American founders and immigration
There's been a new round of U.S. immigration drama on Twitter, this time centered around YC. I'm currently dealing with immigration to attend the upcoming YC batch as a founder, so the online discourse feels very close to home.
Here's my perspective on all this, which I've been meaning to write before this even blew up:
The best path for founders to immigrate to the US is an O-1 visa, meant for individuals with "extraordinary ability". This is the only viable path for most, because the only other pathways into the US are through family sponsorship or employer sponsorship.
I was vaguely familiar with the O-1. What I didn't know is just how onerous the whole process is. To prove extraordinary ability, you need to assemble proof, gather recommendations, and submit hundreds of pages of paperwork, all meticulously verified for consistency. I've heard cases of this requiring 10-15 hours a week for months, on top of tens of thousands in lawyer fees.
That being considered, the fact that almost half of all unicorn founders are non-American was baffling to me. A recent YC batch even had 50% of its founders be non-American.
Why is this? The venture capital business is tremendously meritocratic. There are billions on the table, and investors have every incentive to maximize the chance of funding the right founder. This is unlike the job market, which many argue to not be meritocratic: Most companies hire for some baseline set of skills. It wouldn't matter much if someone is slightly better at the job, and oftentimes one who is worse may well get the job if they knew the right people1. The same is not true of venture capital investing in founders: founding teams are so integral to the success of a startup and are the difference between a multi-billion dollar ROI and a negative ROI.
If you reframe it like that, with VCs highly incentivized to invest in the best founders from a global pool, then even the fact that 50% of successful founders in the US are American is pretty good. It's even an overrepresentation, considering that America only makes up 4% of the world population. Accounting for the fact that not everyone wants to build in the US, and that Americans inherently have an advantage in running a US company, the 50% number seems more reasonable.
##Footnotes
-
Separately, I have come to realize that a lot of what people call "nepotism" is at times about circles of trust. If the founding team of a company is all from Waterloo, and the founders have to decide who to hire as their first couple employees, they will probably be from Waterloo. This is not because they are filtering for people from Waterloo; rather they are seeking out the best people they know and can personally trust for a role that is so integral to their startup. It would just happen to be that a large percentage of who they know are from Waterloo. Then when those hires end up building their own teams within that company, so too will they have an inherent bias. This "bias" is introduced without ever explicitly filtering, the bias is introduced implicitly from the structure of networks of people. When this is applied to universities from different countries, the same result can now visually be seen. From the outside this looks like a racial bias. I'm sure there also exist many cases of true racial bias. But from my (limited) experience at high growth startups and unicorns in the Bay Area seeking top X% talent, it is often not. Opinions are formed through personal experience, and I suspect a lot of people have had different experiences leading to different opinions. ↩