Air Travel Demand Elasticities

AIR TRAVEL DEMAND ELASTICITIES:
CONCEPTS, ISSUES AND MEASUREMENT
FINAL REPORT
David W. Gillen
School of Business & Economics
Wilfrid Laurier University
Waterloo, Canada
William G. Morrison
Waterloo, Canada
Christopher Stewart MBA

Waterloo, Canada
TABLE OF CONTENTS
Executive Summary.
1. Introduction.
2. Elasticity in the context of air travel demand.
3. Measurement issues.
4. Evaluation of elasticity studies.
5. Discussion.
Bibliography.
Appendix A: survey of demand elasticity studies.

EXECUTIVE SUMMARY
This document reports the findings of a review of the economics and business literature on empirically-estimated own-price elasticities of demand for air travel for
Canada and other major developed countries. It does not explicitly measure the impact of the Air Travellers Security Charge on air travel in Canada, nor does it assess
how the enhanced air travel security system has affected the demand for air travel.
WHAT IS A PRICE ELASTICITY OF DEMAND?
The demand for a particular good or service depends on a variety of factors. Key influences include the tastes of consumers, the levels of consumer income, the price
and quality of the product in question and the prices of other goods, especially goods that are close substitutes. In order to obtain useful estimates of the price
sensitivity of demand for a product, researchers must carefully control for all the factors affecting the demand.
As a general rule, when other influences on demand remain unchanged, a higher price for a product results in a lower quantity demanded. However, the price
responsiveness of demand varies from one good to another and from one market to another. The own-price elasticity of demand measures the responsiveness, or
sensitivity, of the demand for a good to changes in its price when other influences on demand are held constant. It is defined as the percentage change in quantity
demanded resulting from a given percentage change in price.
For example, if a 1% increase in price results in a 1.2% decrease in quantity demanded, the own-price elasticity of demand is 1.2. In this case, since the percentage
fall in demand is greater than the percentage rise in price, total spending on the good will decline, and the demand for it is said to be "elastic". If, on the other hand, a
1% price rise causes a smaller percentage decline in the quantity demanded, the own-price elasticity will be less than one, and demand is said to be "inelastic".
KEY FINDINGS OF THE REPORT
This report identifies six distinct markets for air travel. Specifically, it is observed that studies of the demand for air travel demand should distinguish among markets
for: business and leisure travel; long-haul and short-haul travel; and international and North American long-haul travel. Accordingly, to examine the sensitivity of the
demand for air travel to its price, separate estimates of the own-price elasticity of demand are gathered for each of these distinct markets.
Since the availability of alternative modes of transportation that are reasonably close substitutes for air transport diminishes with distance travelled, it is expected that
the demand for air transport will be less elastic for longer flights than for shorter flights. Further, international travel tends to be spread over more time than domestic
travel, so that the airfare is a smaller proportion of overall trip costs, which makes international travel less sensitive to changes in ticket prices. In addition, leisure
travellers are more likely to postpone trips to specific locations in response to higher fares, or to shop around for those locations offering more affordable fares.
Consequently, it is expected that the demand for air transport for leisure reasons will be more elastic than business travel.
Standardized information on 21 Canadian and international empirical studies of own-price elasticities of demand for air travel is provided in this report. This information
includes the author(s), study date, country/countries considered, model type and functional form, type and source of data, and markets examined (e.g., business, long
haul, and international). A scoring system based on desirable characteristics of demand studies is applied to each of the 21 studies to improve the level of confidence
regarding the practical use of own-price elasticity values in assessing air travel demand. For example, recent studies are more highly rated, as are studies that control
for changes in consumer income, and, for short-haul markets, the availability of alternative travel modes. This methodology identifies 11 studies that are expected to
provide particularly relevant estimates for different markets.
Summary results from the report are reproduced in Chart 1. Median [1] elasticity values differ significantly by type of traveller, travel distance and domestic/international
routes, and confirm the existence of distinct air travel markets. They generally accord with expectations that the demand for air transport should be less elastic for
business than for leisure travellers, and less elastic for flights of longer distances, notably for international travel. Based on available studies, however, markets for
business travel do not show a consistently increasing responsiveness as distance travelled becomes shorter. The chart also shows, for each market, the range of
elasticity values within which one-half of the estimates lie.[2]
There are two important caveats to the interpretation and use of these estimates for purposes of assessing the responsiveness of air transport demand in Canada to
the ATSC.
First, the median and range of values for some of the distinct markets for air travel are based on a relatively small number of estimates and studies. For
example, values for the own-price elasticity of demand for long-haul domestic leisure are based on six estimates from two studies.
Second, very few of the studies included in the report use Canadian data to estimate own-price elasticities of demand.
Chart 1 - Own-Price Elasticities of Demand
The ranges of values shown capture the middle one-half of the estimates and encompass the median, which is represented by a black dot.
Consequently, the median value for a particular market is not necessarily representative of the corresponding market in Canada. It is argued in the report that analysts
should instead consider the broader range of values around the median value when examining the impact of price changes on the demand for air travel in Canada.
1. INTRODUCTION
The purpose of this study is to report on all or most of the economics and business literature dealing with empirically estimated demand functions for air travel and to
collect a range of fare elasticity measures for air travel and provide some judgment as to which elasticity values would be more representative of the true values to be
found in different markets in Canada.
While existing studies may include the leisure business class split, other important market distinctions are often omitted, likely as a result of data availability and
quality.[3] One of the principal value added features of this research and what distinguishes it from other surveys, is that we develop a meta-analysis that not only
provides measures of dispersion but also recognizes the quality of demand estimates based on a number of selected study characteristics. In particular, we develop a
means of scoring features of the studies such as focus on length of haul; business versus leisure; international versus domestic; the inclusion of income and inter-
modal effects; the age of the study; data type (time-series versus cross section) and the statistical quality of estimates (adjusted R-squared values). By scoring the
studies in this way, policy makers are provided with a sharper focus to aid in judging the relevance of various estimated elasticity values.[4]
2. ELASTICITY IN THE CONTEXT OF AIR TRAVEL DEMAND.
Elasticity values in economic analysis provide a "units free" measure of the sensitivity of one variable to another, given some pre-specified functional relationship. The
most commonly utilized elasticity concept is that of "own-price" elasticity of demand. In economics, consumer choice theory starts with axioms of preferences over
goods that translate into utility values. These utility functions define choices that generate demand functions from which price elasticity values can be derived.
Therefore elasticities are summary measures of people's preferences reflecting sensitivity to relative price levels and changes in a resource-constrained environment.
The ordinary or Marshallian demand function is derived from consumers who are postulated to maximize utility subject to a budget constraint. As a good's price
changes, the consumer's real income (which can be used to consume all goods in the choice set) changes. In addition the goods price relative to other goods changes.
The changes in consumption brought about by these effects following a price change are called income andsubstitution effects respectively. Thus, elasticity values
derived from the ordinary demand function include both income and substitution effects. [5]
Own-price elasticity of demand measures the percentage change in the quantity demanded of a good (or service) resulting from a given percentage change in the
good's own-price, holding all other independent variables (income, prices of related goods etc.) fixed. The ratio of percentage changes thus allows for comparisons
between the price sensitivity of demand for products that might be measured in different units (natural gas and electricity for example). 'Arc' price elasticity of demand
calculates the ratio of percentage change in quantity demanded to percentage change in price using two observations on price and quantity demanded. Formally this
can be expressed as:
(1)
where:
represent the observed change in quantity demanded and price
represent the average price and quantity demanded. The elasticity is unitless and can be interpreted as an index of demand sensitivity; it is
measuring the degree to which a variable of interest will change (passenger traffic in our case) as some policy or strategic variable changes (total fare including any
added fees or taxes in our case).
In the limit (when are very small) we obtain the 'point' own-price elasticity of demand expressed as:
(2)
where:
Q(P,S) is the demand function

P = a vector of all relevant prices
p = the good's own-price.
q = equals the quantity demanded of the good
S = a vector of all relevant shift variables other than prices (real income, demographic characteristics etc.)
We expect own-price demand elasticity values to be negative, given the inverse relationship between price and quantity demanded implied by the 'law' of demand, with
absolute values less than unity indicating 'inelastic' demand: a less than proportionate response to price changes (relative price insensitivity). Similarly, absolute values
exceeding unity indicate elastic or more sensitive demand: a more than proportionate demand response to price changes (relative price sensitivity).
The ratio of change in quantity demanded to change in price [equation (1)] highlights that elasticity measures involve linear approximations of the slope of a demand
function. However, since elasticity is measuring proportionate change, elasticity values will change along almost all demand functions, including linear demand
curves.[6] Estimation of elasticity values is therefore most useful for predicting demand responses in the vicinity of the observed price changes. As a related issue,
analysts need to recognize that in markets where price discrimination is possible aggregate data will not allow for accurate predictions of demand responses in the
relevant market segments. In air travel, flights by a carrier are essentially joint products consisting of differentiated service bundles that are identified by fare classes.
However the yield management systems employed by full-service carriers (FSCs) also create a complex form of inter-temporal price discrimination, in which some fares
(typically economy class) decline and some increase (typically full-fare business class) as the departure date draws closer. This implies that ideally, empirical studies of
air travel demand should separate business and leisure travellers or at least be able to include some information on booking times in order to account for this price
discrimination, and that price data should be calibrated for inter-temporal price discrimination: for example, the use of full-fare economy class ticket prices as data will
overestimate the absolute value of the price elasticity coefficient. Within the set of differentiated service bundles that comprise each (joint product) flight, the relative
prices are important in explaining the relative ease of substitution between service classes. Given the nature of inter-temporal price discrimination for flights, the
relative price could also change significantly in the time period prior to a departure time.
The partial derivative in (2) indicates that elasticity measures price sensitivity independent of all the other variables in the demand function. However when estimating
demand systems over time, one can expect that some important shift variables will not be constant. It is important that these shift variables be explicitly recognized
and incorporated into the analysis, as they will affect the value of elasticity estimates. This will also be true with some cross-sectional studies or panels.[7] In particular
changes in real income and the prices of substitutes or complements will affect demand. In air travel demand estimations, income and prices of other relevant goods
should be included in the estimation equation. Alternative transportation modes (road and rail) are important variables for short-haul flights, while income effects
should be measured for both short and long-haul. The absence of an income coefficient in empirical demand studies will result in own-price elasticity estimates that can
be biased. With no income coefficient, observed price and quantity pairs will not distinguish between movements along the demand curve and shifts of the demand
curve.[8]
The slope of a demand function, which affects the own-price elasticity of demand, is generally expected to decrease (become shallower) with:
The number of available substitutes;
The degree of competition in the market or industry;
The ease with which consumers can search and compare prices;
The homogeneity of the product;
The duration of the time period analyzed.[9]
Given the implied relationships above, any empirical demand study should carefully define market boundaries to include all relevant substitutes and complements and
to exclude products that might be related through income or other more general variables.
In air travel, ideally market segment boundaries should be defined by first separating leisure and business passengers and second long-haul and short-haul flights. The
reason is that we expect different behaviour in each of these markets. Within each of these categories, distinctions should then be made between the following:
Connecting and origin-destination (O-D) travel;
Hub and non-hub airports;[10]
Routes with dominant airlines and routes with low-cost carrier competition.
In addition, for the North American context, long-haul flights should be further divided into international and domestic travel (within continental North America). These
market segment boundaries are illustrated in figure 2.1 below, which also highlights the relative importance of intermodal competition for short-haul travel.
While distinctions in price and income sensitivity of demand between business and leisure or long and short-haul travel are more intuitive, other distinctions are
perhaps less obvious. If available, data that distinguishes between routes, airlines and airports would provide important estimates of how price sensitivity is related to
the number of competing flights and the willingness to pay of passengers utilizing a hub-and-spoke network, relative to those traveling point-to-point, more commonly
associated with low cost carriers. To the extent that existing studies assume that each passenger observation represents O-D travel, they will not be capturing fare
premiums usually associated with hub-and-spoke networks and full service carriers, nor will they necessarily capture the complete itinerary of travellers utilizing a
number of point-to-point flights with a low cost carrier. For example, a passenger who travels from Moncton to Vancouver with Air Canada, and utilizes the hub at
Pearson International airport, is being provided with a number of services that includes baggage checked through to the final destination and frequent flyer points as
well as a choice in flights and added flight and ground amenities. The fare for Moncton-Vancouver includes a premium for these services. Now consider a passenger
that is travelling with WestJet from Moncton to Hamilton, and then with JetsGo from Toronto Pearson Airport to Vancouver. In this case there are no frequent flyer
points to be attained and baggage has to be collected and re-checked after a road transfer between Hamilton and Pearson International. Although the origin and
destination is the same for these passengers, the itineraries are significantly different. In many cases data used for demand estimates would not able to account for
these differences.
Route-specific data can also capture competition that may exist between airports and the services they offer as well as airlines. This may be especially true for certain
short-haul routes where intermodal competition (road and rail) can play an important role in shaping air travel demand.
Figure 2.1
Market segments in air travel demand.
3. MEASUREMENT ISSUES
Oum et al. (1992) provide a valuable list of pitfalls that occur when demand models are estimated and therefore affect the interpretation of the elasticity estimates
from these empirical studies.
1. Price and Service Attributes of Substitutes: Air travel demand can be affected by changes in the prices and service quality of other modes. For short-haul routes
(markets) the relative price and service attributes of auto and train would need to be included in any model; particularly for short-haul markets. Failure to include the
price and service attributes of substitutes will bias the elasticity. For example, if airfares increase and auto costs are also increasing, the airfare elasticity would be
overestimated if auto costs were excluded.
2. Functional Forms: Most studies of air travel demand use a linear or log-linear functional specification. Elasticity estimates can vary widely depending on the
functional form. The choice of functional form should be selected on the basis of statistical testing not ease of interpretation.
3. Cross-Section vs. Time-series Information: In the long run demand elasticities for non-durable goods and services are larger in absolute terms, than in the short run.
This follows because in the long run there are many more substitution possibilities that can be used to avoid price increases or service quality decreases. In effect there
are more opportunities to avoid these changes with substitution possibilities. Data tends to be cross-sectional or time-series although more recently panels have
become available. A panel is a combination of cross-section and time-series information on several routes for a multi-year period is a panel. Cross-sectional
information is generally regarded as indicating short run elasticities while time-series data is interpreted as long run elasticities. In time-series data the information
reflects changes in markets, growth in income, changes in competitive circumstances, for example. Policy changes should rely on long run elasticities since these are
long run impacts that are being modelled. Short run elasticities become important when considering the competitive position of firms in a highly dynamic and
competitive industry.
4. Market Aggregation/Segmentation: As the level of aggregation increases the amount of variation in the elasticity estimates decreases. This occurs because
aggregation averages out some of the underlying variation relating to specific contexts. Since air travel market segments may differ significantly in character,
competition and dominance of trip purpose, interpreting a reduction in variation through aggregation as a good thing would be erroneous. Such estimates might have
relatively low standard deviations but would be also be relatively inaccurate when used to assess the effect of changes in fares in a specific market.
5. Identification Problem: In most cases only demand functions are estimated in attempts to measure the demand elasticity of interest. However, it is well known that
the demand function is part of a simultaneous equations system consisting of both supply and demand functions. Therefore, a straightforward estimation of only the
demand equation will produce biased and inconsistent estimates. The problem of identification can be illustrated by describing the process by which fares and travel,
for example, are determined in the origin-destination market simultaneously. To model this process in its entirety, we must develop a quantitative estimate of both the
demand and supply functions in a system. If, in the past, the supply curve has been shifting due to changes in production and cost conditions for example, while the
demand curve has remained fixed, the resultant intersection points will trace out the demand function. On the contrary, if the demand curve has shifted due to changes
in personal income, while the supply curve has remained the same, the intersection points will trace out the supply curve. The most likely outcome, however, is
movement of both curves yielding a pattern of fare, quantity intersection points from which it will be difficult, without further information, to distinguish the demand
curve from the supply curve or estimate the parameters of either.[11]
Earlier we identified sources of bias that can arise from problems with aggregation, data quality, implicit assumptions of strong separability among others. Almost all
demand studies have an implied assumption of strong separability in that they only consider aviation markets in the analysis. Such studies in effect constrain all
changes or responses in fares or service to be wholly contained in the aviation component of people's consumption bundle. The paper by Oum and Gillen (1986) is the
one exception where consideration of substitution with other parts of consumption was included in the modelling. It would be difficult to extract a conclusion from this
one study as to existence, degree and direction of bias in elasticity estimates when other parts of consumption are and are not included in the modelling. However,
having said this, an inspection of the elasticity estimates from this study shows they are not significantly different than other time-series estimates.
3.1 DATA ISSUES
Elasticity estimates depend critically on the quality and extent of the data available. Currently, the best data for demand estimation is the DB1A 10 percent ticket
sample in the US, but even this data has some problems.[12] The DB1A sample represents 10 percent of all tickets sold with full itinerary identified by the coupons
attached to the ticket. However with electronic tickets, as more and more tickets are being sold over the Internet, there is a growing portion of overall travel that may
not be captured in the sample. This means that the proportion is not 10 percent but something less.[13] Other important considerations are the amount of travel on
frequent flyer points, by crew and airline personnel.
In Canada we have poor quality data because it is incomplete, even if it were accessible. Airports collect traffic statistics but these data make it very difficult to
distinguish OD and segment data. Airlines report traffic data to Statistics Canada (or are supposed to) but these data do not include fare information or routing.
Knowing the itinerary or routing is important because of differences in service quality and hubbing effects. Fare data is also more useful than yield information since it
identifies the proportion of people travelling in different fare classes. Yet, in many cases yield information is used as a weighted average fare. There is also the problem
that carriers of different size may have different reporting requirements. Some researchers and consultants have been cobbling together data sets for analysis by using
the PBX clearing house information. These data are limited and apply only to those airlines that are members of IATA. [14] The current public data available in Canada
simply does not permit estimation of any demand models.
Besides demand side data it is also important to have supply side information. Elasticity estimates should emerge from a simultaneous equations framework. This data
is more accessible through organizations like the OAG [15], which provide information on capacity, airline and aircraft type for each flight in each market.[16] These data
measure changes in capacity, flight frequency and timing of flights.
One study, which undertook an extensive survey to collect multimodal data,[17] was the High Speed Rail study sponsored jointly by the Federal, Ontario and Quebec
governments. This study, which had three different demand modelling efforts, examined the potential for High Speed Rail demand, and subsequent investment, in the
Windsor-Quebec corridor. The analysis included intermodal substitution between air, rail, bus and car. The study was undertaken in the early 1980s. However, it is not
possible for public access to any of the technical documents that would allow an assessment of the study. Attempts in the past to obtain access to the data have proven
fruitless.
3.2 DISTINGUISHING ELASTICITY MEASURES
As we have stated, price elasticity measures the degree of responsiveness to a change in own or other prices (fares). However, care must be exercised in interpreting
the elasticity since they differ according to how they have been estimated. Many empirical studies of air travel demand estimate a log-linear model. In evaluating such
studies, it is important to keep in mind that the empirical specification implies a certain consumer preference structure because of the duality between utility functions
and demand functions. It is equally important to remember that empirically estimated demand functions should contain some measures of quality and service
differences or quality changes over time. Failure to include metrics for frequent flyer programs, flight frequency, destination choice or service levels in estimating an air
demand function can lead to downward bias in the price elasticity estimates.
Price elasticities can be estimated for aggregate travel demand as well as modal demand. Figure 3.1 illustrates the differences between aggregate and modal
elasticities.[18] Our interest is in modal elasticities not the aggregate amount of travel but it is important ultimately that any policy analysis take account of the impact of
any policy change on aggregate travel as well as modal redistribution. The impact of a change in price on aggregate demand would be measured by the fis in Figure
3.1 whereas the Fiis would measure the impact on air travel demand. The Fiis are a composite or combination of the fis and the Miis.
The Canadian aviation industry has undergone significant change in the last several years. In 2000 Air Canada completed its takeover of Canadian Airlines, which left it
with in excess of 80 percent market share. Market dominance leads to different fare and service quality levels. As a result of higher fares, for example, we should find
higher absolute values of elasticities of demand simply because with higher fares we have moved further up the demand curve. In 1996 Westjet entered the market
and has continued to grow each year. Canada 3000 exited the market in 2001, as did Canjet and Royal (as part of Canada 3000). Roots airline has come and gone but
Canjet has reemerged in eastern Canada and JetsGo is offering some level of service on longer haul domestic flights as well as in the Montreal-Toronto market.
The entry of low cost carriers leads to lower fares for a subset of traffic and competitors will offer a supply of seats to match these fares. Lower average fares should
lead to lower demand elasticity estimates, while increases in the number of competitors in the market will lead to higher demand elasticity estimates.
Figure 3.1
One should not confuse low cost carriers with a seeming lack of exploiting monopoly power. High prices or fares are not synonymous with monopoly and low fares with
competition. Airlines like Westjet where they are the sole airline serving the market may still act as a monopolist but charge low(er) fares. Profit maximizing
monopolists price where marginal cost equals marginal revenue, if marginal cost is low, one should expect to see lower fares but still marginal cost and revenue are
equalized. Monopolists are generally viewed as being high price because they are high cost and the high costs are attributable to some degree from a lack of
competitive discipline in the market. Full service carriers operating with hub-and-spoke systems have a high cost business model while low cost carriers have a low cost
business model.
4. EVALUATION OF ELASTICITY STUDIES
Overall we have collected some 254 demand elasticity estimates from 21 studies. Each of these studies is described, using a standardized summary sheet as illustrated
in Appendix A. To aid our understanding of how existing elasticity estimates might inform policy makers in forecasting air travel demand, we provide a descriptive
meta-analysis of various distributions of estimated values in section 4.1. We next develop a weighted scoring table with respect to generally desirable data, design and
output characteristics of the studies. This allows us to generate a rank ordering of the studies, from which to generate a sub-sample of estimates from studies with a
'passing grade' score. A passing grade is simply defined as 50 percent of the maximum score attainable. From these studies we provide suggested ranges of elasticity
values in six key market segments:
1. Short-haul business travel

2. Short-haul leisure travel
3. Long-haul, domestic business travel
4. Long-haul, domestic leisure travel
5. Long-haul, international business travel
6. Long-haul, international leisure travel
Before we discuss our scoring system for the studies, we first present some more general descriptive information on the distribution of estimated elasticity values in
various categories.
4.1 DESCRIPTIVE DISTRIBUTIONS OF ELASTICITY ESTIMATES
Here, we present for the aggregate and for several important sub-categories, histograms of the estimates in the studies we have researched. We begin with the most
general distribution: the set of all the studies containing some 254 estimates of own-price elasticity. We next present sub-categories in increasing detail defined in
terms of market characteristics. We also present sub-samples of the estimates based on data type (cross-section versus time-series) and the age of the study (less
than five years old, versus between five and ten years old). In each case we report the median value as a measure of central tendency, along with the kurtosis and
skewness of the distributions.[19]
4.1.1. All studies[20]
We generate a histogram for all own-price elasticities with 254 estimates taken from 21 studies.[21] The minimum estimated elasticity value is 3.20.[22] The histogram
demonstrates a crowd of estimates between zero and 2.5. The median, or midpoint, of all estimates is -1.122. We use the median as the measure of central tendency,
as opposed to the mean, in order to remove the effects of outliers in our data set. The skewness of the histogram is (-0.37). This indicates that our data is not normally
distributed.
All Studies
Own-Price Elasticities
5th percentile -1.967
First quartile -1.418
Median -1.122
Third quartile -0.633
Interquartile range 0.785
Number of estimates 254
Minimum -3.200
Maximum 0.040
Variance 0.312
Skewness -0.370
Kurtosis 0.177
The histogram below provides a more detailed depiction of the crowd of elasticity estimates with each column representing a one-tenth (0.1) segment.
4.1.2. All long-haul studies[23]
We subdivide the aggregate data into a subset of long-haul own-price elasticity estimates. The data set includes estimates for distances greater than 1500 miles, or
estimates that are reported as 'long-haul' or 'international' in their respective study. The subset is comprised of 100 estimates with a median elasticity of 0.857. A
majority of the values are bunched up between zero and 2 as indicated by the skewness of the histogram at 0.275.
All Long-haul Own-price Elasticity Estimates
Median -0.857
Number of estimates 100
Minimum -2.234
Maximum -0.010
Variance 0.298
Skewness -0.275
Kurtosis -0.874
4.1.3. All Short/medium-haul studies[24]
The data set for short/medium-haul own-price elasticity estimates includes estimates for distances less than 1500 miles, or estimates that are reported as 'short-haul',
'medium-haul', or 'regional' in their respective study. The subset is comprised of 109 estimates. Note that the sum of long-haul and short/medium-haul estimates
(100+109) does not equal the number of estimates in the aggregate data set. This is a result of the exclusion of elasticity estimates that are not defined by their
distance in their respective reports. The median elasticity in this subset is 1.15. A crowd of estimates is located between zero and 1.5. The minimum value (3.20)
represents a Sydney-Brisbane route taken from Milloy et al. (1985). The skewness of the histogram is 0.434.
All Short/Medium-haul Own-price Elasticity Estimates
Median -1.150
Number of estimates 109.000
Minimum -3.200
Maximum 0.040
Variance 0.329
Skewness -0.434
Kurtosis 0.710
The histogram below provides a more detailed depiction of the crowd of elasticity estimates with each column representing a one-tenth (0.1) segment.
4.1.4. All long-haul international travel estimates[25]
This sub-category of long-haul international travel is comprised of 69 estimates extracted from the aggregate data set. The data set represents estimates for country-
to-country international travel taken from seven studies. The estimates are distributed between zero and 2.7, with some crowding below 0.5. The median elasticity is
0.79 and the distribution is somewhat skewed.
Long-haul International Travel Own-price Elasticities
Median -0.790
Minimum -2.700
Maximum -0.010
Variance 0.407
Skewness -0.672
Kurtosis -0.456
4.1.5. All long-haul domestic studies[26]
This subset is comprised of 36 estimates extracted from six studies. The majority of the estimates are bunched between the maximum value (-0.44) and 2.3. The
skewness is 0.168.
All Long-haul Domestic Travel Own-price Elasticities
Median -1.150
Minimum -1.810
Maximum -0.440
Variance 0.149
Skewness 0.168
Kurtosis -1.078
4.1.6. All long-haul, international business travel estimates[27]
The international business travel subset contains 16 estimates from two studies. A majority of the estimates (15) are calculated by the Bureau of Transport
Communications and Economics (1995) for business travellers to and from Australia. The lowest estimate (-2.0) represents Australian business travellers to the U.K.
The majority of the estimates are bunched between the maximum value (-0.01) and 0.6. The median elasticity estimate is 0.265. The histogram is negatively
skewed (-2.405), which indicates a non-normal distribution.
All long-haul international business travel
Own-price Elasticities
Median -0.265
Minimum -2.000
Maximum -0.010
Variance 0.251
Skewness -2.405
Kurtosis 6.095
All long-haul international business travel

Median -0.265

Minimum -2.000
Maximum -0.010
Variance 0.251
Skewness -2.405
Kurtosis 6.095
4.1.7. All long-haul international leisure travel estimates[28]
The long-haul leisure travel segment contains a total of 55 estimates, representing seven studies. Nearly 50 percent of the estimates (24) are taken from the Bureau of
Transport Communications and Economics (1995) study. The median of the estimates is 0.993 with estimates distributed between -0.14 and 2.7. The minimum
values (-2.7) are taken from Taplin (1980) and represent elasticity estimates calculated by Jud and Joseph (1974) (for travel from the U.S. to Latin America), and from
Straszheim (1978) (for high discount travel). The skewness of the histogram is 0.555.
All Long-Haul International Leisure

Median -0.993

Minimum -2.700
Maximum -0.140
Variance 0.423
Skewness -0.555
Kurtosis -0.393
4.1.8 All Long-haul Domestic Business Estimates[29]
The long-haul domestic business travel subset is comprised of 26 estimates from two studies. The estimates are bunched up between the 0.5 and 1.6. The median of
the histogram is 1.15. The skewness (0.270) indicates a non-normal distribution.
All Long-haul Domestic Business Own-price Elasticities
Median -1.150

Minimum -1.700
Maximum -0.543
Variance 0.113
Skewness 0.207
Kurtosis -1.119
4.1.9 Long-haul Domestic Leisure Histogram[30]
The long-haul domestic leisure travel subset is comprised of seven estimates from two studies. The estimates are distributed between 0.44 and 3.20. The median
elasticity is 1.120.
All Long-Haul Domestic Leisure

Median -1.120

Minimum -3.200
Maximum -0.440
Variance 0.821
Skewness -1.640
Kurtosis 3.265
4.1.10. All short-haul business travel estimates[31]
The short-haul business travel subset is comprised of 18 estimates taken from four studies. The median elasticity is 0.73. The histogram demonstrates some crowding
of values between 0.5 and 0.8.
All Short-haul Business Travel Own-price Elasticities
Median -0.730

Minimum -1.500
Maximum -0.100
Variance 0.106
Skewness -0.151
Kurtosis 1.509
4.1.11. All short-haul leisure travel estimates[32]
This subset is comprised of 19 estimates from five studies. The median elasticity is 1.52 with estimates distributed across the range of values with little crowding. The
histogram is positively skewed (0.158), which indicates that the number of estimates decrease as we approach zero.
All Short-haul Leisure Travel Own-price Elasticities

Median -1.520
Minimum -2.370
Maximum -0.400
Variance 0.307
Skewness 0.158
Kurtosis -0.704
4.1.12 All Cross section study estimates[33]
The subset of all cross-sectional studies is comprised of 85 estimates, of which 80 estimates are taken from Oum et al. (1986) and represent U.S. city-pair routes. All
of the estimates are taken from studies between 1981 and 1986. The median elasticity is 1.33. The histogram is positively skewed (0.314).
All Cross-section Study

Median -1.330
Minimum -2.010
Maximum -0.181
Variance 0.158
Skewness 0.314
Kurtosis -0.563
4.1.13 All Time-series study estimates[34]
This subset is comprised of 136 estimates, twenty-eight of which are taken from studies published within the last five years. The histogram is negatively skewed with a
crowd of estimates between zero and 2. The median elasticity is 0.847.
All Time-series Study Estimates
Median -0.847
Minimum -2.540
Maximum 0.040
Variance 0.313
Skewness -0.542
Kurtosis -0.227
4.1.14 Studies 5-10 years old[35]
Two subsets have been created based on the age of the studies. The first subset is comprised of estimates calculated in studies published between 1992 and 1997. This
subset contains 45 estimates from three studies.
Estimates for all studies
1992-1997
Median -0.560
Minimum -2.300
Maximum -0.010
Variance 0.369
Skewness -0.907
Kurtosis -0.152
The median elasticity is 0.56 with the majority of estimates residing between 0.1 and 1.1. The histogram is negatively skewed with a skewness of 0.907, which
indicates a non-normal distribution. The second subset of estimates based on the age of the study is comprised of estimates calculated in studies published between
1997 and 2002.[36] Four studies qualify for this subset resulting in 30 estimates. The histogram demonstrates no crowding around a small range of values. Instead,
there is a wide distribution of values between zero and 2.3. The median elasticity is 0.847.
In comparing the median elasticity value for 1997-02 studies (-0.847) with the median elasticity for studies produced between 1992-97 (-0.56), it would appear that
own-price elasticity of demand has become more price sensitive (elastic) over time. However, interpreting this change is not straightforward. The completion date of a
study does not map directly to the age of the data employed. For example, the Nairn (1992) study utilizes data from 1983 and 1984, while the Hamal study (1998)
uses time series data from 1974-1996. Furthermore, the comparison becomes less informative when we examine the range between first and third quartiles for each
distribution. The range for 1997-02 studies is -0.5 to 1.4 while the range for 1992-97 studies is 0.3 to 1.2.
Own-price Elasticity
Estimates for all studies 1997-2002

Median -0.847

Minimum -2.234
Maximum 0.040
Variance 0.407
Skewness -0.426
Kurtosis -0.731
4.1.15 All income elasticities[37]
The subset of all income elasticities contains 132 estimates from 14 studies. The minimum estimated elasticity value is 1.21, which represents inbound pleasure travel
to Australia from the United States, as calculated by Hollander (1982). The maximum value is 11.58, which is calculated in the Bureau of Transport Communications
and Economics (1995) report for leisure travel by Australian residents to Taiwan. The median estimate is 1.39. There is a crowd of estimates bunched up between 0.5
and 2.5.
All studies
Income Elasticities
5th percentile 0.249
First quartile 0.840
Median 1.390
Third quartile 2.169

Minimum -1.210
Maximum 11.580
Variance 2.506
Skewness 2.671
4.1.16 Summary
Table 4.1.16 summarizes the median values of estimated own-price elasticities by market segment and study characteristics (data type and age).
Table 4.1.16
Summary of median elasticity values by type
Category Median Own-price Elasticity Value
All estimates -1.122
All long haul estimates -0.857
All long-haul international estimates -0.790
All long-haul international business estimates -0.265
All long-haul international leisure estimates -0.993
All long-haul domestic estimates -1.150
All long-haul domestic business estimates -1.150
All long-haul domestic leisure estimates -1.120
All short/medium haul estimates -1.150
All short/medium haul business estimates -0.730
All short/medium haul leisure estimates -1.520
All cross-section study estimates -1.330
All time-series study estimates -0.847
All estimates from studies 1992-1997 -0.560

All estimates from studies 1997-2002 -0.847
The table indicates that there are significant differences between some market segment elasticities (long-haul international business and short-haul leisure in
particular) and the median value for all estimates (-1.122). Time-series estimates indicate relatively less price sensitivity than those derived from cross-section studies.
Moreover, recent studies have returned relatively elastic values compared with older studies.
4.2 SCORING THE STUDIES
To improve the level of confidence regarding the practical use of elasticity values in forecasting air travel demand, we developed a scoring system based on desirable
input and output characteristics of empirical demand studies. Following on from our earlier discussion of theoretical and measurement issues, we have identified eleven
characteristics that contribute to the quality of elasticity estimates. In each case, the point scores represent our assessment of the relative importance of either the
inclusion or exclusion of the characteristic in question. We readily acknowledge that the subjective assignment of point scores cannot provide definitive scientific
results. Nevertheless we feel that in the absence of time or resources for more sophisticated analysis (meta regression analysis, risk analysis and bootstrapping
techniques for example), the scoring rule provides a useful rule of thumb for comparing the reliability of estimates we feel should be of higher quality with the overall
set of estimates from all the studies surveyed.
Specifically, we have rated the studies based on the following characteristics:
i. Separation of business and leisure travel

ii. Separation of long-haul vs. short-haul travel
iii. Inclusion of an income coefficient
iv. Inclusion of intermodal substitution
v. Data type: panel vs. time-series vs. cross-section
vi. Country focus
vii. Route-specific estimates
viii. Hub vs. non-hub airports
ix. Connecting vs. O-D passengers
x. Age of the study
xi. Adjusted R-squared values
i. Separation of business and leisure travel
We expect business travel to be more price insensitive than leisure travel. Consequently studies that do not distinguish between these market segments are likely to
provide elasticity estimates that would be biased if applied in any detailed analysis whether applied to specific business or leisure market segments, or to routes, which
are predominantly business or leisure oriented.
Scoring rule:
Estimates for both business and leisure = 3 points
Estimates for either business or leisure = 2 points
No separation of business and leisure = 0 points
ii. Separation of Long-haul vs. short-haul travel
We expect less price sensitivity for long-haul flights than for short-haul flights (where more inter-modal substitution is possible). In a similar fashion to the
business/leisure distinction, studies that do not distinguish market segments by flight length will provide elasticities that underestimate price sensitivity for short-haul
flights and over-estimate it for long-haul flights.
Scoring rule:
Estimates for both long and short-haul = 3 points
Estimates for either long or short-haul = 2 points
No separation of long and short-haul = 0 points
iii. Inclusion of an income coefficient
Without an income coefficient, demand studies will confuse a shift of the demand curve with movements along the demand curve. With a positive income elasticity for
air travel, and increasing per-capita real income, demand studies with no income coefficient will overestimate the absolute price elasticity of demand for price
decreases and underestimate it for price increases.
Scoring rule:
Income coefficient = 2 points
No income coefficient = 0 points

iv. Inclusion of intermodal substitution (for short-haul flights)
The shorter the distance comprising a trip, the more road and trail transportation become effective substitutes for air travel. Therefore we would expect the price and
other characteristics of alternative modes to have a more significant (shift) impact on the demand for short-haul air travel, ceteris paribus. Studies of short-haul flights
that do not include intermodal effects are likely to provide bias estimates if the shadow prices of alternative modes change. The scoring rule in this case attempts to
award short-haul studies that incorporate intermodal effects, without penalizing studies of longer-haul air travel.
Scoring rule:
Intermodal substitution in short-haul study = 2 points
Not a short-haul study = 1 point
No intermodal substitution in short-haul study = 0 points
v. Data: panel vs. time-series vs. cross-section
Policy analysis should not be guided by immediate or short-term reactions to prices that result from policy changes. Consequently, policies that impact air travel
demand should rely more on long-term elasticity measures. While panel studies are ideal as they capture cross-section and time-series effects, studies from time-series
data that are sufficiently long in duration will also capture longer-term elasticities.
Scoring rule:
Use of panel data or time-series = 2 points
Use of crosssection data = 0 points
vi. Country focus
There are likely to be many structural details of price sensitivity that relate to the specific national context of the airline industry, including the degree of competition,
the size of the market and the regulatory environment. The impact of policies on air travel prices in Canada can be more readily related to some countries more than
others. The close geographical proximity of international hubs and agreements within the EU make European studies somewhat less relevant to the Canadian context.
US studies are more relevant given the geographic proximity of the US to Canada and the number of US cities to which Canadians travel. Australia on the other hand,
provides reasonably comparable demographic, urban, geographical, governmental and regulatory structures.
Scoring rule:
Study relates directly to Canada = 2 points
Study relates to similar foreign country (US or Australia) = 1 point
Study relates to non-similar foreign country = 0 points
vii. Route-specific estimates
Studies that aggregate the effects of price changes on multiple routes will not capture the effects of market competition in which certain airlines enjoy significant
market power on some routes but not others. A well-known example of this in the US is the effects of low-cost competition by Southwest Airlines on routes flown by
full-service carriers. A related issue is that studies, which focus on multiple short-haul routes run the danger of aggregating effects of routes that are predominantly
used by business travellers with routes that are more leisure-oriented. This latter category often constitutes a significant portion of business for low-cost carriers, who
offer cheap short-haul flights in competition with alternative leisure activities and entertainment. An example of this is the market for special event parties in Dublin
(wedding stags for example) that was created by flights offered by RyanAir from various locations in the UK.
Scoring rule:
Study provides route or airline-specific estimates = 1 point
Study does not provide route or airline-specific estimates = 0 points
viii. Hub vs. non-hub airports
Studies that do not separate out hub from connecting airports will not be able to distinguish "hub premium" effects. Passengers with an itinerary that utilizes a hub
airport may be willing to pay a "hub premium" for the integrated service that hubs provide, including sequenced flight segments that minimize waiting time, and
baggage that is checked through to the final destination. The existence of a hub premium effect is supported by research in the US.
Scoring rule:
Study identifies hub airports = 1 point
Study does not identify hub airports = 0 points

ix. Connecting vs. O-D passengers
There is a difference between an itinerary and the measurement of traffic volumes between city pairs. If a passenger is travelling from Moncton to Vancouver via
Toronto, then their willingness-to-pay and their price sensitivity relates to the trip from Moncton to Vancouver. However, such a passenger could be included in the
data that is measuring price sensitivity on the city pair Toronto-Vancouver.
Scoring rule:
Study identifies connecting vs. O-D passengers = 1 point
Study does not distinguish connecting vs. O-D passengers = 0 points
x. Age of the study
The airline industry is a dynamic and changing industry, in the evolution of business models (full-service versus low cost carriers for example), infrastructure (airport
business practices) and government regulation. Studies conducted in the US prior to 1978 would not incorporate the effects of deregulation. A similar argument applies
to studies that predate 1984 in Canada. Further the National Airport Policy in Canada has led to a gradual devolution of airports from Transport Canada to independent
local airport authorities throughout the 1990's. This devolution has led to important infrastructure and pricing decisions. Only the most recent studies would capture
system-wide effects of this evolution as some local airport authorities have only come into being in the last year or two.
Scoring rule:
Studies completed during 1997-2002 = 3 points
Studies completed during 1990-1997 = 2 points
Studies completed prior to 1990 = 1 point
xi. Adjusted R-squared coefficient values
This last item addresses the quality of output in the studies rather than the quality of inputs. In regression results, a low R-squared value indicates that only a small
portion of variation in the dependent variable (O-D passengers), is explained by the independent variables. The adjusted R-squared value is a weighted measure that
penalizes the addition of a large number of independent variables with low explanatory power.
Scoring rule:
Adjusted R-squared value over 0.8 = 3 points
Adjusted R-squared value between 0.6 and 0.8 = 1 point
Adjusted R-squared value lower than 0.6 = 0 points
The highest possible score under the criteria we have developed is 23 points. Table 4.2.1 below summarizes the scores of each study, from which we have generated
histograms in six sub-categories using only those studies with a 'passing grade' of 12 points or higher. The categories provide separation of long and short-haul,
international and domestic travel and business and leisure travel. Note that the column headings in the table refer to the numbered characteristics discussed above.
Table 4.2.1 Summary of Study Scores (cont'd)

4.2.1 Short-haul business travel[38]
This subset is comprised of 16 estimates taken from three studies, the most recent of which is Battersby-Oczkowski (2001). The median elasticity for the data set is
0.70.
Short-haul Business Travel Own-price Elasticities:
Studies Scoring 12 Points
Median -0.700
Minimum -1.110
Maximum -0.100
Variance 0.072
Skewness 0.697
Kurtosis 1.396
4.2.2 Short-haul leisure travel[39]
Three studies scoring 12 or more points in our scoring system generate 16 estimates of short-haul leisure travel. The estimates are distributed between a range of 0.4
and 2.37. The minimum value is taken from the Bureau of Transport Economics (1986) and represents winter vacation travel in Australia. The median estimate for all
values is 1.520.
Short-haul Leisure Travel Own-price Elasticities:
Studies Scoring 12 Points
Median -1.520
Minimum -2.370
Maximum -0.400
Variance 0.278
Skewness 0.485
Kurtosis -0.116
4.2.3 Long-haul international business travel[40]
The subset of international business travel estimates provides 16 estimates taken from two studies. The median elasticity is 0.265, which is the same value derived
prior to applying the scoring model to the aggregate data set. This occurred because both data sets are comprised of the same estimates.
Long-haul International Business Travel

Own-price elasticities
Studies scoring 12 points
Median -0.265
Minimum -2.000
Maximum -0.010
Variance 0.251
Skewness -2.405
Kurtosis 6.095
4.2.4 Long-haul international leisure travel[41]
There are 49 international leisure travel price elasticity estimates from six studies with at least 12 points in our scoring system. A majority of the estimates (31) are
taken from studies published after 1995. The median elasticity is 1.040 with a large proportion of the estimates bunched up between 0.14 and 1.
Long-haul International Leisure Travel

Own-price elasticities
Median -1.040
Minimum -2.700
Maximum -0.140
Variance 0.420
Skewness -0.465
Kurtosis -0.474
4.2.5 Long-haul domestic business travel[42]
The domestic long-haul business travel subset consists of 26 estimates from two studies. The median elasticity is 1.15. The estimates are taken from Lubulwa (1986)
and Oum et al. (1986).
Long-Haul Domestic Business Travel

Own-price Elasticities Studies scoring 12 points
Median -1.150
Minimum -1.700
Maximum -0.543
Variance 0.113
Skewness 0.207
Kurtosis -1.119
4.2.6 Long-haul domestic leisure travel[43]
There are six long-haul domestic leisure travel price elasticity estimates taken from two studies. The median elasticity is 1.104. The histogram demonstrates no
crowding around a range of values.
Long-haul Domestic Leisure Travel Own-price Elasticities

Median -1.104
Minimum -1.680
Maximum -0.440
Variance 0.191
Skewness 0.015
Kurtosis -0.171
4.2.7 Income elasticities[44]
A subset of 103 income elasticity estimates is generated from the 'passing grade' studies. In similar fashion to the histogram for all studies, a crowding of estimates
around the values of 0.5 to 2.5. The median value of the subset is 1.14.
Income Elasticities
for all Studies Scoring 12 points
First quartile 0.807
Median 1.140
Third quartile 2.089
Minimum -1.039
Maximum 11.580
Variance 2.642
Skewness 3.051
Kurtosis 14.139
4.2.8 Studies that account for Intermodal effects[45]
The data set for studies that include the effects of intermodal competition (e.g. auto, rail, bus, ship) is comprised of 104 own-price elasticity estimates taken from
thirteen studies (including short, medium and long haul routes). The histogram does not demonstrate any bunching around a set of values as supported by the
skewness (-0.641). The median elasticity value is 1.113.
Own-price Elasticities: Studies with Intermodal Effects
Median -1.113
Minimum -3.200
Maximum 0.040
Variance 0.389
Skewness -0.641
Kurtosis 0.614
A subset of estimates is extracted that includes estimates for short/medium-haul elasticities.[46] The data set is comprised of 34 estimates taken from four studies
including those elasticities calculated by Battersby-Ozckowski (2001). The median estimate
(0.720) is lower than the median elasticity calculated for all intermodal studies (-1.113). However, some details of these studies make the interpretation of this result
difficult.
First, the subset includes discount, economy, and business fare-class estimates. The elasticities reflect both the nature of travel on the routes (business or leisure) and
the fare class. For example, the Sydney-Melbourne route is a significant business route in Australia with relatively low elasticity estimates (Discount= 0.07, Economy=
-0.81, Business = -0.1). The dataset contains several estimates from routes that are historically business travel city-pairs.
Secondly, thirteen of the estimates are from city-pair routes with distances of approximately 870 to 1000 km (the high end of the short-haul distance condition). Only
five out of 34 estimates are explicitly defined as short-haul routes of less than 750 km. These elasticity estimates are: Melbourne-Adelaide (-0.46); Australia short-haul
<500 km (-0.728); New South Wales, Australia, routes <200km (-2.54); Short-haul Western and Mid-western, U.S., routes <500 miles (-0.08); Short-haul Eastern
city-pairs, U.S. <500 miles (-0.36). Two of the estimates (Short-haul Eastern U.S., and Western and Mid-Western U.S.) are likely capturing business travel.
Lastly, 28 of the 34 estimates are taken from studies comprised of Australian city-pair routes. These studies do not provide sufficient information about the city-pair
characteristics, such as whether or not a specific route has one or more (or possibly no) competing transportation modes. If the use of a competing mode is infeasible
or highly unlikely then the elasticity estimate is not capturing intermodal effects.
Short/Medium-haul Own-price Elasticities:

Studies with Intermodal Effects
Median -0.720
Minimum -3.200
Maximum 0.040
Variance 0.508
Skewness -1.524
Kurtosis 2.925
5. DISCUSSION
In a number of cases, studies that are focused on the impact of price changes or fees on demand use a single elasticity measure to compute the quantity, revenue and
profit change for a route, market, airline or entire economy (see for example, PODM (Transport Canada) which uses one elasticity of business and one for leisure;
economic impact studies for airports often use this approach as well). Using a single value implicitly assumes that the elasticity measure is transferable across markets
and time. There is a rich and extensive literature that explains the conditions under which such estimates are transferable.[47] The properties or characteristics of the
data in different markets should meet statistical tests in order to have statistical validity in having a common elasticity.
We have shown that elasticity values can and do differ significantly between travel distance, type of traveller and even domestic and international routes. This is
illustrated in Table 5.1 in which we report elasticities for three different route types and two passenger types. We have argued that the usefulness of estimates should
be based upon other criteria such as the inclusion of income coefficients and distinctions between types of passengers and airports. We have also argued that route-
specific data is especially valuable in capturing competitive, geographic and market differences.
In the report we show that for the entire set of studies as well as for categories of studies the distribution of elasticity estimates is highly skewed. Such a distribution
makes the use of the mean or average tenuous at best. The mean may turn out to be a value that was yielded by none of the studies. The variance is also large which
makes the level of confidence we can place in a 'mean' value particularly low. Therefore, we have used the 'median' value of the elasticity estimates as an indicator of
what elasticity value might be used in forecasting changes in revenue, passengers and profit in markets where the elasticity is appropriate short or long-haul,
business or leisure etc. as a result of a policy change.[48] In addition to the reported median values in the various categories, we have also reported quartile information
from the distributions of elasticity values; dividing the observations into quartiles simply means we have divided it into quarters, so the 1st quartile would be the first 25
percent of observations. In particular, table 5.1 draws attention to the first and third quartiles (twenty-five percent of the values in a distribution fall below the first
quartile and seventy-five percent fall below the third quartile).[49] The first and third quartiles form a useful range around the median that widens (narrows) as the tails
of the distribution grow thicker (thinner). This is illustrated in figures 5.1a and 5.1b.
Figure 5.1a: Inter-quartile range with wide tails in the distribution
Figure 5.1b: Inter-quartile range with narrow tails in the distribution

Table 5.1 below, shows median absolute values (meaning we have dropped the negative sign in front of the elasticity value) of estimated demand elasticities along with
the first and third quartiles of the distribution for all studies and for 'passing grade' studies in six categories. [50]
In the long-haul international market, there is no apparent difference between the elasticity values from all studies and the group regarded as having a 'passing grade',
based on our scoring system.[51] The median values are low (0.265) for business travel and close to unity for leisure travel. This seems reasonable, since long-haul
international business travel demand has relatively few close substitutes, making demand insensitive to fare changes. On the other hand, international leisure travellers
are more likely to postpone trips to specific locations in response to higher fares, or shop around for those
Table 5.1
Summary Table of Absolute Elasticity Values
Category Median Median
(1st quartile) (3rd quartile) (1st quartile) (3rd quartile)
Elasticity Values Elasticity Values

All Studies Studies Scoring 12 points
0.265 0.265
Own-price: Long-haul international business 0.475 0.198 0.475 0.198
0.993 1.040
Own-price: Long-haul international leisure 1.65 0.535 1.700 0.560
1.150 1.150
Own-price: Long-haul domestic business 1.428 0.836 1.428 0.836
1.120 1.104
Own-price: Long-haul domestic leisure 1.472 0.887 1.228 0.787
1.520 1.520
Own-price: Short/medium-haul leisure 1.745 0.885 1.743 1.288
0.730 0.700
Own-price: Short/medium-haul business 0.798 0.608 0.783 0.595
Income Elasticity 1.390 1.140
0.840 2.169 0.807 2.0489
locations offering more affordable fares. In the vacation market, international travel competes more directly with domestic travel for vacation destinations.
The long-haul domestic business segment elasticities are the same whether looking at all studies or the sub-set of 'passing grade' studies. The value of 1.15, being
close to unity indicates that domestic business travellers will have higher elasticities (in this case about four times the value) than international business travellers. In
domestic markets, alternatives such as telecommunications are more substitutable than in international markets due to common culture, laws, contracts etc.
International trips are typically planned well in advance, with the travel spread over more time we would expect the airfare to be a lower proportion of overall trip
costs.
The median long-haul domestic leisure elasticity values do not differ significantly between all studies and those rated superior, however the range of elasticity values in
the passing grade studies (as defined by the first and third quartiles of the distribution) is narrower and slightly lower. The value of 1.104 does not seem unreasonable
in comparison to the domestic business travel elasticities.
The median elasticity value from all studies versus 'passing grade' studies for short/medium-haul leisure are identical and characteristically elastic at 1.52. Notice
however that the range of values around the median is narrower in the passing grade studies, excluding the possibility of inelastic demand at the lower bound of the
range.
The estimated median fare elasticity for short/medium-haul business travel is moderately inelastic at 0.73 with a very tight range around the median. Generally, the
price or fare elasticity will decline with distance other things held constant. However, we can see in Table 5.1 that the short/medium haul business elasticity is smaller
in value than the long haul domestic business elasticity. The explanation is quite straightforward, people using aviation as a business tool, will have an especially high
value of time. They will therefore willingly pay high fares to save time on short haul trips. Secondly, there are numerous 'must-go' short haul trips that arise in the
course of business dealings, trips that are not easily planned and can be completed in a morning or afternoon without requiring scheduling meetings or packing or
making family arrangements. These short haul business trips will have low elasticities; for example, in flying to Montreal from Toronto to close a multi-million dollar
deal and sign the contact, the high air fare is low when factored into the overall value of the trip.
Table 5.1 provides ample evidence that using a single elasticity for all market segments is inappropriate just as a single elasticity will not reflect impacts on the
aggregate market. Furthermore, simply segmenting markets by business and leisure is insufficient to provide any degree of accuracy to forecast changes in passengers
with changes in fares. For example, given the differences between short and long-haul market elasticities, using long-haul values to evaluate impacts on short-haul
markets would generally provide an underestimate.
It may be the case that elasticity values reported in the empirical literature are underestimated for both leisure and business travel, particularly in short to medium-
haul markets. The reasoning is based on research, which shows that the entry of low cost carriers into markets leads to a reduction in the average fares on those
routes (Windle and Dresner, 1999).[52] For example, if a low cost carrier like Southwest enters a market, the effect has been a reduction in fares of almost 50 percent.
Table 5.2 compares very recent revenue yields in various US markets that are delineated based on length of flight but also on whether there is competition from
Southwest Airlines, and the form of that competition. The table shows, quite dramatically, that the more direct the competition from Southwest, the lower the yields
(which translates into lower average fares).
Table 5.2
Revenue yields of other airlines (OA) and Southwest (SW)
Yields
(cents per passenger mile)
500 Miles 1000 Miles
Market type
OA-no SW presence 51 26
OA-SW connecting competition 31 20
OA- SW direct competition 26 19
SW-connect 21 14
SW-non-stop 18 12
Sources: US Department of Transportation (2002); D. Gillen "Frills, no frills or Wal-Mart: The future of Canada's Aviation
industry" (Forthcoming, Van Horne Institute, University of Calgary)
There are few studies that have included, as a time-series, the growth in markets where low cost carriers have concentrated their activity. In empirical studies, routes
are usually aggregated so an average elasticity is estimated across short, medium and long-haul routes. Thus, in studies using detailed US data markets served by
Southwest Airlines and more recently by other low cost carriers such as JetBlue and Air Tran are aggregated with those served by other full service carriers even
though the growth in traffic in these markets is quite different.
In order to look at the potential underestimation of demand response in markets where low cost carriers participate we used US data from 1999-2000 2nd Quarter. The
year-over -year changes in passengers and fares are used to calculate arc-elasticities for routes of different length and for fare increases and fare decreases. We found
the calculated arc elasticities did not differ in any significant way from the values we have found from our survey of the literature. This applies to values for long-haul
as well as short to medium-haul markets; short/medium-haul markets are more price sensitive than long-haul markets. We therefore feel the values reported in Table
5.1 fairly reflect the sensitivity of markets including those served by low cost and low fare carriers.[53]
As we note in the discussion of Table 5.1, there is no single elasticity value that is representative of air travel demand. There are several distinct markets and several
different elasticities should be used when exploring the impact on markets from changes to the aviation environment. Furthermore, even given the elasticity for a
market segment, there is a range around this elasticity that should be considered in using the elasticity to forecast the impact of fare changes. The aggregate
elasticities for the market segment reflect the combined effect of demand relationships in each component market. Each market will typically exhibit different
elasticities than that considered for the aggregate market level.
BIBLIOGRAPHY
Abrahams, M., "A Service Quality Model of Air Travel Demand: An Empirical Study", Transportation Research, 17A(5), 385-93, 1983.
Agarwal, V. and W. K. Talley, "The Demand for International Air Passenger Service Provided by U.S. Air Carriers", Interntaional Journal of Transport Economics, 12 (1),
63-70, 1985.
Anderson, James E., and Marvin Kraus, "Quality of Service and the Demand for Air Travel", The Review of Economics and Statistics, Volume 63, Issue 4, 533-540,
1981.
Andrikopoulos A.A., and T. Terovitis, "An Abstract Mode Model: A Cross-section and Time-series Investigation", International Journal of Transport Economics, 10(3),
563-76, 1983
Battersby, B. and E. Oczkowski, "An Econometric Analysis of the Demand for Domestic Air Travel in Australia", International Journal of Transport Economics, Vol.
28(2),193-204, 2001.
Bhadra, Dipasis, "Demand for Air Travel in the United States: Bottom-Up Econometric Estimation and Implications for Forecasts by O&D Pairs", Center for Advanced
Aviation System Development The Mitre Corporation, 2002.
Brons Martijn, Eric Pels, Peter Nijkamp and Piet Rietvela, "Price Elasticities of Demand for Passenger Air Travel: A Meta-Analysis", Tinbergen Institute, Amsterdam,
2001.
Bureau of Industry Economics, "Expenditures by overseas visitors and Australians travelling abroad",Tourist Expenditure in Australia, BIE Research Report 16.
Canberra, AGPS, 1984.
Bureau of Transport and Communications Economics, International Aviation, Report 86, 19-22.
Bureau of Transport Communications and Economics, "Demand Elasticities for Air Travel to and from Australia", Working Paper 20, Department of Transport and
Communications, 1995.
Bureau of Transport Economics, "Demand for Australian Domestic Aviation Services Forecasts by Market Segment", AGOS, Canberra. Occasional Paper 79., 1986.
Fridstrom, L. and H. Thune-Larsen, "An Econometric Air Travel Demand Model for the Entire Conventional Domestic Network: The Case of Norway", Transportation
Research, 23B(3), 213-24, 1989.
Gillen D. W. and Morrison W. G., "Airport Financing, Costing, Pricing and Performance", Report to the Canadian Transportation Act review Committee, April 2001.
Hamal, K., "Australian Outbound Holiday Travel Demand: Long-haul Versus Short-haul", Bureau of Tourism Research, Canberra, BTR Conference Paper 98.2, 1998.
Hollander, G., "Determinants of Demand for Travel to and from Australia", BIE Working Paper no. 26,Bureau of Industry Economics, Canberra., 1982.
Ippolito, R.A., "Estimating Airline Demand With Quality of Service Variables", Journal of Transport Economics and Policy, 15(1), 457-64, 1981.
Jung, J.M. and E.T. Fujii, "The Price Elasticity of Demand for Air Travel", Journal of Transport Economics and Policy, Volume 10, 257-262, 1976.
Lubulwa, A.S.G., "Brandow Demand Functions For Australian Long Distance Travel", Forum Papers, 11th Australian Transport Research Forum, volume (2), 200-218,
1986.
May, T.E., E.W.A. Butcher and G. Mills, "Consumer Responsiveness to Changes in Air Fares, Independent Review of Economic Regulation of Domestic Aviation", vol. 2,
Appendix L, 1986.
Morrison, Steven A., and Clifford Winston, "An Econometric Analysis of the Demand for Intercity Passenger Transportation", Research in Transportation Economics,
Volume 2, 213-237, 1985.
Nairn, R.J. and P. Hooper, "Tourism Related Movement Study Final Report", Roads and Traffic Authority, NSW, Sydney., 1992.
Oum T.H. and D. W. Gillen , "The Structure of Intercity Travel Demands in Canada: Theory Tests and Empirical Results", Transportation Research, 17B(3), 175-91,
1982.
Oum T.H., D. W. Gillen and S.E. Noble, "Demand for Fareclasses and Pricing in Airline Markets, Logistics and Transportation Review", 22(3), 195-222, 1986.
Oum, T.H., Waters, W.G. and Yong , J.S., "A Survey of Recent Estimates of Price Elasticities of Demand for Transport", World Bank Working Paper, WPS 359, 1990.
Oum T.H., Waters, W.G. and Yong , J.S., "Concepts of Price Elasticities of Transport Demand and Recent Empirical Estimates", Journal of Transport Economics and
Policy, vol.26(2), 139-154, 1992.
Oum T.H., A. Zhang and Y. Zhang, "Inter-Firm Rivalry and Firm-Specific Price Elasticities in Deregulated Airline Markets", Journal of Transport Economics and Policy,
vol. 27(2), 171-92, 1993.
Talley, W.K. and A. Schwarz-Miller, "The Demand for Air Services Provided by Air Passenger-Cargo Carriers in a Deregulated Environment", International Journal of
Transport Economics, 15(2), 159-68, 1988.
Taplin, J.H.E., "A Coherence Approach to Estimates of Price Elasticities in the Vacation Travel Market",Journal of Transport Economics and Policy, 14(1), 19-35, 1980.
Taplin, J.H.E., "A Generalised Decomposition of Travel-Related Demand Elasticities into Choice and Generation Components", Journal of Transport Economics and
Policy, Volume 31 (2), 183-191, 1997.
APPENDIX A: SURVEY OF DEMAND ELASTICITY STUDIES
We have adopted a standard reporting sheet to summarize existing studies in the literature. Each sheet denotes (when possible), the publication or completion date of
each study; the country or countries studied; the modes of travel studied and the type and sources of data used. In addition, we note which market segments are
identified (business/leisure; long-haul/short-haul) and the types of elasticities estimated (own-price, cross-price or income). Finally, the summary sheet specifies where
possible, the type and functional form of the model employed along with any relevant statistical properties.
Many studies in the literature were completed and/or published prior to 1990 more than twelve years ago. Consequently, actual estimate values in these studies
provide less relevance for forecasting air travel demand in Canada. For this reason, we have grouped the summary sheets into older studies (completed prior to 1990),
and more recent studies.
A.1 OLDER STUDIES (PRIOR TO 1990)
A.1.1 Title of study
A Coherence Approach to Estimates of Price Elasticities in the Vacation Travel Market

Authors John H. E. Taplin
Date completed/published 1980
General Summary information
Country/countries studied Various
Modes of travel studied Air (Passenger)
Data Sources A summary of results of 8 previous studies.
Specific focus
Identified Elasticities Yes No Comment
Business travel demand No
Leisure travel demand Yes
Long-haul between major centres Yes Only country-pairs are provided.
Long-haul between small communities No
Short-haul between major centres No
Short-haul between small communities No
Own-price elasticity values Yes
Cross-price elasticity values Yes*
Income elasticity values Yes
Technical report
Type of model employed Unknown
Modes of travel included in study
Functional form of the model Unknown
Policy or price change relevant to estimate(s) in the study Unknown
Statistical properties of estimates

Other Comments
* Taplin calculates cross-elasticity values, which include estimates with respect to car operating costs, price of domestic accommodation, price of overseas
accommodation, and the price of other consumer goods and services.
Estimating Airline Demand With Quality Of Service Variables
Authors Richard A. Ippolito
Country/countries studied U.S.
Modes of travel studied Single mode: Air (Passenger)
Data Sources Data sources unknown.
Specific focus
Leisure travel demand Use dummy variables for segments that served Florida, California, or Las
Yes
Vegas
Long-haul between major centres No
Short-haul between small

communities No
Cross-price elasticity values No*
Income elasticity values No
Technical report
Type of model employed Log-linear demand function, allow for inter-modal competition by dummy;
Simultaneous equation model
Modes of travel included in study
Functional form of the model
Policy or price change relevant to estimate(s) Pre-deregulation in the U.S.

in the study
Statistical properties of estimates Only t-stat values
Other Comments
* Ippolito: To account for mode choice, distance (in logs) is added to the demand specification. Moreover, since the preference for the auto mode may be particularly
strong for very short trips, zero-one dummy variables were included in the model, which equalled unity when trip distance was 100 miles or less.
A.1.3 Title of study The Structure of Inter-city Travel Demands in Canada: Theory Tests and Empirical Results
Authors Tae H. Oum and David W. Gillen
Date completed/published July 1982
Country/countries studied Canada
Modes of travel studied Three travel modes (Air, Bus, Rail)
Data Sources Statistics Canada (CANSIM data base); 1961-76.
Specific focus
Leisure travel demand No
Cross-price elasticity values Yes (Includes: bus, rail, goods, and services)
Technical report
Type of model employed Structural analysis demand equations derived from utility maximization.
Estimated by Non-linear least squares.
Modes of travel included in study Air, Bus, Rail

Policy or price change relevant to
estimate(s) in the study
Statistical properties of estimates Air: D.W. Statistics = 1.645; R-square = 0.9048
Determinants of Demand for Travel to and From Australia

Authors G. Hollander
Country/countries studied Australia, New Zealand, UK, US, Japan, Canada, Germany, Italy
Data Sources Bureau of Industry Economics
Specific focus
Leisure travel demand Yes Country of origin to Australia.
Country-pairs are provided.

Long-haul between major centres Yes
Cross-price elasticity values No Unknown if model includes intermodal effects.
Technical report
Type of model employed Double-log; pooled time-series, cross-section 1975-1981
Modes of travel included in study Air (Passenger)
Statistical properties of estimates Unknown

Other Comments
All data was retrieved from the Bureau of Transport and Regional Economics website at:http://dynamic.dotars.gov.au/btre/tedb/tablist_detail.cfm?ID=153
A Service Quality Model of Air Travel Demand: An Empirical Study

Authors Michael Abrahams
Date completed/published April, 1983
Country/countries
studied United States
Modes of travel
studied Air passenger travel (demand equation includes automobile operating costs)
Data Sources C.A.B. Service Segment Data Base (100 most heavily traveled domestic origin-destination pairs in the
U.S.); Measure of price employed is the lowest unrestricted coach fare deflated by the Consumer Price
Index (1973-77); Value of schedule delay time are estimates obtained from a procedure developed by
Ericson (1977) and Swan (1978)
Specific focus
Leisure travel demand Yes Northeastern U.S. major centres to Florida; West Coast to Hawaii. [54]
Long-haul between major centres Yes Transcontinental U.S. (> 1500 miles)
Short-haul between major centres Yes Eastern U.S. (< 500 miles)
Cross-price elasticity values No Effect of fare on demand through changes in the air: auto model split
Technical report
Type of model employed 2 SLS estimation with Cochrane-Orcutt Transformation (used to correct for first order serial
correlation); Time-series (20 quarters)
Models of travel included in

study
Functional form of the model Demand and Elasticity equation: See 'Other Comments'.
Policy or price change relevant The study was conducted pre-1978 Airline Deregulation Act in the U.S.
to estimate(s) in the study
Statistical properties of Durbin-Watson values indicate that the hypothesis that there exists no first order serial
estimates correlation cannot be rejected; R-square for all pools were > 0.96 for dummy variable, and >
0.91 for non-dummy variable;
Other Comments
City-pairs were pooled on the basis of common characteristics.
Demand equation:
Where,
= Lowest unrestricted coach (air) fare between city i and j;
= Expected schedule delay time in hours as estimated using equation;
= Auto costs as described in equation;
= Population of Standard Metropolitan Statistical Area (SMSA) containing city i times the population of SMSA containing j;
= Income per capita of SMSA containing i times income per capita of SMSA containing j;
GGNP = Gross National Product.
Elasticity equation:
Where,
= Average price elasticity of demand;
= Estimated fare coefficient for P;
= Estimated coefficient for AC;
= Average real fare;
= Average passenger traffic;
ADC = average auto driving costs.
An Abstract Mode Model: A Cross-section and Time-series Investigation

Authors Andreas A. Andrikopoulos and Theophilos Terovitis

Country/countries studied Greece
Modes of travel studied Air-ship, Air-Bus, Air-bus-rail
Data Sources Civil Aviation Service, Olympic Airline Time and Fare Tables
Specific focus
Cross-price elasticity values Yes
Technical report
Type of model employed Linear demand; Cross-section (1970-80) and Time-series data (1969-80);
Estimated by OLS
Models of travel included in study
Policy or price change relevant to estimate(s) in

the study Unknown
Other Comments
Number of air passengers per unit of time is taken as the dependent variable. Two sets of explanatory variables: 1) the mode's attributes relative to the close
substitute, 2) Socio-economic variables including population and tourism.
Expenditure by Overseas Visitors and Australians Travelling Abroad

Authors Bureau of Industry Economics (Australia)
Country/countries studied Australia, New Zealand, UK, US, Japan, Canada, Germany Italy
Data Sources Bureau of Industry Economics (Australia)
Specific focus
Long-haul between major centres Yes Between countries of origin and Australia.
Cross-price elasticity values Yes Intermodal, prices in alternative destinations.
Technical report
Type of model employed Unknown; Pooled time-series and cross-section; 1970-1980
Other Comments
All data was retrieved from the Bureau of Transport and Regional Economics website at:http://dynamic.dotars.gov.au/btre/tedb/list_detail.cfm?Ref_ID=21
The Demand for International Air Passenger Service Provided by U.S. Air Carriers
Authors Vinod Agarwal and Wayne K. Talley
Country/countries studied U.S. and undisclosed foreign-country destinations. (city-pairs)
Modes of travel studied Single mode: Air (Passenger)
Data Sources International Civil Aviation Organization (ICAO), Official Airline Guide (World-wide Edition)
Specific focus
Leisure travel demand Yes All elasticities are assumed to be 'excursion' travel.
Cross-price elasticity values No
Technical report
Type of model employed Log linear demand estimated by OLS. Cross-section data (Dec. 1981), 63
flight segments.
Functional form of the model See 'Other Comments'.

the study Post-deregulation data after 1978 in the U.S.
Statistical properties of estimates Only variables P, FF, LF are significant at the .05 level; Adjusted R-square =
.8766
Other Comments
Demand equation:
Where,
= Number of passengers transported by U.S. air carriers from the ith U.S. departure point to the jth foreign-country landing point.
= Average excursion fare for U.S. air carriers on a given flight segment divided by the distance in kilometers of that flight segment.
= Represents the proportion of all air passengers transported by U.S. air carriers from the ith U.S. departure point to the jth foreign-country landing point.
=Average of the travel times for all U.S. air carriers providing service on a given flight segment.
= Total number of flights provided by all U.S. air carriers serving ij flight segment.
= Number of passengers transported by U.S. air carriers (for December 1981) divided by the number of aircraft seats flown by these carriers over a given
flight segment.
Title of study
A.1.9
An Econometric Analysis of the Demand for Intercity Passenger Transportation

Authors Steven A. Morrison and Clifford Winston
Country/countries studied U.S.
Modes of travel studied Rail, Air, Bus (Vacation and Business trips)
Data Sources 1977 Census of Transportation National Travel Survey; June 1977 Official Airline Guide
Specific focus
Business travel demand Yes
Cross-price elasticity values Yes Intermodal coefficients are included for rail, car, and bus.
Technical report
Type of model employed Nested logit model.
Modes of travel included in study Rail, Bus, Air.
Policy or price change relevant to estimate(s) in the study Data taken from 1977 (Pre-deregulation in the U.S.)
Other Comments
Morrison and Winston note that the elasticity results for vacation travellers (Air) are not too large (less than 1.0). They assert that given the large share of the market
that this mode possesses, the results are not too surprising.
Morrison and Winston: More generally, the air destination elasticities indicate that mode choice price elasticities for air may understate the total traveller
responsiveness to changes in airfare. Consequently, such elasticities should be viewed with some caution.
Title of study
A.1.10
Demands for Fareclasses and Pricing in Airline Markets

Authors Tae H. Oum, David W. Gillen and S.E. Noble
Country/countries
studied U.S.
Modes of travel
studied Domestic air passenger travel
Data Sources Official Airline Guides, North American Edition; Domestic Origin-Destination Survey of Airline Passenger
Traffic, Air Transport Association of America; Two hundred intra-U.S. routes were selected from 1978
cross-sectional data.
Specific focus
Identified elasticities Yes No Comment
Business travel demand Yes Aggregate data includes business and non-business travel.
Leisure travel demand Yes Washington D.C. Miami, Pittsburgh Miami, N.Y. - Miami
Short-haul between major centres Yes

Cross-price elasticity values Yes[55] Between Fare classes: first class, economy, and discount.
Income elasticity values Yes Vacation routes
Technical report
Type of model employed First stage: Partial Elasticities, translog demand system.
Second stage: Total Elasticities, Log-linear aggregate demand model; cross-sectional data
(1978), intra-U.S. routes

study Air (Passenger)
Functional form of the model Demand equation: see other comments.
Statistical properties of Total Price Elasticities, R-square = 0.6793

estimates Vacation routes (price-elasticity) t-ratio = 2.08
Vacation routes (income-elasticity) t-ratio = 1.43
Non-Vacation routes (price elasticity) t-ratio = 2.52
Other Comments
Demand equation:
Where,
= Aggregate traffic volume of route r;
= Weighted average fare of route r using the fitted revenue shares Sj's as the weights;
= Dummy variable (equal to 1 for vacation rous);
= Weighted average per-capita income of the two cities on route r;
= Total population of the two cities on route r.
Demand for Australian Domestic Aviation Services by Market Segment

Authors Bureau of Transport Economics
Country/countries studied Australia
Data Sources Bureau of Transport Economics

Specific focus
Leisure travel demand Yes Tasmania and Queensland.
Long-haul between major centres Yes O-D points not given.
Short-haul between major centres Yes O-D points not given.
Cross-price elasticity values Yes Transport mode not given.
Technical report
Type of model employed Double-log; time-series data; 1977-1983
Modes of travel included in study Air (Passenger); Alternative mode for cross-elasticity not given.
Other Comments
Title of study
A.1.12
Brandow demand functions for Australian long distance travel

Authors A.S.G. Lubulwa
Data Sources Compilation of 7 studies.

Specific focus
Long-haul between major centres Yes[56] O-D points not specified for all elasticities.
Short-haul between major centres Yes[57] O-D points not specified for all elasticities.
Cross-price elasticity values Yes Cost of car.
Technical report
Type of model employed Unknown; Compilation of seven studies
Modes of travel included in study Air; (Cost of travel by car is provided as cross-elasticity)
Other Comments
All data retrieved from the Bureau of Transport & Regional Economics website at:http://dynamic.dotars.gov.au/btre/tedb/tablist_detail.cfm?ID=42
Consumer responsiveness to changes in air fares

Authors T.E. May, E.W.A. Butcher, and G. Mills
Data Sources Unknown

Specific focus
Long-haul between major centres Yes O-D points not given.
Long-haul between small communities No Unknown.
Short-haul between major centres Yes O-D points not given.
Short-haul between small communities No Unknown.
Technical report
Type of model employed Unknown; data set from 1977-1984
Modes of travel included in study Air
Other Comments
All data retrieved from the Bureau of Transport & Regional Economics web site at:http://dynamic.dotars.gov.au/btre/tedb/tablist_detail.cfm?ID=43

The Demand for Air Services Provided By Air Passenger-Cargo Carriers In A Deregulated
Environment
Authors Wayne K. Talley and Ann Schwarz-Miller
Date
completed/published 1981
Country/countries
studied U.S.
Modes of travel studied Single Mode: Air (Passenger and Cargo)
Data Sources
22 U.S. air-passenger-cargo carriers for the year 1983: Moody's Transportation Manual and the Air
Carrier Traffic Statistics
Specific focus
Technical report
Type of model employed Log-linear two stage least squares demand function; Cross-section
data
Policy or price change relevant to estimate(s) in the

study Post-deregulation in the U.S.
Statistical properties of estimates R-square: 0.9662, Adjusted R-square: 0.9696; t-stat: -3.453
Other Comments
General form of the demand function:
Where,
= Number of passenger miles of service demanded of the ith carrier for the tth time period.
= Passenger fare per mile of the ith carrier for the tth time period.
= Number of departures by the ith carrier for the tth time period.
= Overall flight stage length (average distance covered per flight hop) by the ith carrier for the tth period.

An Econometric Air Travel Demand Model For the Entire Conventional Domestic Network: The
Case of Norway
Authors Lasse Fridstrom and Harald Thune-Larsen
Date
completed/published July 1988
Country/countries studied Norway
Data Sources Origin-Destination data set (source unknown).
Specific focus
Long-haul between major No Travel time including checking, checkout, access and egress times between city center
centres and airport (in minutes) is a variable in the equation.
Long-haul between small ""

communities No
Short-haul between major ""

centres No
Short-haul between small ""

communities No
Cross-price elasticity values Yes Cross demand elasticity of air travel with respect to surface fares.
Technical report
Type of model employed Gravity Model: the model was estimated applying ordinary least squares regression to a
logarithmic transformation of the equation. Norway Time-series (1972-83 annual) and Cross-
section (95 intercity links) data.

stud
Policy or price change relevant

to estimate(s) in the study
Statistical properties of
estimates Sample size: 1140 annual observations. R-square: 0.6923 < R < 0.7520
Other Comments
The model includes a short-term/medium-term and a long-term variable.
A.2 MORE RECENT STUDIES (1991 - )
Tourism Related Movement Study Final Report
Authors Nairn, R.J. and Hooper, P.
Country/countries studied USA, Australia
Data Sources Pickrell (1984), BTE (1983)
Specific focus
Business travel demand Yes U.S.
Leisure travel demand Yes Australia
Short-haul between major centres Yes U.S. (Centres unknown)

Cross-price elasticity values No Unknown if intermodal coefficients are used.
Technical report
Type of model employed Unknown
Other Comments
Title of study
A.2.2
Demand Elasticities for Air Travel to and from Australia

Authors Bureau of Transport Communications and Economics
Country/countries
studied Australia and numerous destination countries around the world.
Data Sources ABS, Overseas Arrivals and Departures Australia, Cat. 3402.0; Book1 Worldwide Fares, Air Tariff
Publications
Specific focus

Cross-price elasticity values No[58]
Technical report
Type of model employed Double log and linear models are employed; Time-series data
Policy or price change relevant to estimate(s) in the study Unknown.
Other Comments
Leisure travel
A general expression for the dynamic model of leisure travel demand is:
for i = 1, 2, ... , 24 and t = 1, 2, ... , 32
where i is the leisure travel market analysed for foreign visitors and Australian travellers between Australia and each of the 12 countries, t is the quarterly time period
and j is the number of quarterly lags.
The variable LD is leisure demand for air travel, LF is the real leisure airfare, REX is the real exchange rate, and Y is real income.
The coefficients and represent the shortrun effect on demand of a change in the airfare, the real exchange rate and income. The sum of the
coefficients and represents the longrun effect on demand of a change in the airfare, the real exchange rate and income. These
coefficients can be used to derive long-run elasticities of demand.
Business travel
A general expression for the dynamic model of business travel demand is:
for i = 1, 2, ... , 24 and t = 1, 2, ... , 32

where i is the business market analysed for foreign visitors and Australian travellers between Australia and each of the 12 countries, t is the quarterly time period
and j is the number of quarterly lags.
The variable BD is business demand for air travel, BF is the real business airfare, REX is the real exchange rate, AGDP is real Australian gross domestic product
and FGDP is real foreign gross domestic product.

A Generalised Decomposition of Travel-Related Demand Elasticities into Choice and Generation
Components
Authors John H. E. Taplin
Date
completed/published May 1997
Country/countries
studied Australia and unknown overseas O-D pairs.
Modes of travel
studied Air (Passenger)
Data Sources Uses elasticities sourced from other studies cited in Taplin (1980) to calculate one own-price and cross
price elasticity for domestic travel within Australia, and one for overseas travel to/and from Australia.
Specific focus
Leisure travel demand Yes All elasticities are considered 'Vacation Air Trips'.
Long-haul between major O-D pair unknown. Only 'Overseas' si cited in the paper.
centres Unknown
Long-haul between small ""

communities Unknown
Short-haul between O-D pair for domestic travel unknown.

major centres Unknown
Short-haul between small

communities Unknown
Own-price elasticity
values Yes
Cross-price elasticity "Car Operating Costs" is a variable used to infer cross-price. Other cross-price
values Yes elasticities were also calculated for various goods and services (such as hotel price).
Technical report
Type of model employed Unknown. The paper is not explicit about the model employed.
Modes of travel included in study Air.

Functional form of the model Unknown.
Policy or price change relevant to estimate(s) in the Taplin uses elasticities calculated by other papers published prior to
study 1978.
Statistical properties of estimates Not provided in the paper.
Other Comments
Australian Outbound Holiday Travel Demand: Long-haul Versus Short-haul
Authors Hamal, K.
Country/countries studied Australia and Foreign Destinations
Data Sources Bureau of Tourism Research, Canberra
Specific focus
Leisure travel demand Yes Australian outbound holiday travel.
Short-haul between major centres No[60]
Cross-price elasticity values Yes
Technical report
Type of model employed Double-log linear, time-series data (1974-1996)
Modes of travel included in study Air

Functional form of the model Unknown.
Policy or price change relevant to estimate(s) in the study N/A
Title of study
A.2.5
An Econometric Analysis of the Demand for Domestic Air Travel in Australia

Authors B. Battersby and E. Oczkowski
Modes of travel studied Air (Passenger): Three distinct segments discount, full economy, and business.
Data Sources Quarterly data (1992 1998); Bureau of Transport Economics;
Specific focus
Short-haul between major centres Yes[61]
Cross-price elasticity values Yes No specific transport mode is highlighted, just an index.
Technical report
Type of model employed Linear

Functional form of the model Not provided in detail. Demand determinants included: price, income, substitute
prices and seasonality.
Policy or price change relevant to estimate(s)

in the study N/A
Other Comments
The authors note that their own-price elasticity estimates are generally at the lower end of the estimates reported by previous studies. The estimates reported for the
Melbourne-Brisbane route are particularly low. This general divergence from previous estimates may in part be due to the explicit modelling of individual market
segments, which contrasts to the aggregate route analysis conducted by most other studies.

Demand for Air Travel in the United States: Bottom-Up Econometric Estimation and
Implications for Forecasts by O&D Pairs
Authors Dipasis Bhadra
Date
completed/published 2nd Draft (2002)
Country/countries
studied United States
Modes of travel
studied Air (Passenger)
Data Sources Aviation statistics from Bureau of Transportation Statistics (10% Survey) (50,000 records for eight
quarters, 1999 and 2000); and local area data from Bureau of Economic Analysis
Specific focus
Long-haul between major centres Yes Distance (in miles) is used to identify long or short haul.
Long-haul between small communities Yes ""
Short-haul between major centres Yes ""
Short-haul between small communities Yes ""
Own-price elasticity values Yes ""

Technical report
Type of model employed Semi-logarithmic linear model with Limited Information Maximum-Likelihood
(LIML) estimation.

the study N/A
Other Comments
Where,
P = average daily passengers
D and ND = dominant and non-dominant airlines
f = one-way fare
PI = personal income
Density = population density per sq. miles
Interactions = intensity of economic activities are represented by interactions between population and income
Distance = distance traveled between O&D markets
Market Power = share of passenger demand by airlines in total O&D market
Southwest = presence of Southwest Airlines in O&D market
Season = adverse spring and summer weather
Take log for those independent variables for which interpretations are meaningful. Leave out hub status, Southwest presence and season as dummy variables.
FOOTNOTES
1. The distribution of elasticity estimates is highly skewed in each distinct market. Since the estimates are not distributed normally around their average value, the
median is used as a measure of central tendency. (The median is the middle observation: half of the estimates are bigger and half are smaller.) The variance of the
estimates was also found to be large, which further reduces the level of confidence in using the average for this purpose. [Return]
2.The elasticity estimates for each distinct market are divided into four equal parts, or quartiles. The reported range of values comprises those observations lying
within the second and third quartiles. [Return]
3. In some cases separate equations are estimated for these markets; PODM (The Transport Canada air travel forecasting model) for example uses different equations
and variables for leisure and business markets. [Return]
4. Previous surveys (e.g. Oum et al., 1992) provide a listing of the elasticities and their ranges but no basis for choosing from among the values within the
range. [Return]
5. Theoretically an alternative to the ordinary demand function is the compensated demand function, obtained from a resource expense minimization subject to a given
level of utility. Elasticity values from the compensated demand function incorporate only substitution effects, however in practice we can estimate only the ordinary
demand function. Nevertheless the distinction is important since large price changes may yield significant income effects. [Return]
6. The exception would obviously be the constant elasticity demand function. [Return]
7. A panel is a data set that contains both time-series and cross-sectional information. [Return]
8. This will be true for all factors other than own-price. [Return]
9. The exception here is durable goods, for the opposite relationship is expected between long and short run elasticities. [Return]
10. The difference between this point and the previous one is that hub airports will have different service levels and will generally have a hub premium. [Return]
11. Fortunately, several techniques have been developed for the estimation of the structural parameters of an a priori specified system of simultaneous stochastic
equations. These include indirect least squares, two stage least squares, instrumental variables, three stage least squares, full information maximum likelihood, and
limited information maximum likelihood. [Return]
12. The term 'best' means researchers observe this data source to be the most geographically comprehensive, detailed and temporally available. [Return]
13. The growth of the Internet in booking tickets is being integrated into the DB1A database, as is the growing use of electronic tickets. [Return]
14. IATA is the International Air Transport Association. [Return]
15. OAG is the Official Airline Guide. [Return]
16. These data are sold and can be expensive. [Return]
17. Estimates were that in excess of $1 Million was spent on data collection alone. [Return]
18. This figure is adapted from Figure 1 in Oum et al. (1992) [Return]
19. In computing numerically descriptive measures of data we are generally interested in two indicators: the measure of average or central value of the data and the
measure of the degree to which the data are spread out about this average value. The most popular measure of central tendency is the arithmetic mean (known
generally as simply the mean) but this can be unduly influenced by extreme observations; which is the case here. We have therefore used the median as the measure
of central location. The median is the value that splits the data set exactly in half. The measure of spread is called the variance; how widely the data are dispersed
around the measure of central location. Two added descriptors of the distribution of the data are skewness and kurtosis. The first, skewness, describes whether the
data are distributed symmetrically around the measure of central location. If they are not, they are said to be 'skewed' or have a short tail on one side and a longer tail
on the other rather than having equally sized tails in each side. Kurtosis is a measure of whether the data are peaked or flat relative to a normal distribution. That is,
data sets with high kurtosis tend to have a distinct peak near the measure of central location, decline rather rapidly, and have heavy tails. Data sets with low kurtosis
tend to have a flat top near the mean rather than a sharp peak. Since in many cases, the distribution of estimated values is skewed, the mean is not a useful measure
of central tendency and the standard deviation is not useful in providing a range around the mean. [Return]
20. The data-set is comprised of all twenty-one studies located in the appendix. [Return]
21. The study by Anderson and Kraus (1981) is not included in this histogram since they do not calculate elasticities directly. Doubt over the quality of estimated
positive elasticities in Jung and Fuji (1976) lead us to exclude their estimates also. [Return]
22. It is conventional to present own-price elasticities with a negative sign indicating the general negative relationship between price and quantity demanded. Larger
values of the elasticity imply greater price sensitivity while lower values imply less price sensitivity. [Return]
23. Source: Bhadra (2002), Hamal (1998), Taplin (1997), BTCE (1995), Oum et al. (1986), BTE (1986), Lubulwa (1986), May and Butcher (1986), BIE (1984),
Abrahams (1983), Hollander (1982), Taplin (1980). [Return]
24. Source: Battersby and Oczkowski (2001), Bhadra (2002), Nairn (1992), Oum et al. (1986), BTE (1986), Lubulwa (1986), May and Butcher (1986), Abrahams
(1983). [Return]
25. Source: Hamal (1998), Taplin (1997), BTCE (1995), Lubulwa (1986), BIE (1984), Hollander (1982), Taplin (1980). [Return]
26. Source: Oum et al. (1986), May and Butcher (1986), Lubulwa (1986), BTE (1986), Bhadra (2002), Abrahams (1983). [Return]
27. Source: BTCE (1995), Lubulwa (1986). [Return]
28. Source: Hamal (1998), Taplin (1997), BTCE (1995), Lubulwa (1986), BIE (1984), Hollander (1982), Taplin (1980). [Return]
29. Source: Oum et al. (1986), Lubulwa (1986). [Return]
30. Source: Lubulwa (1986), Abrahams (1983). [Return]
31. Source: Battersby and Oczkowski (2001), Nairn (1992), Lubulwa (1986), Oum et al. (1986). [Return]
32. Source: Nairn (1992), Oum et al. (1986), BTE (1986), May and Butcher (1986), Abrahams (1983). [Return]
33. Source: Oum et al. (1986), Talley (1988), Agarwal and Talley (1985), Morrison and Winston (1985), Ippolito (1981). [Return]
34. Source: Bhadra (2002), Battersby and Oczkowski (2001), Hamal (1998), BTCE (1995), Fridstrom (1989), BTE (1986), May and Butcher (1986), BIE (1984),
Abrahams (1983), Andrikopoulos (1983), Oum and Gillen (1982), Hollander (1982), Taplin (1980). [Return]
35. Source: BTCE (1995), Nairn (1992). [Return]
36. Bhadra (2002), Battersby and Oczkowski (2001), Hamal (1998), Taplin (1997). [Return]
37. Source: Battersby and Oczkowski (2001), Hamal (1998), Taplin (1997), BTCE (1995), Fridstrom (1989), Oum et al. (1986), BTE (1986), Lubulwa (1986), BIE
(1984), Abrahams (1983), Oum and Gillen (1982), Hollander (1982), Ippolito (1981), Taplin (1980). [Return]
38. Source: Battersby and Oczkowski (2001), Oum et al (1986), Lubulwa (1986). [Return]
39. Source: Oum et al. (1986), BTE (1986), Abrahams (1983). [Return]
40. Source: BTCE (1995), Lubulwa (1986). [Return]
41. Source: Hamal (1998), Taplin (1997), BTCE (1995), Lubulwa (1986), BIE (1984), Taplin (1980). [Return]
42. Source: Oum et al. (1986), Lubulwa (1986). [Return]
43. Source: Lubulwa (1986), Abrahams (1983). [Return]
44. Source: Battersby and Oczkowski (2001), Hamal (1998), Taplin (1997), BTCE (1995), Oum et al. (1986), BTE (1986), Lubulwa (1986), BIE (1984), Abrahamas
(1983), Taplin (1980). [Return]
45. Source: Battersby and Oczkowski (2001), Hamal (1998), Taplin (1997), Fridstrom (1989), BTE (1986), Lubulwa (1986), Morrison and Winston (1985), BIE (1984),
Abrahams (1983), Andrikopoulos (1983), Oum and Gillen (1982), Ippolito (1981), Taplin (1980). [Return]
46. Source: Battersby and Oczkowoski (2001), BTE (1986), Lubulwa (1986), Abrahams (1983). [Return]
47. This literature grew out of the early demand modeling efforts. See for example, Watson, Peter L. and Richard Westin. Transferability of Disaggregate Mode Choice
Models, Amsterdam: North-Holland, 1975 and Frank S. Koppelman and Eric I. Pas, Multidimensional Choice Model Transferability,.Transportation Research. Part B,
Methodological. Vol. 20B, no. 4 (Aug. 1986) p. 321-330 [Return]
48. We remind the reader the median is the value that divides the sample in half so 50 percent of observations will lie above the median value and 50 percent will lie
below it. [Return]
49. When using mean values as a measure of central tendency, the standard deviation of the distribution can be used to create confidence intervals of plus and minus
one standard deviation around the mean. Since we are using the median as a measure of central tendency, we cannot use standard deviations (which assume a normal
distribution which by definition is not skewed). [Return]
50. Absolute values mean we have dropped the negative sign before the elasticity value. We had included the sign in section 4 because price elasticities were negative
while income elasticities were positive. In a few cases studies reported positive price elasticities, a clear error. Here we are interested in only the degree of difference in
price sensitivity as reported by the magnitude of the elasticity value. [Return]
51. These are not the same set of studies but the superior group is a subset of the total. [Return]
52. Robert Windle, Martin Dresner (1999) Competitive Responses to Low Cost Carrier Entry,Transportation Research. Part E, Logistics and Transportation Review. Vol.
35E, no. 1 (Mar. 1999) p. 59-75 [Return]
53. We have not carried out any analysis on Canadian data since is there is no comparable data set to that available in the US. [Return]
54. We infer that the leisure routes from the Northeastern U.S. to Florida are short/medium haul routes. [Return]
55. No intermodal coefficients are used in the model. [Return]
56. We calculate distances for city-pair routes provided and distribute the estimates accordingly based on short/medium or long-haul. [Return]
57. Ibid. [Return]
58. BTCE: For Australia, there is effectively no intermodal competition to international air travel. In 1993, only 0.3% of arrivals and departures were by ship. P.28-
29 [Return]
60. Hamal differentiates between short-haul and long-haul international travel. Australia-U.K. is determined to be long-haul, Australia-Singapore is determined to be
short-haul. We calculate distances between all country-pairs and determined that these country-pairs all satisfied the long-haul condition. We group all estimates into
long-haul elasticities, as alternative modes of transportation are improbable. [Return]
61. We calculate city-pair distances for the Australian domestic O-D pairs used. Based on these distances all four (4) of the routes studies are short/medium-haul
domestic routes. [Return]

Air Travel Demand Elasticities

Uploaded by

Copyright:

Available Formats

Air Travel Demand Elasticities

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Air Travel Demand Elasticities

Uploaded by

Copyright:

Available Formats

What is an own-price elasticity of demand?

What is an own-price elasticity of demand?

What are the six distinct markets for air travel identified in the report?

What are the six distinct markets for air travel identified in the report?

AIR TRAVEL DEMAND ELASTICITIES:

CONCEPTS, ISSUES AND MEASUREMENT

Christopher Stewart MBA

2. Elasticity in the context of air travel demand.

4. Evaluation of elasticity studies.

Appendix A: survey of demand elasticity studies.

WHAT IS A PRICE ELASTICITY OF DEMAND?

KEY FINDINGS OF THE REPORT

Chart 1 - Own-Price Elasticities of Demand

2. ELASTICITY IN THE CONTEXT OF AIR TRAVEL DEMAND.

represent the observed change in quantity demanded and price

Q(P,S) is the demand function

The number of available substitutes;

The degree of competition in the market or industry;

The homogeneity of the product;

The duration of the time period analyzed.[9]

Connecting and origin-destination (O-D) travel;

Hub and non-hub airports;[10]

3.1 DATA ISSUES

3.2 DISTINGUISHING ELASTICITY MEASURES

4. EVALUATION OF ELASTICITY STUDIES

1. Short-haul business travel

4.1 DESCRIPTIVE DISTRIBUTIONS OF ELASTICITY ESTIMATES

4.1.1. All studies[20]

5th percentile -1.967

First quartile -1.418

Third quartile -0.633

95th percentile -0.190

Interquartile range 0.785

Number of estimates 254

4.1.2. All long-haul studies[23]

5th percentile -1.851

First quartile -1.365

Third quartile -0.495

95th percentile -0.190

Interquartile range 0.870

Number of estimates 100

4.1.3. All Short/medium-haul studies[24]

5th percentile -1.992

First quartile -1.520

Third quartile -0.712

95th percentile -0.112

Interquartile range 0.808

Number of estimates 109.000

5th percentile -1.960

First quartile -1.400

Third quartile -0.349

95th percentile -0.172

Interquartile range 1.051

Number of estimates 69.000

4.1.5. All long-haul domestic studies[26]

5th percentile -1.685

First quartile -1.528

Third quartile -0.828

95th percentile -0.553