Case Study 3 Chapter 5

1
Case Study 3 Chapter 5

• Airline Cost and Production Under Regulation

2
History of Airline Regulation

• 1926-Air Mail Act
• Responsibility for air mail shifted from Army to
  USPSwho contracted it to private sector
• Purpose was to expand nations air routes and
  passenger services
• Achieved its intended effect
• airlines increased from 13 to 38 by 1930
• Revenue passenger miles increased from 1.3 to
  85.1 billion

3
Regulatory AuthorityICC CAA

• USPS responsible for allocation of mail contracts
• ICC given authority to set air mail rates
• ICC set rates below cost, but tied new route
  allocations to acceptance of new rates
• Result, the industry became financial unstable
• Authority shifted from ICC to CAA in 1938
• CAA encouraged mergers in an attempt to increase
  scale of operations and reduce LRAC
• CAAct specified rules for entry/exit, market
  access, route restrictions, rate-setting and
  subsidization

4
CAAact Entry/Exit Controls

• 16 existing intercity airlines were granted
  continuing authority to operate on same routes
• To gain entry, new airlines had to show a need
  existed and
• It could provide efficient, economical and safe
  service
• No new airlines allowed entry between 1938 end
  of regulation in 1970s
• Exit from industry was also strictly controlled
• Airlines were not allowed to discontinue, or
  reduce services, on any of its routes
• Airlines response was to consolidate
• The merger process allowed firms to enter new
  markets and leave unprofitable markets

5
CAAct Route Restrictions

• CAAact granted existing carriers operating
  authority (OA) on their existing routes
• Under CAB (replaced CAA) existing airlines
  allowed to extend service into new markets
• Carrier chosen was usually based on its financial
  condition, rather than
• its ability to provide the most efficient and
  economical service
• The general goal was to equalize profitability
  across airlines

6
Emergence of Local Carriers

• CAB granted temporary OA to provide feeder
  service to trunk airlines 40s-50s
• Feeders were subsidized
• Local feeders were granted permanent OA in 1955
• Over time local airlines granted some
  overlapping authority with trunk airlines
• Reduced the subsidies to local airlines
• Allowed them to evolve into regional carriers
  that competed with trunk airlines
• ExamplesFlorida Air, Piedmont, Eastern,
  Southwest, Northwest, etc.

7
The ProblemDifferential Effects of
RegulationTrunks vs Locals

• Prior to deregulation a two-tier structure
  existed in Airline Industry
• Locals serving smaller communities on shorter
  flights in smaller planes with smaller networks
• Trunks serving major intercity markets with
  longer flights, larger planes and networks
• Significant differences in technology and cost
  structure existed
• Empirical questions
• How significant were cost/technology differences
• What specific factors explain the differences

8
Estimating Airline Cost Functions

• Continuing to use translog flexible cost
  specification
• ln C(Tp,o,nw) function of output, three input
  prices, two airline operating characteristics and
  network size
• C is total cost in 000s and
• T is Revenue-output milesa composite index
  measure
• NW is network size measured by Avg. number of
  points served
• Characteristics representing airlines technology
• Avg. stage length (asl) length of flights in
  miles/number flights
• Avg. loads (ald) seat-miles sold/seat-miles
  available
• Inputs are labor, capital fuel

9
Hypotheses 1 2 Returns to Scale

• H-1Generalized returns to scale are constant
• Generalized economies of scale refers to changes
  in both network size and level of all inputs,
  holding pi and oj constant
• Test by combining the coefficients of T and NW
• H-2Increasing returns to traffic density,
  holding network size constant
• Test by using coefficient of output (T) only

10
Hypothesis 3 Input Prices Positively Related to
Total Cost

• The null hypotheses are
• Side result is
• Each input prices coefficient represents its
  share of total cost
• The sum of shares must equal 1

11
Hypothesis 4 Increases in asl ald Reduce Total
Cost

• Increases in asl, ceteris paribus
• Ground costs are spread over more revenue-miles
• Total stacking delay costs are reduced
• During regulatory period, most airports were
  fixed-slot and stack delays (a.k.a. holding
  patterns) were common
• Use of larger planes on long routes also reduces
  cost per revenue mile
• Increases in ald, cet par reduce avg. costs
• Represents an increase in capacity utilization
  and efficiency
• Empty planes cost as much to fly as full planes
• Ground costs can be spread over more revenue
  miles

12
RTS in Airlines?

• Generalized Returns to ScaleNW all inputs are
  variable
• Coefficients for T NW are .80 and .132,
  respectively
• Sum these and take reciprocal yields 1.07implies
  CRTS
• Returns to Scale With Fixed Network sizeReturns
  to Density
• Coefficient for T only .80 yields RTS
  1.25IRTS
• Means that inability to increase output on a
  fixed network inhibited airlines from achieving
  cost efficiency in regulatory period.
• Returns to Network sizeall inputs constant
• Reciprocal of NW coefficient 7.6 implying
  strongly IRT NW size
• Suggests that if unregulated, airlines could
  significantly reduce their costs by increasing
  number of points served

13
Input Shares and Substitution

• Input price coefficients show proportion of total
  cost represented by each input
• Largest share (48) goes to capitalSuggesting
  the importance of large capital investment in
  planes and facilities
• Labor accounts for 35 of cost
• Fuels share of total cost is 17
• Elasticities of Substitution are mixed
• Elast of Sub between labor capital are positive
  suggesting some substitutiondoes require as
  large a crew to flight a small plane
• Elast of Sub between labor and fuel suggests a
  complementary relationshipthe more pilot hours
  the more fuel required
• Elast of Sub between capital and fuel suggests
  substitutessuggests larger planes are more
  economical per revenue passenger mile

14
Operating Characteristics

• Both asl and ald are negative and significant
• als suggests that a 10 increase in length of
  flight reduces cost by 1.5
• ald had much larger effect on costs than als
• Comparison of characteristics suggests more cost
  savings by flying fuller rather than longer
  flights

15
Trunks vs. Locals

• Both RTS and returns to density were close in
  both sectors
• Average number of points served were also similar
• The biggest difference was between length of
  flights
• Trunks asl was four times larger than locals
• Load factor was also 21 higher for trunks
• These latter two factors account for differences
  in operating costs of the trunks and locals
  during regulation