) a, e Champaign, IL

(Mesaba) Pellston, MI (Mesaba) Montgomery, AL

(Express Airlines I) d Fort Smith, AR (Express Airlines I) Hancock, MI
(Mesaba) Rhinelander, WI

(Mesaba) Tupelo, MS (Express Airlines I) Greenville, MS (Express Airlines I)
Charleston, WV

(Mesaba) Rochester, MN (Mesaba) Duluth, MN

(Mesaba) Columbus, MS (Express Airlines I) Eau Claire, WI

(Mesaba) Toledo, OH (Mesaba) a International Falls,

MN (Mesaba) Lafayette, LA (Express Airlines I) Elmira, NY (Mesaba) Traverse
City, MI

(Mesaba) a Monroe, LA (Express Airlines I) Erie, PA (Mesaba) Fayetteville,
AR

(Express Airlines I, Mesaba) c Paducah, KY (Express Airlines I) Escanaba, MI

(Mesaba) State College, PA (Mesaba) Panama City, FL (Express Airlines I)
Fort Dodge, IA

(Mesaba) Waterloo, IA (Mesaba) Hattiesburg, MS (Express Airlines I) Grand
Forks, ND

(Mesaba) Watertown, SD (Mesaba) Aberdeen, SD (Mesaba) Grand Rapids, MN

(Mesaba) Fargo, ND (Mesaba) a Communities where Express Airlines I replaced
Mesaba service.

b Communities where both Express Airlines I and Mesaba increased service. c
Communities where Express Airlines I discontinued all service, while Mesaba
decreased service. d Community where carriers decreased frequency, but
increased capacity. e Community where carrier decreased frequency, but
experienced no net change in seat capacity.

Source: GAO analysis of airline schedule data from the Kiehl Hendrickson
Group.

Appendix VII: Changes in Air Service at Small Communities Made By Great
Lakes Aviation and Air Wisconsin Between 2000 and 2001

Page 62 GAO- 02- 432 Small Community Air Service

Of the 202 small communities in our study, Great Lakes Aviation served 16,
Air Wisconsin served 15, and both airlines served 9. Both airlines served
United?s Chicago (O?Hare) and Denver hubs. Between October 2000 and October
2001, Great Lakes and Air Wisconsin altered air service in 39 communities.
Of the 39 communities with service changes, 4 lost all of their air service
(all of which was provided by Great Lakes). A total of 30 communities saw
reductions in their service (i. e., capacity, either through a reduction in
departures or by using smaller aircraft) by Great Lakes or Air Wisconsin.
Five of the communities in our analysis gained either new service or
capacity (i. e., number of seats available for purchase).

 Great Lakes altered its weekly capacity at 16 small communities between
2000 and 2001. Of these communities, 4 of them (Dubuque, Iowa; Lafayette,
Indiana; Rhinelander, Wisconsin; and Salina, Kansas) lost all of their
service. Furthermore, Great Lakes reduced service at 11 communities. Only
one community- Telluride, Colorado- gained capacity from Great Lakes.  Air
Wisconsin reduced service at 11 communities and added either new

service or additional capacity in 3 communities.  Great Lakes and Air
Wisconsin both served 9 communities in our analysis.

 Between October 2000 and October 2001, in 3 of those communities (Traverse
City, Michigan; Springfield, Illinois; and Eagle, Colorado), Air Wisconsin
replaced Great Lakes service, and in one community (Cody, Wyoming) Great
Lakes replaced Air Wisconsin service.  Both Great Lakes and Air Wisconsin
provided service to Grand

Junction and Durango, Colorado. Great Lakes discontinued service by October
2001.  Casper, Wyoming; Hayden and Gunnison, Colorado all were receiving

service from both Great Lakes and Air Wisconsin in 2000. By October 2001,
Air Wisconsin had discontinued all of its service.

Table 13 summarizes the changes in service at small communities served by
Great Lakes and Air Wisconsin between October 2000 and October 2001.
Appendix VII: Changes in Air Service at Small

Communities Made By Great Lakes Aviation and Air Wisconsin Between 2000 and
2001

Appendix VII: Changes in Air Service at Small Communities Made By Great
Lakes Aviation and Air Wisconsin Between 2000 and 2001

Page 63 GAO- 02- 432 Small Community Air Service

Table 13: Nonstop Air Service Changes for Communities Served by Great Lakes
and Air Wisconsin Between October 2000 and October 2001

Lost all service (4) Lost capacity (30) Gained capacity (4) Gained new

service (1) No change (1)

Dubuque, IA (Great Lakes) Waterloo, IA (Great

Lakes) Williston, ND (Great Lakes)

Bismarck, ND (Air Wisconsin)

Lincoln, NE (Air Wisconsin)

Billings, MT (Air Wisconsin) Lafayette, IN (Great Lakes) Aspen, CO

(Air Wisconsin) Jackson, WY

(Air Wisconsin) Mosinee, WI

(Air Wisconsin) Rhinelander, WI (Great Lakes) Appleton, WI (Air

Wisconsin) Montrose, CO (Air Wisconsin)

Telluride, CO (Great Lakes) Salina, KS

(Great Lakes) Kalamazoo, MI

(Air Wisconsin) Peoria, IL

(Air Wisconsin) Eagle, CO

(Great Lakes/ Air Wisconsin) b Scottsbluff, NE (Great Lakes) Pierre, SD

(Great Lakes) Bloomington, IL (Air Wisconsin)

Rapid City, SD (Air Wisconsin) Cheyenne, WY (Great Lakes) Riverton, WY

(Great Lakes) Durango, CO (Great Lakes/ Air Wisconsin) d

Santa Fe, NM (Great Lakes)

Fargo, ND (Air Wisconsin)

Sheridan, WY (Great Lakes) Farmington, NM (Great Lakes) Springfield, IL
(Great

Lakes/ Air Wisconsin) a Gillette, WY (Great Lakes) Bristol/ Johnson/ King,
TN

(Air Wisconsin) Grand Junction, CO (Great Lakes/ Air Wisconsin) d

Traverse City, MI (Great Lakes/ Air Wisconsin) a Grand Island, NE (Great
Lakes)

Quincy, IL (Air Wisconsin) Cody, WY (Great Lakes/ Air Wisconsin) c

Gunnison, CO (Great Lakes/ Air Wisconsin) e Casper, WY (Great Lakes/ Air
Wisconsin) e

Hayden, CO (Great Lakes/ Air Wisconsin) e a Communities where Air Wisconsin
replaced Great Lakes service and resulted in decreased capacity.

b Communities where Air Wisconsin replaced Great Lakes service and resulted
in increased capacity. c Communities where Great Lakes replaced Air
Wisconsin service, and total capacity decreased. d Communities where Great
Lakes and Air Wisconsin initially provided service but Great Lakes
discontinued service, resulting in decreased capacity for the community. e
Communities where Great Lakes and Air Wisconsin initially provided service
but Air Wisconsin

discontinued service, resulting in decreased capacity for the community.
Source: GAO analysis of airline schedule data from the Kiehl Hendrickson
Group.

Appendix VIII: GAO Contacts and Staff Acknowledgments

Page 64 GAO- 02- 432 Small Community Air Service

JayEtta Z. Hecker (202) 512- 2834 Steven C. Martin (202) 512- 2834

In addition to those individuals named above, Triana Bash, Curtis Groves,
Dawn Hoff, David Hooper, Sara Ann Moessbauer, John Mingus, Ryan Petitte,
Carl Ramirez, Sharon Silas, Stan Stenersen, and Pamela Vines made key
contributions to this report. Appendix VIII: GAO Contacts and Staff

Acknowledgments GAO Contacts Acknowledgments

Related GAO Products Page 65 GAO- 02- 432 Small Community Air Service

Financial Management: Assessment of the Airline Industry?s Estimated Losses
Arising From the Events of September 11. GAO- 02- 133R. Washington, D. C.:
October 5, 2001.

Commercial Aviation: A Framework for Considering Federal Financial
Assistance. GAO- 01- 1163T. Washington, D. C.: September 20, 2001.

Aviation Competition: Restricting Airline Ticketing Rules Unlikely to Help
Consumers. GAO- 01- 831. Washington, D. C.: July 31, 2001.

Aviation Competition: Regional Jet Service Yet to Reach Many Small
Communities. GAO- 01- 344. Washington, D. C.: February 14, 2001.

Essential Air Service: Changes in Subsidy Levels, Air Carrier Costs, and
Passenger Traffic. GAO/ RCED- 00- 34. Washington, D. C.: April 14, 2000.

Airline Deregulation: Changes in Airfares, Service Quality, and Barriers to
Entry. GAO/ RCED- 99- 92. Washington, D. C.: March 4, 1999.

Aviation Competition: Effects on Consumers From Domestic Airline Alliances
Vary. GAO/ RCED- 99- 37. Washington, D. C.: January 15, 1999. Related GAO
Products

(544000)

The General Accounting Office, the investigative arm of Congress, exists to
support Congress in meeting its constitutional responsibilities and to help
improve the performance and accountability of the federal government for the
American people. GAO examines the use of public funds; evaluates federal
programs and policies; and provides analyses, recommendations, and other
assistance to help Congress make informed oversight, policy, and funding
decisions. GAO?s commitment to good government is reflected in its core
values of accountability, integrity, and reliability.

The fastest and easiest way to obtain copies of GAO documents is through the
Internet. GAO?s Web site (www. gao. gov) contains abstracts and full- text
files of current reports and testimony and an expanding archive of older
products. The Web site features a search engine to help you locate documents
using key words and phrases. You can print these documents in their
entirety, including charts and other graphics.

Each day, GAO issues a list of newly released reports, testimony, and
correspondence. GAO posts this list, known as ?Today?s Reports,? on its Web
site daily. The list contains links to the full- text document files. To
have GAO e- mail this list to you every afternoon, go to www. gao. gov and
select "Subscribe to daily e- mail alert for newly released products" under
the GAO Reports heading.

The first copy of each printed report is free. Additional copies are $2
each. A check or money order should be made out to the Superintendent of
Documents. GAO also accepts VISA and Mastercard. Orders for 100 or more
copies mailed to a single address are discounted 25 percent. Orders should
be sent to:

U. S. General Accounting Office P. O. Box 37050 Washington, D. C. 20013

To order by Phone: Voice: (202) 512- 6000 TDD: (202) 512- 2537 Fax: (202)
512- 6061

GAO Building Room 1100, 700 4th Street, NW (corner of 4th and G Streets, NW)
Washington, D. C. 20013

Contact: Web site: www. gao. gov/ fraudnet/ fraudnet. htm, E- mail:
fraudnet@ gao. gov, or 1- 800- 424- 5454 or (202) 512- 7470 (automated
answering system).

Jeff Nelligan, Managing Director, NelliganJ@ gao. gov (202) 512- 4800 U. S.
General Accounting Office, 441 G. Street NW, Room 7149, Washington, D. C.
20548 GAO?s Mission

Obtaining Copies of GAO Reports and Testimony

Order by Mail or Phone Visit GAO?s Document Distribution Center

To Report Fraud, Waste, and Abuse in Federal Programs

Public Affairs
*** End of document. ***