Outsource screening to colleges

This is an interesting take. From How the Supreme Court Created the Student Loan Bubble by Bill McMorris.

That’s because none has addressed why young Americans are forced to go to college in the first place. Their money and brainpower would be better spent overturning Griggs v. Duke Power Company.

The 1971 Supreme Court decision remains largely unknown, but no ruling of the past forty-five years (except for Roe v. Wade) has done more harm to the American way of life. It changed the way companies hire, pay, and promote workers, ensuring that America would be a country defined by credentials rather than merit. Griggs is why we’re wasting money and time on a dubious good like a B.S. degree—pun intended.

[snip]

The lie that props up our Big Education regime is that the GI Bill, which paid for World War II veterans to attend college, produced the upward mobility and economic boom of the postwar period. It’s a heartwarming story, the veteran who would have been a dust farmer but for the grace of government generosity. But it just isn’t true. Only one out of every eight returning veterans attended college. The rest, the vast majority, benefited from something even more egalitarian: aptitude testing. The format favors raw talent above all else, allowing companies to hire high-potential candidates from any background and groom them to fit the company’s needs.

These tactics came to commerce from a familiar source.The armed services were forced to process hundreds of thousands of recruits during the war, and in order to filter and assign soldiers, the government developed aptitude tests. Businesses witnessed the U.S. defeat the two most efficient peoples known to man, thought there must be something to this whole testing thing, and followed suit. The chief hiring metric in the postwar era was not whether someone had a degree, but whether he had the aptitude that would enable him to succeed. Every industry from blue-blooded high finance to immigrant-heavy manufacturing employed testing to determine who would rise through the ranks, regardless of lineage, heritage, or education. Testing enabled men who set out to be blue-collar workers to ascend based solely on their ability.

Above all, tests are effective.

“Despite their imperfections, tests and criteria such as those at issue in Griggs (which are heavily…dependent on cognitive ability) remain the best predictors of performance for jobs at all levels of complexity,” University of Pennsylvania Professor Amy Wax has found.

[snip]

The Griggs decision has made that organic rise through the ranks impossible, as disparate impact left businesses liable for those who failed to pass hiring tests.

“Most legitimate job selection practices, including those that predict productivity better than alternatives, will routinely trigger liability under the current rule,” Wax wrote in a 2011 paper titled “Disparate Impact Realism.”

The solution for businesses post-Griggs was obvious: outsource screening to colleges, which are allowed to weed out poor candidates based on test scores. The bachelor’s degree, previously reserved for academics, doctors, and lawyers, became the de facto credential required for any white-collar job.

The article goes on in to some far reaches but I think that core argument, the Griggs undermined the ability of companies to select based on merit rather than credential, probably has some merit. It wouldn't be the first time a good intention resulted in a bad outcome.

I think what McMorris is really saying is that the root issue is that the government essentially, but in round about ways, outlaws any qualification tests which result in disparate impacts. Since all such tests always demonstrate disparate impacts, companies are no longer able to use valid tests for selection and advancement. Instead, they short circuit the process and go to those who are still sanctioned to use such tests - universities.

Its not that IQ tests are all that excellent at forecasting. Better than nothing, less than perfect. But would there be an improvement in company performance that was material enough to be noticed in macro-economic numbers were companies allowed to use targeted testing, even if it had disparate impacts? I don't know but I suspect so. In an increasingly competitive and uncertain global economy, I suspect that that lost productivity from non-testing may get much more enticing.