> The problem with your theory is that the disparate impact criteria in Griggs creates exactly the same burden for college degree requirements as for IQ tests.

> The problem with your theory is that the disparate impact criteria in Griggs creates exactly the same burden for college degree requirements as for IQ tests. (And, secondarily, that that burden for IQ tests is low enough that if they were as suited for purpose as your argument suggests, it would be routinely and easily met.) A lot of business activity is guided by risk-management decisions that try to minimize the likelihood of dumb and costly lawsuits. An example is companies not wanting to write down feedback for job applicants that are rejected. We all know it's a shitty and dehumanizing practice to not give some feedback and encouragement, but companies can't be human because that creates a possible risk. My response to what you said above is that no employer would face the risk of a lawsuit for trying to hire college graduates, so this system is 100% safe and desirable for them. In fact, this college-based system works great for employers because their employees are now addled with college debt so employees might feel a bit more cautious about risking their jobs to stand up and challenge working conditions. So I guess most existing employers aren't heavily incentivized to change this system too much.