Ignoring the Middle Fallacy: overlooking long-term trends

I don’t get this common/popular notion where just because you lose a lucrative job or opportunity that suddenly you become poor. What happens to all that accumulated money? Does it just go away? I think we want to believe that it does because it gives us comfort knowing that some sort of balance has been restored to the universe–that people who are so successful have to be reset to zero.

This observation led me to conceive the ‘ignoring the middle’ fallacy, which describes the tendency of pundits and commentators to only look at the endpoints or negative events that punctuate an otherwise long-term trend, ignoring or overlooking accumulated earnings (or in mathematical parlance, the integral) or the long-standing trend in-between the endpoints, so thus one is inclined to overestimate risk and underestimate reward. I originally described this earlier, but I want to expound on it because it’s so common.

It’s also related to negativity bias. It’s not newsworthy when things are going well. Incidents of ‘public freakouts‘ (which has become a meme in and of itself, as evidence of the travails of life in a post-Covid ‘boring dystopia’) and police misconduct go viral on social media, but hardly anyone shares clips of police doing their jobs well, or of people behaving well in public. Or at least people are more receptive to negative stories compared to positive ones, especially on social media.

For example, from Nassim Taleb regarding Tucker Carlson losing his TV show:

So in the case of Tucker it pays to do BS in the sense he earned a lot of money, which he presumably gets to keep. Tucker’s show ran for over six seasons, which represents considerable accumulated earnings. Tucker’s annual salary was purportedly $8 million, and he has a net worth of $30 million. It would only not pay if he had to forfeit his accumulated earnings.

Yes, Alex Jones was ordered to pay $1.4 billion after losing a defamation lawsuit, yet his Info Wars show is still running, and presumably he is earning a lot of money still (around $70 million/year). And there is no evidence he has paid anything out either, and bankruptcy law makes it hard to collect. (As any civil lawyer can attest, it’s easy to sue, harder to win, and even harder to collect.)

So far, the only money Mr. Jones has paid in the Sandy Hook matter has been to his lawyers and the courts, as he has run up millions of dollars in legal fees and sanctions for abusing the judicial process. Another trial for damages — potentially on top of the more than $1.4 billion Mr. Jones owes — is slated for this year.

And from Freddie deBoer regarding recent tech layoffs:

More to the point, even the mighty computer-science degree is not immune to the forces of industry contraction and fierce competition. Major technology firms have made huge job cuts in the face of rising interest rates and shrinking pools of loose cash, and the number of computer-science majors has grown substantially in the past decade. A New York Times article from December of last year laid out the inevitable consequences: an increasingly difficult job market for recent graduates.

But data going as far back as the ’70s shows engineers earning a lot more compared to other majors. A year or two of weakness does not change the overall trend. Even in 1976 engineers earned twice as much as humanities majors. [1] And it’s not like other sectors/fields are immune to downturns either, such as the 2007-2009 global financial crisis, which mainly affected MBA/finance-types.

And from the early 2000s, STEM is still way ahead:

Of course there is a lot of variance within majors, with some engineers earning less than history majors, but we’re talking averages here. Of course, we don’t need to be reminded that earning more money is not everything…there is value to careers that are less remunerative than ‘learn to code’.

Yes, the tech market has softened, but it needs to be put into perspective relative to the long-term trend, which is still strongly positive. Pundits tend to have terrible track records at predicting this sort of stuff, too. People said the same in 2012, 2016, 2020, etc. “The market has topped!” “Bubble burst!” For example, from Seeking Alpha in 2011 “We May Be Nearing A Third Tech Bubble Collapse”. AMD is singled out as a bubble stock, which given how well it has performed due to the recent boom in GPUs, it’s safe to say that prediction aged poorly.

Companies have layoffs as the business cycle crests, and then rehire on the way up. That’s how business cycles work. It happens over and over. It’s not newsworthy when tech companies hire workers, but laying off a thousand workers is suddenly the End Times.

There is no reason that the challenges of unemployment should be uniquely bad to people in the technology industry. If anything, unemployed FAANG workers are going to be in a better position, all else being equal compared to other jobs, by virtue of having more money and being more qualified overall (being qualified enough to land such a prestigious job presumably would be helpful when looking for a new job), or having better compensation/severance packages.

In invoking the fallacy again, if you’re making $200k+/year in tech, keep expenses low, and invest diligently, even if you get fired after only a few years because you happened to time the cycle poorly, that still represents a decent amount of accumulated savings, which you get to keep. It does not just go away. Your score is not reset to zero, but rather you’re temporarily out of the game.

Regarding inflated tech salaries, I expect them to remain high. Bigger tech companies and winner-take-all markets, means bigger salaries and more perks to attract top talent when the stakes are so high. The difference between getting a product out a few weeks early or late can mean losing or winning invaluable first-mover advantage. The next big arms race is AI. You can be sure big tech companies are offering huge salaries for top talent there too.

Maybe we need to stop inventing excuses or justifications for how people who are by quantifiable, objective measures better off, are not actually better off. Temporary misfortune of the otherwise successful does not change this.

[1] As far back as data exists, engineers earn well above average, and humanities majors earn below average.

1 comment

Leave a comment

Your email address will not be published. Required fields are marked *