In this issue:
- TSA behavior-detection blasted by GAO
- Airport slots and deregulation
- Increased TSA security fee—or tax?
- Expanding U.S. airport capacity
- Armed TSA screeners?
- News Notes
- Quotable Quotes
In one of the hardest-hitting GAO reports I've ever read, Congress's auditing organization has, in effect, said that the TSA's Screening of Passengers by Observation Techniques (SPOT ) program does not work and should be defunded. Members of Congress asked GAO to answer two questions:
- To what extent does available evidence support use of behavioral indicators to identify aviation security threats?
- To what extent does TSA have data necessary to assess the effectiveness of the SPOT program in identifying threats to aviation security?
The answer to the first is that there is no such evidence, and to the second is that TSA does not have such data. This is laid out in 55 pages of text plus seven appendices. ("TSA Should Limit Future Funding for Behavior Detection Activities," GAO-14-159, November 2012)
TSA says the purpose of the program is to identify high-risk passengers based on behavioral indicators that indicate "mal-intent." Accordingly, its cadre of Behavior Detection Officers (BDOs) are trained to size up passengers as they await screening using a memorized checklist of behaviors indicative of stress, fear, or deception. Passengers with a sufficiently high point score are taken aside for an interview, a pat-down, and a search of their belongings. Assuming nothing bad is found, and the person's behavior does not "escalate," that's the end of the process and the passenger gets back in line. But if the behavior reaches a pre-defined threshold, a law enforcement officer (LEO) is summoned to further question the passenger and decide if an arrest is warranted. The initial (pre-LEO) encounter takes an average of 13 minutes. The program started in 2007 and has grown to about 3,000 BDOs working at 176 airports, at a current annual cost of around $200 million.
The GAO team reviewed two TSA studies of the SPOT program and found both to be non-rigorous, with considerable flaws in their methodology. It then carried out a literature review and a meta-analysis of research studies on "whether nonverbal behavioral indicators can be used to reliably identify deception." And the answer is that "research from more than 400 separate studies on detecting deceptive behavior based on behavioral cues or indicators found that the ability of human observers to accurately identify behavior based on behavioral cues or indicators is the same as or slightly better than chance." GAO provides excerpts from several of these studies, by entities such as RAND Corporation, DOD's JASON program, and MITRE Corporation. Another section of the report documents the wide variation in referral rates by BDOs at various airports, as well as presenting evidence on the subjective nature of some of the behavioral indicators BDOs are taught to look for.
But the most damning information of all is who actually gets identified as "high risk" and referred to a LEO. Not a single potential terrorist was identified by the BDOs. Those who ended up arrested were for such matters as possessing fraudulent documents, possessing prohibited or illegal items, having outstanding warrants, being intoxicated in public, being in the country illegally, or disorderly conduct. While all those things may be law violations, not a single one is, per se, a threat to aviation security. And yet the only measure TSA has for the alleged effectiveness of the program is the referrals to law enforcement.
Nonetheless, TSA has recently conducted a "return-on-investment analysis" which it claims justified the SPOT program. Despite zero evidence that the program can detect or deter aviation-oriented terrorists, the analysis assumes that the BDO "layer of security" prevents a catastrophic (9/11-type) attack. GAO dryly notes that "the analysis relied on assumptions regarding the effectiveness of BDOs and other countermeasures that were based on questionable information."
In response to previous GAO and Inspector General criticism, TSA is developing a new set of metrics about SPOT, but says it will require at least an additional three years and additional resources to report on the program's performance and security effectiveness. Meanwhile, it is asking for a budget increase to add 584 more BDOs so the program can expand to smaller airports. In other words, to paraphrase a familiar line about the recent federal health-care law, we have to keep running and expanding the program to see if it works.
GAO sums up this comprehensive assessment as follows: "10 years after the development of the SPOT program, TSA cannot demonstrate the effectiveness of its behavior detection activities. Until TSA can provide scientifically validated evidence demonstrating that behavioral indicators can be used to identify passengers who may pose a threat to aviation, the agency risks funding activities that have not been determined to be effective."
Thirty-five years ago Congress enacted the Airline Deregulation Act of 1978, removing the power of the now-defunct Civil Aeronautics Board to decide which airlines could serve which airports at which fare levels. Competition would replace the heavy, arbitrary hand of government in deciding who flew where. This major policy change followed several decades of research that demonstrated that airlines are not a monopoly like electricity and water supply. Rather, airline service is "contestable."
Alfred Kahn, a principal architect of airline deregulation, often cautioned that freeing airlines was only part of the job. The infrastructure of airports and air traffic control also needed to be opened up to market forces if deregulation was to produce sustainable competition. The fact that government still defines and enforces slots at key airports, as dramatized in the recent Justice Department (DOJ) settlement with American and US Airways, illustrates how much government still intervenes in the market for airline service.
Under the terms of the settlement, it's not just that American and US Airways are required to divest slots and gates at a number of airports. In the cases of Reagan National (DCA) and LaGuardia (LGA), DOJ is attempting to restructure the airline market. First, it is requiring the merged carrier to divest to JetBlue the 16 DCA slots AA currently leases to that airline, and also requiring it to divest to Southwest the 10 slots it leases from AA at LGA. Second, the merged carrier must sell a further 88 slots at DCA and 34 at LGA, but only to airlines approved by DOJ, which says it wants these transactions to "guarantee a bigger foothold for low-cost carriers at key U.S. airports."
This attempt by DOJ to restructure the airline industry is completely at odds with airline deregulation. Instead of allowing market forces to work out the highest and best uses of existing airport capacity, DOJ's central planners seek to impose their vision of what is best. There are all kinds of countervailing views of what is best. The four leaders of the House and Senate transportation committees sent a letter to DOJ last month arguing that all airlines, not just DOJ's favored LCCs, should be able to bid on the divested slots. Their view of the public good favors legacy carriers that use smaller regional jets to link small cities to DCA and LGA, rather than LCCs like Southwest that serve larger cities. But there is no non-arbitrary way to resolve such questions administratively. The DOJ approach inevitably means government picking winners and losers.
And that is an inherent defect of allocating the scarce resource of airport capacity via slots. A far more transparent and market-based way to do this is via runway pricing. To some academic economists, slot auctions and runway pricing are seen as equivalent. But in the real world they are profoundly different. Any proposal to make slots a more effective means of allocating airport capacity runs into the obstacle of incumbents versus new entrants. Although DOT has always maintained that slots are not property and can be changed at any time, incumbents will fight tooth and nail to retain the slots they already have, dooming any slot auction proposal to be tinkering around the edges.
By contrast, airports already charge for runway use, so if a capacity-short airport like DCA or LGA decided to exercise its legal right (under DOT's airport rates and charges policy) to charge congestion-based fees for runway use, those fees would apply to all users—incumbents, new entrants, and foreign carriers. When DOT changed its rates and charges policy in 2008 to allow such pricing, the airlines challenged it in court—and lost. So that issue has already been decided.
Here is how a hypothetical system at DCA and LGA might work. For each schedule season (e.g., quarterly), an airport's pricing body would propose a pricing schedule aimed at minimizing delays while complying with FAA guidance on maximum hourly runway use. Airlines would revise their schedules in response, and the pricing body would model the new schedules' impact on congestion, adjusting the price schedule and resubmitting it to the airlines. After several rounds of this discovery process, the final prices and schedules would be set for that season. This iterative process is based on strategic simulations carried out by the FAA-sponsored NEXTOR project on congestion at the New York airports. The process functions as a kind of periodic auction, not of "slots" but of runway capacity. It would also be better for the small numbers of business jets and fractional operators using airports like LGA than any kind of slot system, since those non-scheduled users could gain access in real time based on being willing to pay the current market price to use the runway. All the increased runway revenues could be dedicated to capacity increasing projects.
In my ideal world, an entity such as DOJ would not be telling merging airlines that they must reduce service at any airport, since the pricing approach above would be making those decisions. But even in the current world of DOJ-required reductions, the process of deciding how the freed-up capacity would be used could be done via this kind of pricing system. I am not wise enough to predict whether the JetBlues and Southwests would be willing to pay the price for expanded service at DCA or LGA—or whether legacy carriers like Delta and United would pay more. But I do think a pricing system would shift some of the air service now using DCA and LGA to outlying airports that also serve the respective metro areas. And that is how it should be. The runway capacity at DCA and LGA is a scarce resource, and should be allocated via market pricing, not central planners' designs. A runway pricing system is the only practical way to do this.
And were Alfred Kahn still with us, I'm sure he would agree.
Both the House Republicans and the Senate Democrats are supporting an increase in the current $2.50 per segment passenger security fee, which generates revenue that pays for a portion of TSA's $5.2 billion aviation security budget. The airlines are up in arms, attacking the proposed increase as a tax increase in disguise. Airlines for America has recruited the Air Line Pilots Association, the Global Business Travel Association, IATA, Consumer Travel Alliance, and the Regional Airline Association to help fight the proposed increase.
The Senate has passed a 2014 budget that includes the White House proposal to replace the current $2.50 per segment fee with a flat $5.00 per one-way trip, increasing by 50¢/year through 2019, when it would reach $7.50. That would raise $25.9 billion over 10 years (averaging $2.9 billion/year). The House Budget Committee's 2014 budget includes a milder version—just a one-time switch to $5.00 per one-way trip, with no planned increases thereafter. CBO estimates that the House version would raise an additional $11 billion over 10 years ($1.1 billion/year).
A4A attacks both proposals as only adding to the tax burden on airlines and passengers, reducing (at the margin) the number who choose to travel by air. (I will leave aside the inconsistency of airlines' significant increase in fees for things like checked and carry-on bags, which apparently don't suppress demand.) A4A president Nick Calio's recent op-ed in The Hill makes two other points. TSA provides some degree of security for other modes of transportation, but only airlines and their passengers have to pay for this; the others get TSA services at no charge. Also, TSA would be providing no additional services to airlines or passengers in exchange for the increased fee level.
But nowhere in any of the airline material have I seen any mention of what I think is the strongest argument against the proposed fee increase. Jeff Davis of Transportation Weekly pointed out the dirty little secret of these proposals in his Nov. 26th issue. As he explains, the existing $2.50/segment fee is categorized as a "discretionary offsetting collection." That means the $2 billion a year it generates is subtracted from TSA's $5.2 billion aviation security budget, for a net budgetary cost of $3.2 billion. "But in order to make sure that any fee increase goes to deficit reduction, the Ryan budget assumed that the increased portion of the fee would be classified as mandatory receipts or revenues, deposited in the general fund, and could not be used to offset any TSA spending."
Bingo—the increase is a tax, not a user fee. The proceeds of the increase go into the general fund, to reduce the deficit. The same is true of the Administration proposal adopted by the Senate majority. I know the explanation is arcane, but there it is in black and white. Both House Republicans and Senate Democrats actually are trying to dress up a tax increase in user-fee clothing. They are singling out this one industry for a special tax to reduce the deficit. The airlines and their coalition should make this point explicitly.
The Eno Center for Transportation last month released a good report called "Addressing Future Capacity Needs in the U.S. Aviation System," funded in part by the U.S. Travel Association. (Go to: www.enotrans.org/store/research-papers/capacity-needs-aviation)
Based on the latest FAA aviation forecast, it identifies a need for significant expansion of capacity, at major airports as well as in the air traffic control system. After providing some useful context, the report presents case studies of the capacity problems at four major international hubs: Kennedy (JFK), Newark (EWR), Los Angeles (LAX), and San Francisco (SFO). All four will need additional capacity, but have serious physical as well as political constraints. The report offers a useful discussion of policies, in addition to adding physical runway and terminal space, that could help.
In terms of operational changes, it contrasts the regulatory approach with a pricing approach, and finds advantages in the latter, while acknowledging likely opposition. One way to counter opposition might be a congestion-relief discretionary grant program (as part of a reformed Airport Improvement Program). Like the DOT's generally successful Urban Partnership Agreement competition from last decade for surface transportation, it would offer competitive grants for airports willing to implement some form of pricing for congestion relief. The report also highlights how various airspace and procedural improvements from FAA's NextGen program could increase the effective capacity of existing airport runways. And it points out that AIP grants go disproportionately to "non-primary" airports that handle just 0.25% of all airline passengers.
The report concludes by offering four main policy recommendations. First, restructure the AIP grant program to focus more on major airports where improvements are most needed and would yield the greatest national benefits. Second, use part of AIP monies for a competitive program to help airports implement congestion-reducing measures such as pricing. Third, explore separating the ATC function from the FAA, to enable the NextGen modernization to be done in a more commercial manner. And fourth, reduce current restrictions on airport use of locally imposed passenger facility charges (PFCs)—at a minimum, give FAA the discretion to permit increases above the current $4.50 cap for projects that would accommodate increased demand.
By and large, these are sound recommendations, though all of them face serious political obstacles. While I think the odds of commercializing the FAA's Air Traffic Organization are increasing, the airport recommendations look less likely to gain a critical mass of congressional support. The most likely one, in my view, would be to allow airports to implement higher PFCs. As the exact opposite of a "federal tax increase," an expanded local self-help charge should appeal to the growing number of those in Congress who want to devolve functions from the federal government to state and local entities.
After the tragic shooting last month at a checkpoint at LAX, in which three TSA screeners were shot, one fatally, the immediate response of TSA union president David Cox was that TSA's workforce should be expanded to include a new category: armed officers with law enforcement status, to protect all checkpoints from such attacks. The new officers would be stationed on raised platforms allowing them to survey the whole checkpoint area. And on top of that, the union is calling for more Behavior Detection Officers, despite there being zero evidence that BDOs can spot people with mal-intent (see previous story).
As many observers pointed out, those who are intent on mayhem can easily choose other targets than screening lines if they want to cause trouble. The advantage of existing law enforcement officers on duty at airports is precisely that they can and do roam, covering ticket lobbies, curbside areas, baggage drop-offs, etc. in addition to checkpoints.
The union's knee-jerk reaction is typical of political responses to aviation security incidents. The focus is on making 100% certain that this particular incident cannot occur again. But as numerous security experts keep reminding us, there is no 100% security at any conceivable amount of spending, If airports were made 100% secure at huge costs, other targets would be favored: railroad bridges, electrical transmission lines, sports stadiums on game days, shopping malls (as in Kenya), etc. As MIT aviation safety expert Arnold Barnett told the New York Times after the LAX shooting, "Wherever you establish a security perimeter, by definition there is stuff outside it."
TSA has already grown far beyond the original concept, and mission creep is an ever-present temptation for a host of interests—not just screener unions but equipment vendors, the TSA bureaucracy, and congressional oversight committees. Let's hope everyone can keep their heads and not implement yet another so-called layer of costly, ineffective security theater.
Brazil Privatizes Two More Airports. Last month Brazil announced the winners in the bidding for the Galeao International Airport in Rio de Janeiro and Tancredo Neves International in Minas Gerais. The Rio airport will be leased for 25 years to a consortium of construction giant Odebrecht and Changi Airport Group, which bid $8.6 billion. The Minas Gerais airport went to Brazil's CCR, for 30 years, for a bid of $827 million. State-owned airport agency Infraero will hold 49% of each airport enterprise.
Las Vegas Solution to TSA Exit Lane Decision. The TSA's recent decision to stop providing staffing for the exit lanes from the secure portion of airport terminals, while still requiring that entry there be prevented, has left many airports complaining about a costly unfunded mandate. McCarran International Airport in Las Vegas has decided to automate this function, with TSA's approval. The automated exit lane gate system is already in use at the airport's newest terminal (T-3) and will be added at the other terminals. The system provides real-time notification to security personnel of any attempts to enter the terminal via the equipped exit lanes.
Moody's Sees Growth at Major Hubs. Rating agency Moody's last month released its FY 2012 US airport medians report on airports whose bonds it rates. The figures show continued growth at large hubs but slight down-trends continuing at medium and small hubs. There was also a slight improvement in airport liquidity compared with FY 2011, with cash on hand increasing to 522 days of operating expenses, compared with 511 days the year before.
Value of Time Lost to Passenger Screening. For a recent Reason Foundation blog post, I discussed the time passengers have to spend waiting in TSA screening lines. Currently, about 550 million passengers begin trips at US airports each year. If the average delay, compared with pre-TSA days, is 15 minutes, that amounts to 137.5 million hours per year. What is the economic value of that lost time? Let's assume that the average business traveler's lost time is valued at $50/hour and the average leisure traveler's time is $15/hour. Since business travelers are about 50% of the daily airport throughput, that leads to a total imposed cost of just under $4.5 billion a year. And that does not count the additional time that many people have learned to add to the beginning of their trips (buffer time) to minimize the odds of missing their flight if the lines are longer than usual on the day they fly. If that is an additional 15 minutes on average, the total economic cost of the time lost due to TSA screening is $9 billion a year.
Landmark Noise Settlement Approved in Fort Lauderdale. The battle to achieve a workable settlement in the long-running noise compensation case over FLL's runway extension has been settled, thanks to an 8-0 vote by the Broward County Commission approving the deal worked out with the city of Dania Beach. The runway extension has been under construction for most of this year, but the terms of the settlement were still being ironed out by county and city negotiators, since the FAA had rejected the initial settlement. Airport Director Kent George told local media that FAA was involved in crafting the settlement, which he hopes will mean agency approval.
Traffic-Short Rome Airport Gains New Airline. Cutbacks at struggling Alitalia Airlines have threatened planned expansion plans at Rome's privatized Fiumicino Airport. But last month saw an announcement by Spain's up-and-coming low-cost carrier Vueling that it will establish its second-largest base at the airport. Vueling will base eight A320s there, and plans to serve up to 550 city-pairs via Fiumicino. Its current Barcelona hub serves 1,200 routes.
PreCheck Expanding to All Military. TSA announced on November 13th that PreCheck screening will be available to all armed forces service members effective Dec. 20th. This includes members of the Coast Guard, Reserves, and National Guard. PreCheck is now available at 100 US airports. Active-duty military will submit their DoD identification number when making flight reservations, and the TSA system will define it as their Known Traveler Number.
Privatize TSA Functions, Says Cato Institute Report. A policy study from the Cato Institute recommends that the TSA's functions be divested to other entities, including the devolution of airport screening to individual airports. That would follow policies in most of Europe that make airport screening the responsibility of each airport, under the supervision of national security authorities. Most European airports outsource screening to certified screening companies. (http://www.cato.org/publications/policy-analysis/privatizing-transportation-security-administration)
"This report [GAO-14-159] represents a serious indictment of TSA's Screening of Passengers by Observation Techniques program. With a single report, GAO has displayed that the science behind the program is non-existent and that the study TSA cites in defending the program was fundamentally flawed. Given TSA's unwillingness to concur with GAO's recommendation that the agency limit funding for the program until it could be proven, it is now up to Congress to take a hard look at reprioritizing the funding for this program. I trust that my colleagues will do just that."
—Rep. Cedric L. Richmond (D, LA), quoted in "TSA Behavior Detection Program Should Be Defunded," news release, Committee on Homeland Security – Democrats, Nov. 13, 2013
"The FAA defines primary airports as having more than 10,000 passenger boardings each year. . . . While only 0.25 percent of enplanements occur at non-primary airports, . . . 35 percent of AIP grants are obligated to non-primary airports. This is a clear misallocation of resources, given that the federal interest in the nation's aviation system should be in making investments with the greatest national benefits. It is highly unlikely that these non-primary airports, which are receiving approximately $1.35 billion annually while carrying only 1.8 million annual passengers, are the best investment of scarce resources."
—Joshua Schank, et al., "Addressing Future Capacity Needs in the U.S. Aviation System," Eno Center for Transportation, November 2013
"[G]iven the rarity of attacks like the one at LAX, none of these steps may save a single life. They make about as much sense as putting the National Guard in movie theaters in the wake of the Aurora, Colorado massacre. Actually, they make less sense, in light of the risks of introducing thousands of guns into small, crowded spaces filled with people who really hate being there. . . . Equipping screeners with deadly weapons would also heighten the sense of coercion and intrusion that makes air travel resemble admission to a medium-security prison. Being ordered around and physically groped by a uniformed officer is bad, but it would be worse if he had a Glock on his hip."
—Steve Chapman, "The Folly of Arming TSA Agents," Nov. 7, 2013 (www.reason.com/archives/2013/11/07/the-folly-of-arming-tsa-agents)