> I know if programmers making 100k find a new job in a week, and factory workers making 30k take 2 years to find a job. Yes, unrepresentative outliers exist.

> I know if programmers making 100k find a new job in a week, and factory workers making 30k take 2 years to find a job. Yes, unrepresentative outliers exist. I don't see your point. This is a frequently cited rule of thumb by professional recruiters and those who work as specialists to help candidates improve their job search and find a job. It's been around for a while and applied to many industries, though certainly some are different. I'm not claiming it always holds, or even that things will always be the way they are now, just that it's a good rule on average for most jobs, and it's definitely a good rule of thumb for software jobs, even in the middle of SV. The reason it takes a long time is that you're assessing goodness of fit. The case when you're just looking to immediately jump ship and you have essentially no standards besides compensation (this happens a lot in SV) is not being considered and is irrelevant in a discussion about severance anyway. You're also speaking from the perspective of someone looking solely in a confined geographic region -- meaning you're not considering the vast amount of job searching that involves considering relocation, in which case just up and leaving your old life behind on a whim after finding a job in one week's time is a ludicrous outlier. What about people with families? You're thinking about this from only a single, narrow perspective that is an outlier in terms of the total volume of jobs. The volume of jobs in SV for which people apply who have low search standards and who are in life positions to just accept possibly relocating or significantly changing their commute on a whim is a tiny fraction of the total volume of jobs, so again I just do not see the relevance of these anecdotes. > Salary is not related to search time, demand is. This is very naive. Salary is loosely a measure of how irreplaceable your labor would be, and is also an expression of demand, both of which affect search time. Salary is definitely causally related to demand, so it's at least a proxy. Again, no one is claiming there is some hard theory to propose an economic law relating search time to salary. It's just a heuristic, most likely because salary is a proxy for demand -- one of the most ubiquitous ones. You seem to just glibly deny this but don't even pause to consider that salary is in part a proxy for demand. The bigger thing is that after you begin earning a salary that you're happy with, you know what you're worth and you know the rough salary range you'll be seeking, and you'll spend some time interviewing at places that can't come near it, but eventually you'll figure out how to target the right places. When you do, then it's not about salary -- it's about work/life balance, flexibility, whether you'll do the kinds of projects you want to do, commute, relocation, whether you like the area you'll move to, how bad are the company politics, etc. This is a massive and difficult combinatorial optimization problem and even conditioning on finding places that offer in your desired salary range, if you take these other life factors at all seriously, it takes a long time to find an acceptable fit. Given that increasing salary is generally related to increasing seniority (thus age, and thus also more complex life situations with spouse and children or family elders in need of care, etc.) and to increasing responsibility (thus politics and subtle factors will greatly affect your experience in the new job, much more than compared with lower salaried workers in general) it's very reasonable to propose that as salary increases, search complexity increases, and so search time to find an acceptable role also increases. Again, outliers exist. Deluded folks who just want to maintain a certain kind of status lifestyle in an urban center may just jump to any job that sustains that lifestyle -- that's a super uncommon way to search for a job in the market overall, even if it's not uncommon in a market like SV. It's just not useful for general heuristics. I mean, an even more obvious outlier example would be getting hired due to a professional connection. You might spend zero time looking for a job and have it just offered to you. But we don't go around making general heuristics based on extreme outlier experiences like that -- and people living in those outlier experiences know that they don't need to use the heuristic.