Because if that anecdote is true, Morgenbesser is clearly insane. Drill down on your intuition about why he's insane and you arrive at "he's violating IIA".
> You only take a job in business operations if you aren’t smart enough to be an engineer (or designer, or product manager, or…)
The article claims that this is false. But in fact, it's pretty obvious that the average non-engineer is less intelligent than the average engineer. See for instance this table of IQ by college major: http://www.statisticbrain.com/iq-estimates-by-intended-colle...
(And, yes, IQ tests do quite accurately capture what is meant by "intelligence".)
You seem to be implying that we should focus on reducing demand. That could be true, but "X implies Y" does not necessarily imply that the best way of eliminating Y is to eliminate X. For instance, "We don't have a murder problem, we have a "people exist" problem."
So politicians are in a conspiracy to not stop problems? Sounds unlikely to me, I think it's more like:
* Many politicians, like most people, are idiots. Partly because many voters are also idiots.
* Politicians don't have strong incentives to stop problems. (But they don't have strong incentives to preserve problems, either. They're only incentivized to be popular.)
* Principal-agent problems. Even if a politician comes up with a brilliant solution, the people lower in the chain won't execute it well, because they often don't have strong incentives to.
Anyway, there was a conspiracy to preserve problems, any particular politician could start solving problems. This would likely place him above his peers. Then his peers might backstab him, but they also might join him, since he'll have increased popularity. Essentially, we have a sort of iterated n-player prisoner's dilemma; or rather, at each step we choose m people (non-randomly) to together participate in an m-player prisoner's dilemma. It might be interesting to try to model this mathematically.
The article uses misleading numbers. Sure, that 2% bonus is worth $60k if you compound it over 40 years. But that number is almost irrelevant. What's relevant is the opportunity cost. If you spent that time (likely hundred or thousands of hours) working on more valuable skills, you'd likely get much more than 2%. Put another way, the value of that time, when compounded over 20 years, is enormous, probably more enormous than $60k. To compound the value gained from language but not compound the value of the time is extremely misleading - either use present values for both, or use comparable future values for both.