Hacker News Reader

Comment by ragona | Hacker News Reader

This article neglects to understand the massive college-hire pipelines that top companies set up. It's not that companies aren't hiring junior developers, it's that they _only_ hire from their college recruiting pipeline for their entry level roles. It's become a common policy to not make "industry" hires at SWE-1; that's basically the college hire rung.

Tell you what, I would not want to be a brand new developer without university connections trying to get a role right now. If you start looking without top internships and a degree from a school with connections it seems like it'll be rough.

View on HN

What's the systemic issue that keeps doing wealth gatekeeping?

It crops up over and over. If X is a path to upper-middle class or above, it's almost inevitable that filters and gates are put in to place that primarily work to ensure it's available only to children of the upper-middle class or above.

There's nothing inherently within most software development that makes it out of reach of being treated like a technical skill like electricians, carpentry, or plumbing.

Personally, I suspect it's more to do with "cultural fit" issues than skill issues. College is a great filter to find people who self-select and are trained to fit into the global corporate world.

Unfortunately, the side effect of this is concentrating wealth into the hands of people whom already have significant wealth and advantages. Efforts to counteract this (like the GI Bill, and college loans) have tended to have additional expensive barriers to weed out people deemed a "bad fit". For example, requiring more advanced degrees or hard to get internships (that may even be unpaid!).

scale. big companies set up basically mechanical processes that remove or reduce variance from human decision making, not just in recruiting but also their software dev, marketing, factory assembly lines, this is how scale works. a side effect of increased process is you can no longer achieve outlier results but the average result is predictable. and predictable results (predictable revenue) is precisely what big companies want.

I think this is actually a great argument for incredibly stringent anti-trust that prevents companies from getting very big at all. Successfully scaling inevitably produces extremely undesirable outcomes for anyone not invested in the company, including customers who have to integrate the heavy level of process into their usage, employees who have to deal with all the process but don't personally reap much of its benefits, competitors who are edged out by economies of scale more significantly than other competitive edges, and reactionary regulation to harms caused by all this that tend to create both more entry barriers for competitors and more process. Scale and automated processes are a means of reducing risk for shareholders and rigging competition, and preventing them should be a priority for everyone who isn't already wealthy enough to angel-invest (And probably those who are too, ultimately, although this does not bear out in the short term)

economies of scale are why we are rich, imagine if you had to grow your own vegetables

Most of the vegetables I eat are grown in my backyard. As for "why we are rich", who is the "we" in that sentence? You? Me? Some aggregate across millions - maybe even billions - of people that's meaningless to most of them?

Dumb anecdote not replicated in the real world.

It's entirely possible that many people satisfy the claim "most of the vegetables I eat are grown in my backyard".

But we can safely read advael's claim as stating "I don't eat very many vegetables", not "I produce enough food to feed myself".

I'm lucky to have a decently large yard in a decently arable part of the world. The garden in my home produces enough fresh vegetables for my (moderate) needs, but it certainly doesn't produce all the food I eat.

You're right, since you grow your own vegetables the entire global economy doesn't make any sense and is stupid.

You chose the example, not me. Also, my contention isn't that the global economy doesn't make sense, of course it makes sense, otherwise it wouldn't be there. What I'm saying is that prioritizing economies of scale is good for the bottom line of companies that do it and generally bad for everyone else on long timescales

I didn't choose any example.

If you just want to tax success highly enough that the end result is the same as failure why not be explicit about it?

I’m sorry, did the parent edit their comment? I don’t see how you get “taxing success” out of this.

They want to punish people and companies for getting good at their jobs to such an extent that there are no rewards for doing something well. The argument is fully analogous to economic growth being bad and by reversal poverty being good.

> a great argument for incredibly stringent anti-trust that prevents companies from getting very big at all. Successfully scaling inevitably produces extremely undesirable outcomes for anyone not invested in the company

> Scale and automated processes are a means of reducing risk for shareholders and rigging competition, and preventing them should be a priority for everyone

That's not a fair reading. You're conflating success with scale, something the parent post explains has a fractal of unwanted externalities.

The only way we have ever discovered of widely making more of anything is scaling processes and organisations that work. If you punish scale you punish success. This is true for manufacturing, agriculture and education and those feed into everything.

Taxing success is part of what got us here. We let these rogue optimizers get really big because of this dumb theory that everyone can somehow automagically benefit from these entities that accumulate massive power and influence. What actually happens is what always happens in such arrangements: When the entity gets powerful enough to ignore or alter the deal, they will, inevitably, whether that deal is taxes, checks and balances, or being subject to any sort of market forces, like there being competition in your market or consumers caring about (and being able to make choices about) the quality of good or service you are selling them. Risk-aversion means that such entities will try to insulate themselves from threats to their power by whatever means they have available, and we have seen time and time again that this includes subverting all the high-minded idealism of both democracy and free market economics.

Taxing "success" doesn't make it failure even with top bracket paying 90% of it back. The hyperbole and melodrama of economic winners complaining about taxes tends to make me unsympathetic. Meanwhile for the median earner, oftentimes admirable work is effectively taxed into stagnation or failure, not by the government per se, but by the upwards money funnels controlled by rent seekers further up the hierarchy, the negative lottery of savings-draining misfortunes, useless mandatory insurance, taxes they actually pay because they don't have access to as many accounting tricks, employers that collude to pay them less than they're worth, and so on.

Gatekeeping effectively boils down to using a heuristic to make a decision (good or bad).

Once a heuristic is known, people who want $thing will optimize for $arbitrary_heuristic. Those with more resources will use them to optimize better. Toss in natural human network effects (eg, giving a friend advice), and those with resources will tend to dominate.

You need active countervailing forces to avoid that natural tendency.

The systemic issue in my experience is the extraordinary flood of poor-to-mediocre candidates you get for entry-level roles...

I personally really hate college-gating admissions and would rather legally forbid that kind of discrimination (as far as, not allowing college to be on a resume). But it's impossible to deny that it helps solve a real filtering problem tech companies face.

It may have at one point, but Goodhart's Law is always looming in the background, waiting to fuck with any filter you care to come up with. In the last few generations, the shift to college being viewed (and subsidized) as an effectively-required vocational training program has diminished the quality of education in degree programs expected to turn out workers while pushing a lot of people through who would not otherwise have been interested in higher education. It's no longer a very effective filter for quality applicants

Could the solution to this be some type of exam/testing process similar to a bar exam that lawyers are obliged to pass before becoming a lawyer?

By this I mean an applicant funded testing (not educating) process operated by a respected institution that can be used as a heuristic for competency.

That was TripleByte's whole idea. They recently pivoted though :(

I recommend a book that discusses this topic in detail: Pedigree.

Basically discusses the mechanisms of how connections and wealth affect hiring.

> If X is a path to upper-middle class or above, it's almost inevitable that filters and gates are put in to place that primarily work to ensure it's available only to children of the upper-middle class or above.

Once X is widely known as a path to wealth, entry level positions get flooded with applications (e.g. a junior posting at my company could easily receive 1000 resumes) beyond what they are capable of judging on an individual basis. Once companies hit this point, they start looking for heuristics to cut down on the amount of searching they need to do. "Gatekeeping" is these heuristics.

Consider that for most top law firms, if you're not coming out of the top 3 law schools they're not interested. Why? Because those schools do a good job of filtering out bad candidates, and offer up enough good ones to fully meet the hiring needs of the company, figuring any generational geniuses they miss will make a name for themselves some other way and can be hired later.

So why does the gatekeeping all seem to help upper-middle class people? As a former educator I'm going to say it's because when it comes to the skills companies are looking for on average children of upper-middle class people are better if for no other reason than they've generally had much more money spent on their education and training, to the tune of over a million dollars compared to lower class peers.

I used to get kinda miffed that super popular restaurants like Border Cafe in Harvard Square do not take reservations, so you basically have to wait an hour or two to get a spot. But from their point of view it makes perfect sense. Why would they go through the trouble if they fill their tables at 100% with a line around the block to boot? That's how I think of the hiring pipelines as well.

p.s. Yes this is out of date somewhat but I promise Border Cafe was a hopping place 20 years ago :-)

Reservations are worse. You have cancellations, and fragmentation if any party leaves early. This can more than make the difference between a restaurant being profitable or not.

Although I’m not sure this can be extended by analogy to only hiring from “top” schools. Beyond, they don’t don’t need to do something so they don’t.

And Casa Mexico 20 years before then :-)

Pu pu hotpot, anyone?

My place!!

You heard it shut down right?

>As a former educator I'm going to say it's because when it comes to the skills companies are looking for on average children of upper-middle class people are better if for no other reason than they've generally had much more money spent on their education and training, to the tune of over a million dollars compared to lower class peers.

This is funny to me, because in most debates about the function and efficacy of public schools, I'm assured by the other side that money isn't the issue.

It's not the state money. It's all the private money the parents dump on top. The tutoring, the extracurricular, the sports, the summer camp, etc, etc, etc, all the shit that teaches them the norms of white collar society so they can fit in enough to do ok in college and beyond.

They're partly right. The US has some aggressively stupid educational policy such as teaching reading via 3 cueing (which I won't go into here), but they've also fallen for a bit of accounting trick.

Most obviously, schools being funded via property tax in most states means rich districts tend to receive more money than schools in poorer districts. Less obvious is that because the US doesn't offer many services generally those that are offered are often routed through the schools and attributed to the education budget e.g. many kinds of therapists and aides are part of school budgets in the US but are part budget under various health department budgets in other countries and thus those countries' total spend on students is undercounted.

Similarly, countries that invest in making sure there are enough social services to ensure a good life for even the bottom 20% of society need to spend a lot less fixing the problems that arise from growing up with a bad home life and get more bang for their buck.

>So why does the gatekeeping all seem to help upper-middle class people? As a former educator I'm going to say it's because when it comes to the skills companies are looking for on average children of upper-middle class people are better if for no other reason than they've generally had much more money spent on their education and training, to the tune of over a million dollars compared to lower class peers.

the extra education didn't go into developing useful skills, but into teaching them how to navigate the UMC status signaling games. that is where the entire premium goes. if they are smarter it is bc of direct filtering by SAT at the admissions stage and indirect genetic confounds, not because elite institutions actually teach useful things better.

> the extra education didn't go into developing useful skills, but into teaching them how to navigate the UMC status signaling games. that is where the entire premium goes

As someone who taught kids like this I can't agree. I taught hundreds of kids real math, writing, and programming skills that I now use in my professional life as a highly paid software engineer. Some of those kids have reconnected with me as adults to tell me I made a difference in their lives and credit my teaching (which their parents paid thousands for) as a very meaningful part of their success.

The problem is that the UMC status signalling game is heavily correlated with success in the real world.

Especially in law, where the status signalling game is a prerequisite in a very serious sense for doing the job, attracting wealthy clients.

If you hire someone from a working class background you basically need to teach them that on the job. Even in engineering it happens.

status signalling games are correlated with success in the real world almost by definition, but that does not mean they are useful to society or that we should tolerate ever greater shares of resources ploughed into them. it is a zero sum game!

What are the KPI's of the people doing the recruiting strategy for these organizations?

They're incentivized to hire people who will fit in, and perform well. After a certain level of competency, from any school, this is no longer a distinguishing feature. So they double down on the culture fit, and the easiest way to eliminate risk is to hire from a pool of people who have already demonstrated that culture.

I has a very illuminating chat once with a hiring consultant developing the onboarding/induction process and activity package for graduate engineers starting at a global oil&gas company (different industry, but the same level of gatekeeping).

When taking the grads through these activities, she was struck by the fact they all looked and sounded nearly identical(even with obligatory racial diversity). She later found out that the two managers in charge of graduate program were hiring exclusively from the two engineering schools they themselves had attended, and that they were running into a terrible time retaining graduates after the program had finished, as they usually stepped off into more boutique consultancies or went to work on passion projects as money was never the motivator.

The recommendation from her consultancy was to broaden their hiring practices, and that drawing exclusively upper crust engineers who just saw the role as a stepping stone was going to hurt them in the long run, whereas they'd get much better loyalty from middle class graduates.

>Consider that for most top law firms, if you're not coming out of the top 3 law schools they're not interested.

Yes, and this is certainly not a new phenomenon. I used to work with a recruiting department at a large upper-end law firm. They only took new associates from the top 10 universities. This was 35 years ago. So there's nothing new under the sun.

I don’t see an “upper-middle class” filter in this particular case though. The company college hiring pipelines are often not limited to Ivy League or “top 10” universities. For example my company was looking fOr summer interns, and any college would work. We’ve had an intern on our team from college which was ranked around #80... gave him full time offer btw.

You've found a feature, not a bug. In my view, wealth and power concentration is the point of nations and empires, not the side effect. Keeping all that power within a few like minded hands is how the upper classes keep their heads attached to their shoulders. From there it's turtles all the way down.

> What's the systemic issue that keeps doing wealth gatekeeping?

UMC wants these jobs reserved for its children.

Btw, what's wrong with that? If the job itself doesn't require any particular skill or competence but is very rewarding, why should they not reserve it for themselves?

There still exists jobs that can't be filled through these networks and they are open to anyone.

FWIW, I’ve worked with some very talented college students from lesser-known universities who were able to land FAANG jobs at graduation. They weren’t directly recruited, but their applications were still taken seriously, at least at the time (few years back)

A parallel problem I’m seeing is that much of the internet advice for juniors is written as if the world revolves around FAANG companies. A lot of the local college students I’ve mentored slate convinced that a reasonable starting compensation is $200K for 1st year CS graduates at every company and that they’re going to get it by playing a game where they interview at 5 different companies simultaneously and make them bid against each other. It’s a shock when they enter the job market and realize companies aren’t in a rush to pay them Google-level salaries for basic webdev or app development work.

It’s bad enough that when I start talking to new mentees I can almost guess which internet communities and even blog posts they’ve been using to gather their view of CS jobs.

> FWIW, I’ve worked with some very talented college students from lesser-known universities who were able to land FAANG jobs at graduation. They weren’t directly recruited, but their applications were still taken seriously, at least at the time

Can confirm. I come from an average state school and landed a FAANG internship for next summer. I always try to get my foot in the door early so there’s still undeniably plenty of spots left because if it comes down to two equally qualified candidates, they’re probably picking the Stanford student over the other student.

> A parallel problem I’m seeing is that much of the internet advice for juniors is written as if the world revolves around FAANG companies.

I see this sentiment on Reddit so much that at times I wonder how much of "everyone is rejecting juniors" can be attributed to juniors applying to only FAANG's and getting rejected.

Regarding the unreasonable salary expectations, I think that's just case of students seeing insane salaries in the Bay Area and thinking that it's reflective of the entire industry rather than a response to the insane cost of living in the bay area. A new-grad making $200k per year sounds insane on paper, until you realize that a studio apartment near the office is going to run you $4k/month

> A new-grad making $200k per year sounds insane on paper, until you realize that a studio apartment near the office is going to run you $4k/month

The Bay Area is expensive, but studio apartments do not cost $4,000/month.

https://sfbay.craigslist.org/pen/apa/d/mountain-view-downtow...

https://sfbay.craigslist.org/pen/apa/d/mountain-view-studio-...

https://sfbay.craigslist.org/pen/apa/d/mountain-view-top-flo...

https://sfbay.craigslist.org/pen/apa/d/menlo-park-almost-700...

https://sfbay.craigslist.org/pen/apa/d/menlo-park-charming-s...

$4,000/month can easily get you 1-3 bedrooms depending on how "upscale" you want to be or not.

https://sfbay.craigslist.org/sby/apa/d/mountain-view-beautif...

https://sfbay.craigslist.org/sby/apa/d/mountain-view-weeks-f...

https://sfbay.craigslist.org/pen/apa/d/menlo-park-furnished-...

https://sfbay.craigslist.org/sby/apa/d/mountain-view-apt-for...

Hi, author here, that's a very fair point that I didn't consider. It totally makes sense that the number of available jobs looks smaller than it is, because some of them are never really advertised in public. Thanks for sharing that!

That does suggest that the doom-and-gloom predictions are wrong, but I'd hesitate to call those jobs "available" when they only exist for a tiny slice of the job-seekers.

It's even more nuanced than you describe. Even within the hiring funnel, which school you go to often determines where you end up in the company. I will take my old employer (Intel) as an example. MIT, Stanford, etc. had a monopoly on "integration" engineers, which were the high visibility/high impact roles that coordinated all of the wafer fabrication steps among the engineers in process, yield, metrology, etc. The rank-and-file process, data analysis, etc. engineers were pulled from the next tier of schools. Finally, metrology, reliability, and QA came from all over, usually 2nd or 3rd rung state schools.

It’s absolutely true, and interviews at most companies are set up like “did you take an algorithms course”, with questions like “develop a path finding algorithm (a*)” for web frontend work.

I am self taught and got “lucky” finding a small company paying peanuts that had a lot for me to do. I interviewed multiple people with several yoe and degrees and found they couldn’t do very simple things that anyone who’d built any frontend could have done. That was over 8 years ago now and I’ve moved on to much bigger companies and much more responsibility.

But I tell my friends nowadays you have to get very lucky to find any company that will hire you based on ability to code rather than ability to pass a theory heavy interview.

The credentials are a much bigger barrier than the interview format. You can easily self-teach the algorithms / data structures theory.

Not that easily, especially when there are things that are rarely used in actual development that CS courses force you to memorize - it’s hard to know what is important and what is not. Code interview prep books being 400+ pages kind of show this.

Yes, exactly. One of the things we had going for us as an industry is that our interviews might be BS but at least anyone had a shot. You didn’t have to go to Harvard. Now that’s ending.

I can't believe that the biggest (and supposedly "best") tech companies recruit from only a small selection of schools, and completely ignore 99% of applications from other schools.

I used to think it was a total waste of talent, but the more experience I get in industry, the more I realize that it's probably for the best that so many of us are forgotten by the big companies. It gives me hope that we can escape the Google/Facebook duopoly that we live in now.

Most of the well funded fast growth startups that could meaningfully challenge big tech are actually even more picky about pedigree for entry level hires.

Perhaps that's why they fail? They are trying to play by the same rules, on the same field, with the same types of people. A well-funded, fast-growth startup that hired from a pool of people who don't think and believe the same things as the big-tech hires might do better at addressing the problems that plague big tech.

Part of it is just incredible laziness. One year my team was late in selecting intern resumes to interview for joining our team the next summer, and getting later because prospects who were contacted had already accepted internships elsewhere. When I was asked to help out selecting resumes to contact, they were all stored in a google drive folder with subfolders for each school. I asked for a list of who had already been contacted so I don't submit duplicates, and learned that the former guy doing it had just gone into the UC Berkeley folder and was picking files from there. One, no wonder they were taken, but two, come on... I downloaded everything into a flat local folder, and reviewed resumes by a mix of random selection with shuf and pdfgrep'ing some keywords I thought might be unique.

> tech companies recruit from only a small selection of schools

I don't really think it's true. At least a few years ago, I saw statistics that one of the highest represented universities at Apple was San Jose State — simply due to being large and local.

I recently had the opportunity to interview 3 fresh college graduates for Apple jobs, so out of curiosity, I looked up their colleges to see what kind of filters the recruiters are applying (I thought all 3 were well qualified). 2 of the 3 went to universities ranked between 10-20 in Computer Science in the US, the 3rd went to a college that is not ranked in any discipline in the US News rankings.

So I highly doubt that the filtering is as narrow as is claimed. Unpredictable and not particularly accurate, I'll easily believe.

I’ve found apple to be less prestige oriented in their hiring.

I made this mistake. I did undergrad research instead of internships in school and was sure I wanted to do a PhD -- until my final semester.

I ultimately wound up having to do a much worse internship than I would've gotten partnered with my school. It took me nearly three months to find it. Big mistake that I recommend anyone in University reading this right now to avoid. Diversify your experiences!

I didn’t know this was the case, took a vacation/sabbatical for a year and suddenly couldn’t apply to FAANG anymore because I missed the “graduation cycle”.

I feel so stupid, if I had known then I would have never taken a sabbatical.

In the years that followed I went almost into a depression because of it. Because I worked so hard (obtaining multiple masters) to try to get a shot and never got one. In retrospect, I feel my approach to FAANG was a much riskier endeavor than I realized.

And then it got complicated. People on HN offered to refer me. It never fully worked out. Here is why: * Sometimes I screwed it up by not strongly chasing after it

* I wasn’t a recent enough grad anymore, I needed to interview for E4

* One time an acquaintance of mine offered to give a referral and then ultimately didn’t but before he didn’t, I declined a referral from a HN’er to the same company because of it

I am working at an okay company, but I feel stuck. I want to work with people that want to treat the company they work for as if it is their own. I want fellow colleagues to be entrepreneurial. I am missing that.

You should really talk to someone about this, because you attitude of “I missed out on an internship, so my life is ruined” isn’t healthy.

People at FAANGs don’t “own the company by any means any more.

And the fact that you have a job (and probably a good job in software) shows how privileged US developers can be.

Having an office job, in a growth industry, in an in-demand field, in a country with lots of opportunities is a huge privilege that most don’t have have.

Sure you don’t work at an ad-tech FAANG company earning obscene rates, but it’s not the end of the world.

I am Dutch, average wage is 40K to 50K in euro’s. Definitely not the end of the world but the US has a lot more growth opportunities.

> I want fellow colleagues to be entrepreneurial

How entrepreneurial is someone who will work only in companies created 20+ years ago, enjoying the comforts of a monopolistic market position created off of the work of other engineers? Some people go work at a big company, some others go work a company big.

Fair point.

You should interview at a few startups to see what they are like. Good chance to meet with people and ask questions, in my experience, they will have a much bigger emphasis on culture/Q&A interviews

> It's not that companies aren't hiring junior developers, it's that they _only_ hire from their college recruiting pipeline for their entry level roles.

As someone who graduated shortly after the dot-com bubble burst, I'm extremely bitter about this behavior. I basically got screwed out of that pipeline by economic circumstance, then was far enough removed to no longer qualify for it once things recovered.

(And that pipeline isn't just the hiring of new grads, but also the system of internship and co-op opportunities that lead to such hiring.)

Yep, this is the “credential” that you buy when you go to college. You also buy knowledge but the credential is nontrivial.

Boot camps seem to be another path for non-credentialed people to get on the first rung of the ladder. They typically do a similar kind of “hiring funnel” as a college would with big companies.

Somewhere, from across the bay, blows a wind, and on it wafts the faintest whisper, an echo of an echo, which seems to say, if you listen closely, closely,

Institutional racism~

...Jokes aside - and not to make anyone feel guilty, or to be divisive, but to point out a natural consequence of the above described dynamic - I went to an HBCU. Companies did not visit campus with the frequency that I came to understand they did PWIs of similar cultural or academic standing. I imagine that if companies don't visit your school, you're not on the shortlist for, er, "meritocratic" recruitment preferences.

This especially sucks because the CS Academia track is much less diverse than outside of it: Coding bootcamps, self-taught, etc. have all the diverse candidates companies claim they're looking for and I can guarantee those folks are hungrier and want it more than most college grads.

Yeah but to be a good programmer, it helps a lot to have a real CS education. "Hunger" and "wanting it more" is no substitute for understanding CS fundamentals

Realities and realities. In most EU countries, companies would hire even garbage men if they prove they can code a little bit, because of such high demand for programmers.

Most companies I know like to hire junior devs and form them to their company culture rather than to deal with the "baggage" of senior devs with 15 - 20 years of experience.

More and more senior devs complain that they are essentially "unhireable" because of their experience and age.

For new grads in recent times, TCS and the like can be a great career starter if you can settle for the job not being glamorous, having to move where they put you, and high variability in career development depending on what you get put on (possibly paid to do nothing at all for stretches of time, which are on you to do something productive with or not). Good beginning pay for a fresh grad and high job security. Not even sure how relevant having to move is anymore with so much more being remote.

I basically did that out of a small university with no internship experience or strong university/personal connections, in 2013.

It was definitely a struggle at start and I've had to struggle through some really crappy roles, but I'm doing pretty solid now. Making almost double now than I was when I started out, I'm a team lead now and very happy with the company I am with now.

But coming from a small uni in Normandy, France, it's also possible not to go to a "top" internship, shit that up and be let go with a kick in the ass.

You can go to a small do-it-all no-money startup for a few years, suffer your ass through it, climb up a few more job and end up exactly at the same place with a very good story and a lot of experience from doing more with less that top companies can appreciate, and actually be the one bossing the silver spoons into work.

It's a fallacy to think people are disadvantaged by not going straight to the top with no real world experience.

How does hiring immigrants fit into this pipeline? Every team I've worked was majority immigrant (myself included ;), from entry level all the way to a few layers up the management chain. Most didn't go to any US schools.

This is something I’ve been experiencing in my quest to switch from IT to SWE.

Entry level jobs are for new grads, and absolutely nobody cares about personal projects or professional experience that doesn’t have “software engineer” in the job title.

Reply on news.ycombinator.com