[tdees40]

A thought experiment: you have a job that pays exorbitantly well, and you want to offer it to someone. It isn't particularly difficult, and any reasonably well-informed, self-motivated college graduate can handle it. How do you hire for that job? The demand is likely insanely high.

Spoiler alert: you create an insanely complicated funnel to weed through huge quantities of candidates. And if talent isn't the primary determiner of success (remember that the job isn't that hard), how do you distinguish people? Simple, you work them nearly to death and see who comes out alive.

I'm not saying that is exactly the situation here, but I think as a mental model it's not that far off.

[Chinjut]

If the demand is so high, why not simply lower the pay to winnow the candidate pool? Why bother with the insanely complicated funnel instead?

[tdees40]

Because your pay is high too! If you can find people to work beneath you for cheaper, why can't they replace you for less money?

Edit: also, they do that with intern/entry-level banking jobs. Internships pay well, but it's nothing like what higher-level bankers make. Same with entry-level analysts.

[whatok]

The senior bankers who actually bring in clients are worth their weight in gold. 0 chance anyone in a position like that is going to stick around to get underpaid while bringing in business.

[dasil003]

Because it winnows the wrong ones. They want the extremely tenacious money-driven type, because those are the ones that will help them make billions once they work their way up—and gladly clap on the golden handcuffs as they do.

[ThrustVectoring]

If you lower the pay, the "best" candidates get grabbed by competing firms. The funnel works because you're not just selecting a subset of people, but because you're selecting a particular group - specifically, those who are willing to give the firm a better share of the organizational rents.

[georgehotelling]

Well it's already $0 in some situations, so... charging people to be interns? You just might have a future in finance.

[dkuntz2]

Not in finance they're not. Finance wants the best interns who can actually perform at least some work, and doesn't want to get busted for breaking employment law in regards to unpaid interns.

[Jdoemk2]

GS interns get paid loads

[stephengillie]

Instead of having one job pay $200,000 per year (let's say total compensation is $350,000) and require 18-24 hours a day of work, you create 2 jobs that each pay $100,000 (total compensation $175,000 each) and only require 9-12 hours per shift.

OR

You could create 3 shifts at $66,667 (total compensation $117,000 each) and have true around-the-clock coverage!

[001sky]

That's not a useful solution. Consider they are trying to get 3 candidates for 1 job promotion; having 6 doesn't help it just dilutes the ability to observe. On the other hand, you now have to manage/babysit double the number of over-qualified, highly ambitious people with too much time on their hands. Thats a recipe for trouble.

Its the same logic as why PhD programs don't simply double their size when there are plenty of people with enough credentials to pass entrace exams. Its a people-management issue, and maintaining the incentives of the rank-higherarchy. It has nothing to do with talent levels.

These positions are only loosely meritocratic. The higher-order filtering is almost always based on some form of social privledge. Its not dysfunctional, at least in a litereal sense. Privledge opens doors and changes the economics (ie, productivity) of the industry materially. That is why it is selected for very heavily.

[stephengillie]

So this is a form of "labor-theater" that signifies a job that shouldn't exist? It's just a show, a play put on, to legitimize the transfer of wealth between the privileged? You make it sound like a false-problem, a social invention where any effort to resolve it is just wasted.

[001sky]

There is a set of legitimate economic issues raised about the cognitive overhead of alternative selection models, etc. Its not like its all in somebody's head. That being said, the trivial view of "meritocratic" selection is not the right model to articulate. These jobs weed out people who don't have a certain level of technical competence, but technical competence (level) is trumped by the speed and consistency (of application) in many cases. Also, the ability to "avoid doing work" is a skill or a feature, not a bug or a liability. Again, this is not really going to line up with the notion of merit being who can do the "most work" at the "highest level". But it turns out one way to test who can "avoid doing needless work" is to force such a high workload that some work must by necessity be avoided. If that makes any sense. I dunno.

[stephengillie]

It sounds like you're describing the Bill Gates quote: "I choose a lazy person to do a hard job. Because a lazy person will find an easy way to do it."

To which others counter: "The lazy person will just not do the job." And you're describing something closer to the ideal of only doing the important part of the job, and being able to discern the important work from the unimportant.

[whatok]

So people always bring up outsourcing part of the job to some developing country but that would run into the same problem as your third option. Having too many cooks in the kitchen will introduce a lot of operational error while having a few people who know everything about a project will run things smoother even if they are worked to the bone.

[stephengillie]

Outsourcing to some remote part of some other country, and outsourcing to some remote part of the country you live in, aren't much different.

Many large companies are able to maintain a large number of server operators and perform essential functions without needing one person to do everything. With sufficient communication you can avoid the "too many cooks" problem. I've personally lead highly-communicating teams which avoid this problem. In short what you do is put one of the 3 in charge of different work as it comes up, load balance and collaborate.

But maybe it's cheaper and easier to pay one person a huge amount of money, and beat them up, than to create a highly-communicating organization of people (including email, IM, ticket system, face-to-face time, regular meetings, bridge calls, etc). It certainly looks simpler.

[whatok]

Thank you, I'm well-aware that other large companies are able to do all these wonderful things but it really is a different beast.

I think a huge part of it is that the tools do not lend to a very collaborative workflow. Excel and Powerpoint do not exactly have a robust version control system or any great way of leaving any useful comments so that alone makes things difficult.

And honestly, it's probably around the same amount of money if not more expensive to keep a leaner crew but for the type and nature of the work, it's probably more efficient.

And don't get me wrong, there is some outsourcing going on, just not on some "well if all they're doing is powerpoint and excel, why don't they outsource the whole thing" level.

[jsprogrammer]

It's probably not only easier, but the alternative requires having considerable technical expertise. GS and ilk probably identified such expertise as a massive cost center long ago and don't keep much stock in people that could actually execute such an organization.

[qq66]

The funnel is not that complicated; it's set up to find the people who are willing to make any amount of personal sacrifice to make huge amounts of money. Those are the people they want to be future Managing Directors. Startups that have insane working hours are doing the same thing.