What happened to Southwest?
If you had to pick only one adjective to describe Southwest Airlines, you might go with “unconventional”. Southwest rose from nothing to become the most successful airline in the country, and at every step of the way, they took an unconventional approach. Here is Thomas Petzinger, Jr. describing the early history of the company in Hard Landing:
Inside the airline industry, however, everyone knew Southwest and only too well. It had never lost money, from the time it was fully established in business. And it had flourished while defying almost every success maxim of the post-deregulation world: it had no computer reservations system, offered no frequent-flier program, did not conduct yield management, and had never organized its flight schedules around anything remotely approaching a hub.
Southwest would belatedly add a computer reservations system, a frequent-flier program, yield management, and even something approaching a hub. But to this day, Southwest is still the most unconventional airline in America. They only fly 737s, they don’t charge for bags, and they don’t assign seats.
It makes sense that the most successful companies would be the most unconventional. An organization that does everything conventionally can only be above-average at best; even if they have best-in-class execution, it will by definition be coupled with a conventional, average strategy. You would expect that the most successful companies to have the best execution and also pick some good unconventional strategies, avoiding the mistakes of their more conventional peers. Indeed, this is the premise of William Thorndike’s The Outsiders, which profiles a set of unconventional CEOs who achieved outstanding success.
Southwest had an operational meltdown over Christmas, stranding thousands of passengers over the holiday. This has been catnip for pundits, each pushing their pet theory explaining how the most successful airline in America could have failed so badly. Was it the route network? Stock buybacks? Incompetent management?
To answer this, let us revisit the collapse of another unconventional company, General Electric.1 GE was one of the most successful and admired companies in the country, and from the mid-1990s to the mid-2000s, it was the most valuable company in America. Jack Welch, their iconic CEO, was one of the most unconventional business leaders of his generation, and his maverick approach frequently paid off. He decided that GE should get out of businesses where they were not #1 or #2, an idea so successful that it has become almost conventional today. He aggressively deployed capital, making big internal investments in airplane engines, gas turbines, and cable television that all paid off handsomely. He demanded that managers hit their numbers, and his rigorous operational reviews were legendary.
The downfall of GE was in Welch’s unconventional strategies that didn’t work. He made GE into a sprawling conglomerate with major interests in media and finance, instead of focusing on industrials as his peers did; the resulting company was unfocused and difficult to manage, and after 2008, the finance side of the business became a millstone for the company. GE had always been unusually acquisitive, which worked well early on when Welch was making good deals, but backfired later on under Welch’s successor, a less talented dealmaker who faced a more competitive deal environment. Finally, Welch insisted on finding a way to report smooth earnings to investors at all costs (often known as “earnings management”), an approach that was going out of style by the 1990s. These accounting shenanigans were a major distraction, and caused GE to make major strategic decisions based on whether they produced results that could be manipulated to smooth GE’s reported earnings, instead of being based on whether they provided value to GE shareholders.
A few years ago, when it became clear that GE was in deep trouble, everyone lined up with their takes about how GE’s failure told us something important about the broader failings of Corporate America. GE was an example of corporate greed, incompetence, or hubris. One journalist even wrote a book about Jack Welch somehow supposedly destroyed capitalism, because apparently no one ever thought about doing layoffs before he came along. (I wonder what he makes of the problems at Southwest, a company that on principle has never done a layoff.) Whatever you thought was wrong with American business, you could spin a story that GE was proof of it.
The problem with all of these theories was that GE was an unconventional company that failed in unconventional ways. GE was admired more than it was truly imitated. No other industrial company was building up lending, real estate, and aircraft leasing subsidiaries funded with commercial paper, nor were they expanding into entertainment. After Enron, most companies improved transparency and stopped trying to report earnings smoothed to the penny, but not GE.
When you read about companies that failed during the financial crisis, you find that they all got into trouble in similar ways, namely by blindly following each other into the similar bad bets on real-estate backed loans. By contrast, when you read about GE, you find a story of a company that got into trouble in mostly novel and unique ways.
Which brings us back to Southwest, which appears to have melted down due to the failure of their antiquated scheduling software. Airlines have been known to have high-profile IT problems from time to time, but no one seems to be able to recall a similar situation involving scheduling software. Paul Krugman cites Southwest’s point-to-point network structure as a possible culprit, but while point-to-point networks are less common, there are plenty of other point-to-point airlines in the world that have operated for decades without problems. Michael Hiltzik of the LA Times says that the problem might be Southwest’s tight scheduling, which leaves less room for error, but again, other airlines in the world operate in the same way just fine.
Zeynep Tufekci writes in the New York Times that this is a case of perverse incentives, that corporate executives will inevitably be tempted to postpone investments in software because the cost will reduce reported earnings today, whereas the payoff will be in the future, when someone else might be in charge. This is absolutely something that happens sometimes at companies, but this is unlikely in a company like Southwest, where executives tend to stay in their roles for decades. Gary Kelly was CEO of Southwest for eighteen years before retiring in February 2022, and he was their CFO for many years before that. Even today, he is still Southwest’s Executive Chairman, and very much in the crosshairs for what happened. This is not an organization where the CEO serves for five years before hitting retirement age and passing the baton; executives absolutely expect to be around to suffer the longer term consequences of their decisions.
Tufekci also advances the popular theory that stock buybacks are to blame; Southwest has spent $8.5 billion on share repurchases in recent years, supposedly enriching management while leaving the company without enough money to upgrade their software. The buyback theory makes no sense. Before returning cash to shareholders in the form of buybacks and dividends, executives choose how much cash to reinvest into the business. For an airline, this might involve investing in planes and software. They figure out how much to invest in planes and software based on the returns they think those investments will generate, and only invest in projects that meet their hurdle rate; sometimes they err in their judgment. The important thing to remember is that executives are absolutely incentivized to reinvest the maximum amount that can be justified, because their compensation (and status) is linked not just to the stock price, but also to the size of the company. As a rule, big company CEOs get paid much, much more than small company CEOs.
No one understands this reinvestment game more than Southwest, which has reinvested most of their profits into new planes for decades. This is how they went from three planes flying in Texas in the 1970s to flying nearly 800 planes all over America today. Their stock is up something like 2,000x since they became publicly traded, and their executive salaries have soared. In the early 1990s, Southwest was a $1 billion a year company, and their top executives were making around $400,000 per year; by the late 2010s, Southwest was doing over $20 billion of revenue annually, and the top executives were making closer to $4 million per year.
Eventually they achieved a size where they could no longer easily grow, and at this point, they started returning excess cash to shareholders. This is what companies have always done with excess cash at that stage; it’s the only thing they can do with excess cash, without making big acquisitions or lighting it on fire. Every sufficiently big company returns most of the cash they generate to shareholders, because they have to. Now, in hindsight, Southwest probably wishes they had invested more in scheduling software, but that’s an error of judgment, rather than an expected consequence of the incentives at play.
(As an aside, this is not to say that open market buybacks shouldn’t be banned. Not because buybacks are somehow detrimental to corporate value or to the economy: they are not, because they are basically functionally equivalent to dividends, which is the main way companies returned capital before buybacks became easier to execute in the 1980s. A billion dollars “spent” on buybacks and dividends both leave investors as a whole with a billion dollars of cash and full ownership of a company that is a billion dollars lighter. No, we should consider banning buybacks because there is no other topic on earth that causes so many otherwise smart people to contort themselves into logical pretzels and embarrass themselves in public. People have convinced themselves that buybacks somehow constitute “waste”, when a buyback is just a company sending investors the cash that already belongs to them in the first place, and that buybacks disincentivize investment, when we have seen that nothing could be further from the truth. Cliff Asness of AQR once accurately termed this phenomenon as “Buyback Derangement Syndrome”. From a societal standpoint, we want companies to return capital to investors when they run out of internal investment opportunities, rather than hoarding it or wasting it, so investors can turn around and redeploy it into more promising productive opportunities (e.g. venture capital or real estate development) if they like. Anyway, everything was fine back when everyone was mostly just paying out dividends, so maybe we should think about just going back to that and saving everyone the brain damage.)
Southwest’s pilots advance the bean counter theory, that Gary Kelly is an accountant who hired other accountants to fill out senior management and none of them understand how to run an airline. People love simplistic narratives about what professional background makes for a good executive, and none of them are supported by evidence. The legendary founder of Southwest, Herb Kelleher, was a practicing lawyer with no operating experience when he took over as CEO in 1981, a decade after Southwest started flying, and he stayed in the role for two decades, leading the company through its glory years.
Jack Welch was a PhD chemical engineer who spent his entire career as an operating manager. His successor, Jeff Immelt, was also a longtime GE operating exec, albeit with more of a background in sales. At the same time that GE was flailing, another industrial conglomerate, Honeywell, was on the rise. Back in the early 2000s, Honeywell was set to be a small bolt-on acquisition for GE, but the deal was blocked by the EU on antitrust grounds. After that deal fell through, Honeywell became one of the top performing industrial companies in the world, growing to eventually become more valuable than GE. The CEO who turned around Honeywell and led it during its ascension was David Cote, an accountant and finance executive who actually spent most of his career at GE, before being pushed out. Some executives are more effective than others, and it seems to have little to do with education or early professional background.
What happened at Southwest? Usually we would like to examine the evidence, but since the post-mortem is still ongoing, let’s indulge in a bit of speculation, based on what happened at GE. Imagine you’re a senior executive at Southwest. You and your team built Southwest from nothing into a colossus, while your competitors have done little more than parade through bankruptcy court. A big part of your success has been your willingness to be unconventional, to zig when everyone else zags. You’ve looked at upgrading your scheduling software, but you studied it and determined it wasn’t yet worth the investment. Your company has always been comfortable with being slow to upgrade technology, and it always worked out in the past. It doesn’t bother you that other airlines have been more aggressive here; why would the industry leader care about benchmarking themselves against the also-rans?
We admire unconventional companies, but an unconventional approach can cut both ways. Unconventional strategies are great when they improve on convention, but detract from performance when they do not. Even the smartest managers, the ones who employ a lot of unconventional strategies that work, will also (unintentionally) employ some unconventional strategies that hold them back, and initially there will be no easy way to tell the two apart. They observe that their strategies are working as a bundle, they conclude that everything they are doing is brilliant, and arrogance and complacency set in. To make matters more complicated, a lot of bad strategies appear to be good strategies as long as the tide is rising. The critics and skeptics are always there, but they are easy to dismiss as long as the results keep rolling in. Until, of course, one day the tide goes out, and the truth is revealed.
Thanks for reading MD&A! Subscribe for free to receive new posts and support my work.
We will revisit GE in more depth in an upcoming essay.