Tarmac delay rules to cost the public welfare USD4 billion


Based on a new study released yesterday, consumers could rightfully say that the Department of Transportation didn't do them any favours by imposing the three-hour tarmac rule that became effective in Apr-2010. For the first time, a four-month-long study from Airline Zone Co-Authors Darryl Jenkins and Joshua Marks, both former executives with the George Washington University Airline Institute, has quantified the impact of the new rule, finding far more passenger disruption from increased cancellations occurred in the first month of the new rule, than ever happened with protracted delays prior to the rule.

  • A new study suggests that the three-hour tarmac rule imposed by the Department of Transportation (DOT) in 2010 has caused more passenger disruption and increased cancellations than previous protracted delays.
  • The study estimates that the rule has resulted in a net public cost of $4 billion over 20 years, compared to the DOT's projected public welfare gains of $69.1 million over the same period.
  • The authors of the study propose a four-hour cap on tarmac delays, which they believe would reduce cancellations and have greater public net benefit.
  • The study highlights the outsized fines imposed by the DOT, which can be 200 to 300 times the revenue generated from a flight.
  • The DOT issued a press release dismissing the study's findings, claiming that the data used was too narrow and did not support the authors' conclusions.
  • The authors call for the DOT to reassess the rule and clarify its policy on fines, which they argue is driving airlines to preemptively cancel flights to avoid penalties.

See related report: Airlines prepare for stranding rule

They suggest the rule actually constitutes harm to the public welfare to the tune of USD4 billion over 20 years compared to DOT's projected public welfare gains of USD69.1 million over 20 years.

"The DOT estimates that new tarmac rules will prevent 113,441 passengers from spending an average of 3.64 hours on the tarmac waiting for departure each year," said Jenkins and Marks. "Our models, corroborated by real-world experience of major airlines, indicate that 406,000 passengers will be on 5,200 flights that cancel solely because of tarmac delay rules. The May 2010 operating data suggest these long-term models are overly conservative given risk-averse airline behavior. We offered a range between USD187 million to USD3.9 billion because we just don't know what the ultimate impact will be," said Jenkins. "At 5,200 annual cancellations, the net public cost is $3.5 billion over 20 years. At 5,924 annual cancellations (using 4x DOT's baseline of 1,481 tarmac delays) the net public cost is $3.9 billion over 20 years."

They suggested a four-hour cap is much easier on all concerned. "A choice of four hours for taxi-out limits would reduce cancellations," said the authors. "A 4-hour standard was clearly demonstrated by DOT to have greater public net benefit than a 3-hour standard. It estimated that a 4-hour tarmac limit produced higher net present value, USD73.4 million versus USD69.1 million, but the Department chose to implement the more aggressive 3-hour standard. The choice of three hours came under significant lobbying and pressure from Congress. For that reason, we believe that DOT is unlikely to change its 3-hour standard for tarmac delays without significant outside pressure from consumer groups and legislators."

Jenkins noted that the airlines brought the rule on themselves through inaction. "The real loser in all this, however, is the consumer," he said. "Now, the DOT and airlines need to work together to find a solution that doesn't harm the customer. We're just trying to make travel better for everyone. The three-hour rule never made sense."

The authors also took a look at the fines, demonstrating how outsized they were in relation to the revenue that could be made from a given flight. They indicated they would be 200 to 300 times the revenue that could be expected from the flight.

Maximum fines

The battle is joined

Even as the two authors were wrapping up their webinar on their study, the DOT issued a press release - right on cue - that the study "offers a misleading and premature assessment of the impact of the new passenger protections." The DOT statement called the use of only one month of airline on-time data "far too narrow to yield defensible conclusions about future airline trends," it said, in a desperate attempt to change the subject.

It latched on to cancellations but failed to counter the study's conclusions. "The data reported in May 2010 does not support the industry consultants' claims about rising numbers of airline cancellations. While there was a slight increase in airline cancellations in May 2010 compared with May 2009, an analysis of May cancellations over the last 15 years shows that cancellations in May 2010 were below average. In fact, the rate of airline cancellations in May 2010 was below the average rate for all 15 previous Mays with comparable data: 1.24% in May 2010 compared to the average of 1.51%."

But DOT arguments are clearly designed to mislead. It described the authors as two business consultants for airline companies, when, in fact, the two academics did the study on their own without the financial participation of airlines. Indeed, Marks and Jenkins suggested that some airlines would not want such a study issued since it might distract the DOT from other issues before the department.

More importantly, however, DOT completely ignored the results of the study which said the department underestimated cancellations resulting from the rule. What it needed to do is counter the argument made by Jenkins and Marks that the department predicted in its regulatory analysis that there would only be 21 additional cancellations per year as the result of the rule when in fact there were 140 just in May alone.

"It assumed minimal taxi-in/out time for return to gate and that gates would be available," said the authors. "DOT assumed no material change in diversions and passenger re-accommodation from disrupted flights would max out at nine hours. Actually, that real world, based on May 2010 statistics indicates there is a 14.16% increase in cancellations or a 4:1 overall cancellation ratio from DOT assumptions. More importantly, it has to be noted that May is not even the peak season for tarmac delays. That comes during June, July and August.

"In May 2009, there were 35 tarmac delays greater than three hours," the Airline Zone study continued. "Of those, 34 occurred on taxi-out and one occurred after landing, contrary to the DOT's assumption that they occur in all phases of flight. In May 2010, there were five, of which four occurred after diversions and only one on taxi-out. That means 140 flights were cancelled to achieve the elimination of 35 tarmac delays in May. DOT's analysis is based on 1,481 annual tarmac delays. Multiplying that by four get's you nearly 6,000 annual cancellations based on May's observations."

Recalculating Welfare

"Limiting cancellations and delays was at the core of DOT reasoning," said Marks. "But our study shows they have far surpassed their projections just in the first month."

Jenkins agreed. "If assumptions made by the DOT are off by a factor of 10 in only a month of data, what will it be with a year's worth of data," he asked.

While it is true that they only have one month of post-rule data, that data showed, "the assumptions DOT used in its regulatory analysis missed by a mile," he concluded. Perhaps the answer will come when the General Accountability Office issues its report on the impact of the rule, but for now, the smart money should be on Marks and Jenkins.

The department's obfuscation and rhetoric, perhaps, signals better than anything else how receptive the department will be to revisiting or revising the rule. It also signals the answer to a major question raised by the study - how heavy handed the department will be in levying fines.

Dead on arrival?

The report was delivered yesterday to DOT Secretary Ray LaHood and called for the department to re-assess the rule and, at the very least, clarify its policy on fines which Marks and Jenkins said is driving the overly-cautious actions of airlines to avoid them by pre-emptive cancellations just as they did with the massive snowstorms in February which resulting in a net cost to the industry USD170 million.

What is interesting about the study's estimated cost of the rule is the fact it is so low. For instance, the study did not calculate the costs of cancellations such as occurred in February nor did it count the missed opportunity costs incurred by passengers from flight cancellations.

Marks and Jenkins indicated airlines are cancelling any flight that is perceived to have risk exposure to the new rule and far more than would be the case were the rule to be relaxed. "The threat of punitive fines is driving extreme risk aversion," said Jenkins. "So the DOT needs to clarify its position immediately and if it does impose the maximum fines it needs to acknowledge the real world consequences."

The study found that not only was the DOT's economic analyses woefully inadequate but it underestimated the average time it would take for a cancelled passenger to be rebooked by half. Its assumptions were based on outdated studies done at a time when load factor was 70% versus the 85%+ experienced in today's world, said the study. Marks and Jenkins found the wait for a new seat after a flight was cancelled was 19 hours rather than the nine presumed by the DOT.

"Nineteen hours is just unacceptable," said Jenkins. "The reality for most of these 406,000 passengers is a day or more of time spent waiting for an available seat."

Delays: May-2010 vs May-2009

DOT's regulatory analysis also did not account for indirect cancellations nor did it account for any changes in the less-than-three hour expected taxi time which airlines have unilaterally capped at two hours and 30 minutes in order to avoid the fines.

DOT guidance on fines

"If the department does not intend to seek punitive and disproportionate fines for relatively minor infractions of the tarmac delay limit, it needs to clarify its position urgently in a public and transparent manner," they said. "This will reduce unnecessary flight cancellations and diversions."

Jenkins and Marks now have their answer it seems which is not unexpected since their study even quotes Secretary of Transportation Ray LaHood's attitude. "We just leveled a $16 million fine which was the maximum fine we could level against Toyota," said LaHood in April, three days before the tarmac rule became effective. "So I don't think anyone thinks that Ray LaHood is not going to have strong enforcement."

A day later, as an example of the mixed messages the authors are trying to clarify, the DOT guidance indicated fines would be based on harm, compliance disposition, economic condition, and ability to pay. There were five 3+-hour delays in May and it will be very instructive to see how the department will handle them. The authors noted that DOT's public posture and recent record illustrates its intention to levy maximum fines and the DOT statement seems to confirm this.

DOT assumptions

The study's most compelling arguments take apart DOT assumptions one by one. But it is clear from the DOT response that the department's is not about to let the facts get in the way of a good political opportunity.

  • DOT assumed tarmac delays were caused by the same thing but the authors found that tarmac delays on taxi-in, taxi-out and diversions are caused by different factors, occur at different frequencies and have differential impact on passenger welfare.
  • DOT suggested they were not seasonal and are stretched out throughout the day but the study found they are highly seasonal and are peaky - occurring in large numbers on single days. They are primarily afternoon events during the summer.
  • DOT suggests airline scheduling is the culprit while the study found that tarmac delays are driven by severe weather that block airports and airspace. They also concluded that flight scheduling cannot reasonably reduce exposure to tarmac delays.
  • DOT suggests tarmac delays are spread throughout the system while the Airline Zone report found they are concentrated at specific airports and relatively rare elsewhere and end in a completed flight rather than a diversion or a cancellation.
  • Tarmac delays on taxi-out are primarily three to four hours in length, caused by weather and airspace factors that often lessen within a four-hour window.

"The DOT said tarmac delays are not seasonal but that is a fundamental misunderstanding," said Marks. "These are mostly summer events peaking in June, July and August. In addition, when they occur they do so in clusters as part of a regional event mostly associated with weather. This is critical to understand in terms of gate resources. That means there are few options for airlines because these events usually paralyze an entire airport. Based on May 2010, at a time when operations were at similar levels to May 2009 with improved weather conditions year on year, there were 40% more cancellations and 25% more diversions."

Time of day tarmac delays: by month

3 hour taxi‐out delays by day: May-2008 to Apr-2010

Tarmac 3+ hour events: 2009 (Blue Columns) vs national weather service
severe events (Red Line) (Correlation Coefficient = 0.84)

2009 departures, 2+ hour tarmac delays and delays caused by airspace/NAS factors

They indicated that for every one cancellation of an outbound flight, there is a corresponding cancellation of that same aircraft's inbound or onward flights meaning that one cancellations have a cascade effect throughout the system. Marks and Jenkins based their conclusions on a case study of storms hitting DFW.

Their other case study described the scenario that occurs with JetBlue when weather hits the northeast where a disproportionate number of disruptions occur. What is interesting is the fact the FAA has found it increasingly difficult to resolve the problems surrounding JFK, Newark, LaGuardia and Philadelphia owing to political pressure from Congress. Perhaps the tarmac rule is also DOT's answer to dealing with the real issues.

3-hour taxi-out delays by airport: May-2008 to Apr-2010

In the JetBlue scenario, severe weather comes it at 4 pm, just as flights from the north and south are inbound to the airport. Meanwhile, flights to the west are awaiting takeoff but are blocked completely by the storm. That means gate space is unavailable for the inbound flights forcing diversions and, later, cancellations.

The northeast accounts for only 16.5% of departures but is responsible for 62% of delays but the impact is not confined to the northeast since it ricochets throughout the system leading to the only conclusion - airspace is a major factor, said Marks.

"You cannot schedule your way out of tarmac delay exposure," he said, adding airport and gate constraints force airlines to choose between cancellation and tarmac delays. "Having an extra gate or two is not going to solve the problem. Flight schedule reduction won't get you where you need to go and NextGen may have an impact but it won't solve the impact of the rule."

The study compared operations before and after the rule. "Before the rule," they said, "the westbound flights would be pushed to the tarmac waiting for the storm to pass which usually takes a couple of hours. That makes gates available for the inbound flights and allows westbound flights to depart once the weather clears. But after the rule, cancellations and diversions go up as inbound flights are diverted to other cities and westbounds are cancelled to use the aircraft for return north-south flights. For an international carrier, this can be especially critical since a diversion could result in an even bigger disruption such as what happened to Virgin Atlantic at Hartford where there were no clearance personnel. That could be addressed by greater government cooperation but it would be nice not to have the diversion in the first place."

Pop‐up storms over PA (62% of NYC tarmac delays cross this area)

The authors reported that there are now up to six times the number of gate returns versus flights that would statistically have three-hour delay. "We estimate 75 departure cancellations were due to gate shortages and tarmac rules," they said. "The key lesson here is that tarmac-related cancellations have at least a 1:1 impact on follow-on cancellations due to aircraft availability or network recovery."

Completed flights by taxi time (>120 minutes): May-2009

Cancellations and returns to gate for flights with 2-3 hour taxi‐out times

He added that they were gratified the department took note of the study, adding they look forward to sharing the information with the DOT. Good luck with that.

Want More Analysis Like This?

CAPA Membership provides access to all news and analysis on the site, along with access to many areas of our comprehensive databases and toolsets.
Find Out More