I'd be willing to bet that in the next 15 years someone is going to go Billy Beane on Bain McK and BCG, pick up the talent that seems to be falling in between the cracks, apply technology to the consulting process, and make a killing.
Actually, I think those firms are pretty good at selecting the type of person they want. Sure, the 90th-percentile Penn State grad is much smarter than the 30th-percentile Harvard graduate. For technology, you'd rather have the PSU grad who can code. On the other hand, for McKinsey, the latter is arguably a "better" hire. Why? Because investment banking and management consulting are about image and the ability to sell oneself; being 15 IQ points smarter doesn't do much. Not all Ivy League students are very bright (some are, some aren't) but what can be said of almost every one of them (with their admission to the most selective universities as strong evidence) is that they're good at selling themselves.
There's something else I just thought about. The hardest part of consulting isn't the actual work-- it's the sales cycle. That's why at a certain point in the hierarchy, these firms don't promote people unless they have MBAs-- because they know the people with MBAs have been formally taught to network and are vetted.
It's the reason the partners earn 30-50%+ of the case fees ($500k/month) even though they're 1/6 to 1/10 of the FTEs.
Anyway, the pitch process is gated right now by whether the partner knows the CEO or executive at the potential client. They may know them from b-school, through a friend-of-a-friend from b-school, through another partner at the firm, etc.
Perhaps, much like AngelList has democratized deal-flow for early stage investors, something similar may crop up for corporations looking for strategy work that would let more lightweight firms bid for the work. And those firms could attract the most capable because they'd distribute the fees much more evenly.
Who knows though this is all speculation and thought out in a matter of minutes.
Given that Ivy admissions these days is based almost entirely on high-school GPA and standardized test scores, I wouldn't wager that the PSU grad is 15 IQ points smarter than the Harvard grad.
Standardized tests, depending on the test, often test for conformance to expected grading criteria and aren't really any better at correlating with IQ.
In fact, it's often shown that in many cases, high IQ is negatively correlated with GPA (the smartest kids by IQ often get lower grades because they spend an awful lot of time being bored out of their skulls in school).
I'd be perfectly willing to take the bet that a randomly selected 90th percentile PSU grad is smarter on an IQ test than a randomly selected 30th percentile Harvard grad (by at least 15 points).
The difference between PSU and Harvard students is SAT scores, not really GPA. Penn state's 25-75 range is 1750-1990 (77-93 %-ile), while Harvard's is 2070-2350 (96-99+ %-ile). The SAT is about as well correlated with IQ tests as IQ tests are with each other. Statistically speaking, the 25 %-ile of Harvard's class in terms of IQ is probably right around the 80th %-ile at PSU+. Do you think a 90th %-ile PSU student has a full standard deviation higher IQ than a 80th %-ile PSU student?
Note I'm not using the word "smart" because as Malcom Gladwell points out there are diminishing returns beyond top 10% IQ's. However, you're the one who made the claim based on IQ.
+ It's probably more pronounced than that because of the way admissions is handled individual schools do not have bell-curve distributions of SAT scores. There are sharp drop-offs below the 25th and above the 75th because schools don't have to report this numbers to USNWR.
I think you are predicating your bet on Harvard relying on the SAT as a principle acceptance criterion -- which I've read several times only makes up about 30% of the acceptance scoring criteria.
Top schools often put as much more weight on extra-curriculars and other social conformance factors when selecting students, this is important in "certifying" graduates with that school's brand name.
i.e. "We certify that this student will do any possible nonsense imaginable to jockey their way into a noticeable and expected position."
For example, kids who've spent their High School in the Key Club, Model U.N., running for an student office position and had a spot in at least one sports team (benched or not), plus volunteered once a week at a retirement home and highway trash pickup once a month are more likely to get in then on raw GPA or SAT scores.
For some social segments, like East Asian students or Ashkenazi Jews, who've focused almost exclusively on GPA and SAT (I'm generalizing quite a bit I know), they're finding that top schools will exclude them in favor of a High School football team captains/student body president with lower scores. Tiger moms are finding that they are having to turn into soccer moms to get their kids into Harvard.
To the firms that make a point of selecting graduates from these schools, this kind of acceptance criteria ensures that they'll be able to get as much grindwork as possible out of recent grads when they hire them. They know that they can squeeze 80-100 hours a week out of these folks, churning out endless slide decks and grinding numbers on spreadsheets without actually having to put any sort of intelligence to work.
Or to put it another way. If the SAT were really all that important, and considering the SAT's positive correlation with IQ (I think it's around .8), wouldn't the schools just accept an IQ test instead?
(note: you are correct in not calling the measure of IQ "smartness", it reflects capacity not the current state of intelligence)
I'm not assuming Harvard gives determinative weight to SAT scores. But its 25-75 is what it is. At least 75% of Harvard students have above a 96th %-ile SAT. In practice it's a necessary but not sufficient qualification.
Another big difference between PSU and Harvard is that PSU has 38,000 undergraduates, and Harvard has about 6,500. I'm not sure how to compare such different population sizes. Think of it this way - if you took only the top 15-20% of Penn State, you'd have a class size about the same as Harvard.
Probably 5 to 15 points if we're comparing 90th vs. 30th percentiles, depending on how we define "percentile" (do we just use GPA, or also include major and research contributions?). Performance at college is more indicative (of talent and skill) than what college the person attended... at least when we're talking top-100 schools (this includes most state flagships).
Actually, I think those firms are pretty good at selecting the type of person they want. Sure, the 90th-percentile Penn State grad is much smarter than the 30th-percentile Harvard graduate. For technology, you'd rather have the PSU grad who can code. On the other hand, for McKinsey, the latter is arguably a "better" hire. Why? Because investment banking and management consulting are about image and the ability to sell oneself; being 15 IQ points smarter doesn't do much. Not all Ivy League students are very bright (some are, some aren't) but what can be said of almost every one of them (with their admission to the most selective universities as strong evidence) is that they're good at selling themselves.