= - percent change in quantity demanded / percent

change in price, or -(dQ/Q)/(dP/P). The minus sign is often omitted because price elasticity of demand is presumed to be negative. If Ed = 0, it is perfectly inelastic, a change in price does not affect the quantity demanded. If 0 >Ed

>-1, it is relatively inelastic, the quantity demanded does not increase at the same rate the price falls. If Ed = -1, there is unitary elasticity, both price and demand change equally. If -1>Ed, it is elastic, demand increases more than the fall in price. It is presumed that the changes in price are small. If the price elasticity of demand is not greater (more negative) than -1, a drop in price can actually reduce overall revenue received by the seller.

Airline industry observers have generally assumed that the demand for airline travel is price elastic. Indeed, one of the primary benefits expected with airline deregulation was a fall in the fare level and increased passenger traffic were regulatory price and service restrictions removed. Economists also generally view the effect of price changes in inflation-adjusted terms, e.g. "real" prices. This paper examines the effect of changes in price on the demand for U.S. domestic air travel, in both nominal and inflation-adjusted terms. The paper shows, contrary to general economic belief, that the overall price elasticity of demand for air travel has been inelastic (e.g. more positive than -1.0) since the early 1970's in both real and nominal terms. Domestic industry price elasticity estimates are developed using multiple linear regression with demand measured in revenue passenger-miles as the dependent variable and several selected independent variables (price, the economy, consumer confidence, and several "dummy" variables). The paper is divided into four subsections: The Change in Air Travel Demand, The Regression Model Form, Model Variables, and Results and Conclusions. Part A. The Change in Air Travel Demand

Graphed below is the year-to-year change in total domestic revenue passenger-miles (RPM's) from 1951 to 2007 (An RPM is one passenger travelling one mile). As shown, the domestic industry RPM's from year-to-year have shown a general decline in growth rate from 1951-1991, with relative stability since then. For purposes of analysis, this paper divides the general decline in growth of RPM's into three "steps": 1951-1968, 1969-1987, and 1988-2007. Two separate groupings of 1951-1991 and 1992-2007 are also provided, but use only one equation in real monetary terms and one equation in nominal terms for each period. 2 Figure 1. Annual Change in Total Domestic Revenue Passenger-Miles ANNUAL CHANGE IN TOTAL DOMESTIC REVENUE PASSENGER-MILES

0.90

0.95

1.00

1.05

1.10

1.15

1.20

1.25

1.30

1.35

1951

1953

1955

1957

1959

1961

1963

1965

1967

1969

1971

1973

1975

1977

1979

1981

1983

1985

1987

1989

1991

1993

1995

1997

1999

2001

2003

2005

2007

Year-Year Change (Ratio)

Source: Ratio of actual data from U.S. Department of Transportation, Bureau of Transportation Statistics, and U.S. Civil Aeronautics Board, Handbook of Airline Statistics, 1972.

Part B. The Regression Model Form

In general, industry price elasticity estimates are derived from economic models using data series, as opposed to point or arc elasticities using far fewer data points. Selection of the type of model form used in part depends on the structure of the data and the statistical results of initial model tests. Econometric models are generally one of two forms, additive or...