> if tech is a meritocacy then why are most jobs made through social connections?
Two reasons:
1) There is no such thing as a true meritocracy (we can generalize this to the claim that there is no such thing as absolute efficiency, which is what a true meritocracy would ostensibly be a manifestation of in employment).
2) Human beings have limited attention and productivity bandwidth which typically (though not always) becomes less than the sum of its parts in groups.
The second point is what's really important here (and it has a causal relationship with the first observation). If you have 100 candidates applying for a position, you literally do not have time to adequately review each and every candidate in a reasonable amount of time while maintaining productivity in your professional capacity.
Theoretically we shouldn't use referrals. But in practice it's a good heuristic because it models a web of trust: if person A trusts person B's abilities, they might also trust person B's ability to judge person C's abilities. Obviously this can't be formalized well enough to be extremely rigorous, which is why it doesn't always work.
It's probably not even the most optimal solution! You could develop sophisticated organizational processes to scale up the candidate review process without leveraging trust in other parties as much. But people default to it because it works often enough and because it's intuitively an attractive idea. That combination tends to be viral for human interaction, even if it's technically suboptimal.
>if tech is a meritocacy then why are most jobs made through social connections?
How do you judge someone in a meritocracy?
Interviews? Not reflective of actual work. Also trivial to bias in favor of one group due to culture fit elements.
Short term projects? Exclude people who have limited free time (generally the more senior/older ones).
Resume? Trivial to lie on.
References? How do you know the reference is competent?
Social connections are merely another, imperfect, attempt to judge someone's merit. Specifically you trust the judgement of someone whose merit you have already proven in an actual job environment.
Because the best way to know if someone has ability in engineering is if you've worked directly with them before. Don't confuse friends and relatives with someone you've worked with before in a professional setting.
Not really really. It happens in other industries as well. It's not corruption. Corruption would be hiring someone without due diligence interviewing just because you know them.
I can't think of one company which refuses referrals. Rather, even the largest of public companies have referral programs.
Two reasons:
1) There is no such thing as a true meritocracy (we can generalize this to the claim that there is no such thing as absolute efficiency, which is what a true meritocracy would ostensibly be a manifestation of in employment).
2) Human beings have limited attention and productivity bandwidth which typically (though not always) becomes less than the sum of its parts in groups.
The second point is what's really important here (and it has a causal relationship with the first observation). If you have 100 candidates applying for a position, you literally do not have time to adequately review each and every candidate in a reasonable amount of time while maintaining productivity in your professional capacity.
Theoretically we shouldn't use referrals. But in practice it's a good heuristic because it models a web of trust: if person A trusts person B's abilities, they might also trust person B's ability to judge person C's abilities. Obviously this can't be formalized well enough to be extremely rigorous, which is why it doesn't always work.
It's probably not even the most optimal solution! You could develop sophisticated organizational processes to scale up the candidate review process without leveraging trust in other parties as much. But people default to it because it works often enough and because it's intuitively an attractive idea. That combination tends to be viral for human interaction, even if it's technically suboptimal.