THE FOLLOWING IS AN EARLY, LONGER VERSION OF AN ARTICLE PUBLISHED BY DBEDT. IN THIS FORM IT DOES NOT NECESSARILY REPRESENT THE VIEWS OF DBEDT NOR OF THE STATE OF HAWAII.
Christopher Grandy
July 1999
The Issue
One of the issues facing Hawaii’s tourism industry is the ability to bring visitors to the islands. The development of the jet-powered airplane was one of the key technological developments that helped establish tourism as the leading economic engine for Hawaii. Maintaining adequate air service has been seen as critical to sustaining the state’s economic vitality.
The number of seats devoted to Hawaii has been an issue over the years. Some have argued that the number of air seats to Hawaii is inadequate. They see the limited number of flights to Hawaii as a constraint that prevents visitor counts from rising. If only airlines would increase capacity, the argument goes, visitor arrivals would rise and with it the economic activity that sustains Hawaii’s standard of living.1
Another perspective argues that air seat capacity is a function of profitability. If flight revenues generated by visitors to Hawaii are too low, or the costs of flying to Hawaii are too high, then airlines will reduce the number of seats devoted to the market. Rather than an artificial constraint, this perspective argues that the number of air seats is a function of supply and demand factors.
This article presents evidence bearing on the issue. The evidence is primarily confined to U.S. domestic carriers on westbound routes because of insufficient data on eastbound routes. Overall, it appears that the number of air seats to Hawaii reflects underlying economic factors.2
Some Evidence
Airfares
One of the pieces of evidence against the air seat-constraint hypothesis is that airfares to Hawaii have been falling after adjusting for inflation. If airlines were restricting the number of air seats to Hawaii, one might expect upward pressure on airfares. This does not appear to be happening. At least since the early 1980s, soon after the U.S. airline industry was deregulated, inflation-adjusted airfares to Hawaii have fallen. Figure 1 reports average airfares on domestic westbound flights to Honolulu, weighted by the share of seats from various Mainland cities. In nominal terms, airfares have been relatively constant at about $210 one way (based on roundtrip fares) since 1987. Adjusting for inflation, the real price of a one-way ticket has fallen from about $200 to $130 in 1982-84 dollars.3
Average real airfares have also been falling to other destinations in the U.S. This probably reflects the effect of deregulation and increased competition since the late 1970s. However, there is some evidence that even adjusting for distance and capacity, airfares to Hawaii are significantly lower than the U.S. average. For example, Figure 2 shows that revenue per available seat mile (RASM) to Hawaii is dramatically lower than the U.S. average.4
RASM will generally be lower for "leisure" destinations because passengers can plan in advance and take advantage of discount programs. But the figure also shows that compared to Orlando and Las Vegas, Hawaii RASM is low. This conclusion is clouded because the average distance to Orlando and Las Vegas from most domestic origins is lower than the distance to Hawaii, suggesting that Hawaii’s RASM may be relatively understated. Evidence from distance-equivalent city pairs is mixed however. For example, RASM for the New York-Las Vegas city pair (roughly the same distance as from the West Coast to Hawaii) approximately equals Honolulu’s RASM. On the other hand, the RASM for the Los Angeles-Orlando city pair (another equivalent distance) lies above the overall Las Vegas RASM illustrated in Figure 2.
All this suggests that, other things equal (such as cost), it would not pay a carrier to add capacity to Hawaii if it can earn $0.092 per seat per mile on an average alternative route versus $0.059 per seat per mile on a Hawaii route.
Load Factors
Load factors provide another (but mixed) piece of evidence on the air seat capacity issue. If air seats were constrained, one would expect load factors to rise to high levels. In fact, load factors have been rising, and some have interpreted this as evidence in favor of the constraint hypothesis.
The load factor is the ratio of passengers to seats on a flight. The average load factor to Hawaii has increased more or less steadily since the late 1980s (Figure 3). In the early 1980s, load factors hovered just above 70 percent. Since 1994 they have remained just below 85 percent.
However it remains unclear what this evidence means. While rising or high load factors are consistent with the constraint hypothesis, they may also be consistent with the "economic" hypothesis. Given the evidence on declining real fares to Hawaii, one way that airlines can improve profitability is to increase capacity utilization. Low load factors imply few customers across which to distribute the flight costs; high load factors imply more customers. It may be that high load factors simply reflect airlines’ attempts to boost efficiency in the face of low fares.
Interestingly, load factors in Hawaii are significantly higher than load factors elsewhere (see Figure 4). On average, between 1982 and 1997, load factors were 14 percentage points higher on domestic flights to Honolulu than on average domestic flights. Hawaii’s load factors are even high relative to those of Las Vegas and Orlando.
Air Seats
While the number of air seats to Hawaii is lower today than in the late 1980s, the number of seats has been rising since 1994 (Figure 5). However, all of the increase has occurred on the Neighbor Islands—principally Maui. Average load factors to Maui peaked in late 1993 and early 1994 at about 90 percent. Since then, they have drifted downward somewhat to about 85 percent. Interestingly, the number of seats flown to Maui increased at the same time that the load factor peaked, from about 20,000 per month in early 1994 to roughly 80,000 per month today. This suggests that once load factors reach a high level, near capacity, increased air seats follow.
This phenomenon appears in the statewide figures as well: Load factors reached 85 percent for the state in early 1994, just about the same time that the total number of air seats began to rise. Given declining real air fares, apparently load factors must remain high before the number of air seats to Hawaii begin to rise; once those load factors are achieved, the number of air seats do, in fact, rise.
Conclusion
Air seat capacity concerns are often expressed by those interested in "leisure" destinations. For example, officials in Las Vegas and New Zealand have also expressed concern over adequate capacity. At least with respect to Hawaii, the evidence reviewed for this article suggests that the provision of air seats is largely an economic decision. Unfortunately direct data on costs are not available, so that one cannot compare profitability across routes. Nevertheless, relatively low revenue per available seat mile reduces the incentive to increase capacity. Airlines have compensated by increasing load factors, thereby spreading costs over a greater number of travelers. However, it appears that load factors must rise to fairly high levels before the number of seats begins to increase. Fortunately, as the data on Hawaii air seats indicates since 1994, airlines appear to increase the number of seats once load factors rise sufficiently.
Notes 2Comparisons of profitability by route would be the most direct measure of evaluating whether economic determinants explain air seat capacity to Hawaii. Unfortunately data on costs by route are not available.
3The airfare data reported here include only paying passengers. Passengers traveling on frequent flier programs and other "free" status are excluded.
4RASM is yield (revenue per passenger per mile) multiplied by the load factor (passengers divided by seats); it is a standard industry measure of revenue.
1Officials of other "leisure" destinations, such as Las Vegas, Orlando, and New Zealand have also expressed concerns over air seat capacity.