A Complex Balance

The Boeing 737 MAX Crashes

The recent tragic crashes of Boeing 737 MAX airliners have raised concerns that commercial aircraft are too complicated to operate due to the increasingly complex technologies involved. Notably, Donald Trump tweeted “Airplanes are becoming far too complex to fly. Pilots are no longer needed, but rather computer scientists from MIT.” Others disagree, saying technology has contributed to airlines’ astounding safety record. So as the investigations continue, more attention is paid to how airlines regulate safety measures in the first place. Have regulators defaulted too much power to corporations?

For economists, the relationship between people and technology is not driven by technical needs, but by costs and business strategies. As Harry Braverman showed in Labor and Monopoly Capitalism, the varieties of capitalism mean that technology and the organization of labor and production (including flying airplanes) always involves a conflict over who controls the time, pace, and content of work, and who gets the profit. Moreover, there’s an ongoing historical tension between public regulation of products and services versus government enabling lower costs and higher profits at the expense of worker and consumer safety.

Still, technological improvements can support what economist Bill Lazonick calls “sustainable prosperity,” where the increase in productivity provides both corporations and the economy with higher output linked to stable employment and equitable sharing of profits. But those goals come with relentless pressure to reduce costs, and commercial air travel is no exception. Consider jet engines. On your next trip to the airport, notice how few commercial planes have three or four engines, compared to just two. Fuel is one of the most significant expense items for airlines, comprising around twenty percent of the total operation cost

Boeing is sharply limiting future productions of its iconic four-engine 747 to cargo planes, while Airbus is canceling its four-engine A380 entirely in favor of smaller two-engine planes.

This illustrates the complex (and costly) relationship between technology and design. There is a significant premium on using existing aircraft when possible. Delta Airlines has perhaps the oldest fleet among major commercial airlines, using planes with an average age of 17 years. But age is not necessarily a barrier, and planes are often retired for economic reasons (especially inefficient engines) rather than safety ones. As long as safety standards are met by inspection, they can keep on flying. Using the same aircraft also reduces the cost of maintenance and labor training. Also, the costs of developing brand-new aircraft are huge; Boeing’s new “mid-market” airliner, the 797, may cost the company well over $15 billion in development expenses before it ever flies.

In contrast, the first Boeing 737 was put into service in 1967. The 737 MAX is the latest iteration, introduced in 2017. Boeing had planned an entirely new aircraft to replace the 737, but when Airbus pushed ahead with a redesigned 320 with greater fuel and flight efficiency, Boeing shifted to redesigning the 737. This decision paid off; the 737 MAX became Boeing’s best-selling commercial aircraft in history.

The Crash

The cause of the crash is apparently due to a combination of sophisticated product design and commercial pressures. Boeing added new, larger, and more fuel-efficient engines to the MAX. But the wing locations for the new engines changed the plane’s long-proven aerodynamics. In response, Boeing then added new software, the Maneuvering Characteristics Augmentation System (MCAS).

MCAS is triggered by two sensors that assess the flight angle, pushing the nose down when the angle is too steep. But if those sensors give inaccurate data, it can lead to a false MCAS triggering, repeatedly forcing the plane’s nose down, even if it was in level flight. Right before the crash, records indicate that MCAS operated repeatedly and unnecessarily, pushing the aircraft nose down despite the pilots’ attempts to maintain it under control. There was no indication that the data might be faulty.

Moreover, Boeing has two more warning systems to alert pilots to inaccurate sensor readings, but they were optional, and adding them would have increased the purchase price. MCAS in the basic 737 MAX could be triggered by only one sensor and does not have these additional warning systems as standard equipment.

Another suspicion was directed towards MCAS’ interaction with human pilots. Despite misinformed commentary, airplanes don’t simply fly themselves. Human pilots control them, just as humans carry out laparoscopic surgery and scan your groceries. And increasing technological complexity can lead to unexpected dangers. As decision scientist Gary Klein notes, automation “often creates new demands, new types of errors, and new paths towards failure.” [1].

Flying is complex, especially when things go wrong, and pilots sometimes have little time to respond to problems that could be caused by a multitude of factors. There is a standard operating procedure for a 737 “pitch” problem, but having just taken off and faced with repeated automatic dives, the Lion Air pilots might not have responded correctly. Even more worrisome, the Ethiopian Airlines pilots reportedly followed recommended procedures, but also lost sight of other factors like leaving power at full throttle.

The last line of defense in an airliner isn’t technology — it’s the human pilots. But highly trained and experienced pilots are costly. Overall, all labor (not just pilots) makes up around 30 percent of airline operating costs, so there is constant pressure to reduce costs, and automation is a key strategy. Pilot unions, naturally, want higher requirements, more training, and higher wages. Airlines aggressively bargain down wages and working conditions, reducing the number of on-board pilots and cutting their rest periods and training.. They are also facing pilot shortages, leading some to lower requirements for pilot training and experience. Technologically, manufacturers seeking to lower labor costs are working on designs for future jetliners that will only need one pilot.

Critics say Boeing failed to alert pilots about the MCAS and how to handle to it, and that Lion Air allegedly skimped on pilot rest periods, training, and maintenance to increase paid flight time. Unions also have noted the insufficient experience for some pilots in the two crashes. Although the Ethiopian Airlines flight captain received many hours of training and the airline had a good safety reputation, the first officer reportedly received only 200 hours of flight time (U.S. regulators require 1500), leading one expert to say “you basically put a student pilot in there.”

A spokesman for the American Airlines pilots union says there should have been more training and alerts from Boeing, including expensive flight simulators with MCAS, which the Federal Aviation Authority (FAA) did not require. Pilots did not receive important instructions about MCAS and didn’t know that it could be activated even when the autopilot was turned off, a step taken by the Ethiopian pilots to no avail. Boeing did not want extra costly pilot training to be a requirement for the MAX, instead preferring that existing “type rating” be used. That is, if a pilot was certified to fly a previous “type” of 737, they would automatically be qualified for the MAX, reducing training costs to airlines.

And who ultimately oversees such decisions? Government. We rely on government regulating a product or service for safety — a new drug, food, a financial product, a car, or a commercial airliner. But economists worry about regulation becoming too costly and cumbersome, fearing it will reduce innovation and raise prices, harming consumers in the long run. Hence their support for cost-benefit analysis; an evaluation of the benefits of regulation vs. its costs (which can include the number of accidents, injuries or deaths). Harvard law professor Cass Sunstein, head of President Obama’s regulatory efforts for three years, endorses this view: “the right stuff is what cost-benefit analysis says is the right stuff.”

But public policy scholars also worry about “regulatory capture,” where the supposedly impartial public regulator comes to argue for the interests they are supposed to be regulating. Inquiries will focus on whether that happened with the FAA and Boeing. In several industries, one way to balance costs and safety in the regulatory process allows companies to self-certify on some issues, leaving more critical decisions for government personnel.

While this might seem an easy process in the abstract, delegation decisions are faced with the same tensions as technology and labor costs. Concerned about potentially excessive bureaucratic delays, Boeing wanted to certify the MAX as quickly as possible, to compete with Airbus. According to the Seattle Times, “several FAA technical experts said that as certification proceeded, managers prodded them to speed the process.” Reports say that FAA management put “constant pressure” on the government’s technical staff to delegate “even more items down to the Boeing Company.”

This can sound familiar to anyone who has worked in government regulation. When I was Assistant Secretary of Labor for pensions, we were continuously lobbied by industry and Congress about regulatory decisions that could result in additional costs for financial firms, even if consumers were left with less information about their pensions.

Congressional hearings will now explore whether such pressures compromised the MAX certification. However, every regulatory process involves balancing complex cost and safety considerations, often in cases where the technical issues are unclear or where reasonable people have different views, in contrast to popular stereotypes of corrupt officials always doing industry’s bidding. Government regulatory agencies have difficulty keeping up with modern technology and must make their own “cost-benefit” decisions about where to deploy their limited resources, which drives the question of how much certification can be “outsourced” to firms themselves.

This concern is now heightened by concerted pressure from the Trump Administration to strip regulatory agencies of expertise and capacity. Journalist Michael Lewis details in his new book The Fifth Risk how conservative anti-government regulatory ideologues are compromising food safety, nuclear waste management, and even weather forecasting. And former Federal Reserve Chairman Paul Volcker is advocating longer-term protection and capacity building for the federal civil service, which has eroded under Democratic and Republican administrations.

But the critical issue for citizens is to not treat regulatory decisions and processes, or technology and labor costs, in isolation from the underlying dynamics of profit-making corporations. Technological innovation must be understood in terms of costs, control of work, how people and technology interact under the conditions and restrictions of the work environment, and how regulation operates in an economically competitive and technologically sophisticated world.

Some suggest that the government should bear all potential safety costs, from regulatory procedures to nationalizing airlines. But extending such reach to all industries will be very expensive and could hinder innovation. And there is no such thing as total protection from technological risks. Instead, safety and reliability must come from a combination of competent government regulation, worker training along with sufficient wages and fair working conditions under union protection.

We still don’t know the definitive reasons behind the MAX crashes, and it may lead to some combination of tighter, less delegated regulation, better training, or more empowerment of front-line workers such as pilots and maintenance personnel. But we will also witness more problems emerging from the same ongoing tensions. Whether it is food additives, breakthroughs in genetic engineering, financial innovations, social media privacy concerns, or measures to address the threat of climate change, our society is left to grapple with these complex issues, with no simple formula for navigating the benefits and risks of technological innovation and market logic.

Rick McGahey is a Senior Fellow, Schwartz Center for Economic Policy Analysis.

[1] Klein, Gary & Shneiderman, Ben & Hoffman, Robert & Ford, Kenneth. (2017). Why Expertise Matters: A Response to the Challenges. IEEE Intelligent Systems. 32. 67-73. 10.1109/MIS.2017.4531230.

Also for you:

Rick McGahey

Previous post

The Moby Dick Problem of War: An Interview with Steve Coll

Next post

The Blessing of Invisible Pain