from
It was a Tuesday in November, around 9 p.m. at the bus terminal in downtown Hamilton, Ontario. A driver had left the doors open on his bus so passengers could board while he took a short break. One of those taking refuge from the cold was a homeless man – or as the equity-oriented Canadian Broadcasting Corporation put it, a man “of no fixed address” – who promptly sat down in the driver’s seat, closed the doors, and proceeded en route with about 10 passengers aboard.
The “driver” got off to a great start, making every stop and allowing passengers to get on and off. He drove flawlessly, which wasn’t easy considering it was an extra-long articulated bus (i.e., accordion-like structure in the middle). And he was a stickler for the rules, at one point denying boarding to someone with an expired pass. No one suspected he wasn’t a real bus driver until he began making wrong turns. Rather than freaking out, the nice Canadian passengers gave him directions and got the bus back on track. According to the Hamilton Police, “there was not a ding on the bus. He did a great job.”
The story received international attention with some outlets referring to him as the “most polite bus thief.” Whether he was polite was beside the point; he had the skills to be an effective bus driver. Several Reddit commentators suggested Hamilton Street Railway (HSR) should offer him a job. Notably, the bus drivers’ union did not respond to requests to comment.
While we don’t have the man’s CV, it’s unlikely Hamilton Street Railway would ever have seriously considered his application. Which explains not only the excruciatingly slow joyride – he saw an open position and, as another Reddit participant put it, “you’ve got to be aggressive in this job market” – but also the parlous state of skills-based hiring.
***
Skills-based hiring was white hot about three years ago. It was the refreshing, democratizing contrast to pedigree- and degree-based hiring, which is all about where you went to school and what degree you earned. But the impact has been miniscule; according to Burning Glass Institute, “for all its fanfare, the increased opportunity promised by skills-based hiring was borne out in not even 1 in 700 hires last year.” And media references to skills-based hiring have trailed off. The problem was that no one knew what it meant or how to do it. Everyone agreed on removing degree requirements from job descriptions, and important progress has been made on this front. But that wouldn’t have been enough to hire our homeless Hamiltonian.
Employers have always attempted to assess the skills of candidates in job interviews. That’s the point of interviews – that and ascertaining “fit,” whatever that means. The question has been who among the hundreds (and in many cases thousands) of candidates gets interviewed. Up until a few years ago, that was the province of applicant tracking systems which attempted keyword matches between the job description and the résumé. Recently employers have begun adopting AI filters, now drawing criticism and litigation – not because AI does worse or discriminates more than keyword matching, but because the odds keep getting worse and candidates are increasingly ignored, prompting The Atlantic to ask When Did The Job Market Get So Rude? According to one plaintiff with a computer science degree and decades of tech work experience, of the thousands of jobs she has applied to, only 0.3% progressed to an interview.
It’s even worse for aspiring career launchers. Because AI filters are trained on datasets of mid-career professionals, inferring skills from titles, roles, and industry taxonomies – not from coursework or clubs – they may be wholly ineffective in screening candidates who haven’t had much, if any work experience (a growing problem given the internship gap). While AI assumes titles and taxonomies, new grads have signals rather than work history. Further, AI filters may interpret age-appropriate career exploration as signals of lack of focus or commitment.
So I wasn’t surprised to read a Wall Street Journal report that entry-level hiring has swung back to pre-skills-based-hiring norms. A recent survey found a 50%+ increase in the number of companies recruiting from a shortlist of schools. GE Appliances has winnowed its recruiting down to “15 select universities” and attends four or five events each semester at those schools. McKinsey is taking a recruiting GLP, slimming down to about 20 schools. According to one executive who led entry-level recruiting at Target, most companies now recruit at fewer than 30 colleges – usually a combination of top-ranked universities and local schools. Local schools like University of Louisville, where the head of career services acknowledged that “a lot of those [national] employers who were new to us [a few years ago] didn’t stick around.” According to the former Target executive, if you’re not graduating from a top school or a local school, “God help you.”
The same thing seems to be happening at business schools. In another Journal article from the same reporter (Lindsay Ellis on the reality check beat), 3-month employment rates for MBAs from top 10 schools like Harvard and Columbia remain high and are holding steady or increasing. Meanwhile, business schools like Georgetown are seeing a sharp decline: “a quarter of M.B.A.s were still looking for work three months after graduation, up from about 16% the year before and 8% in 2019.”
The retreat couldn’t come at a worse time. Job seekers and aspiring career launchers were already befuddled at their prospects. According to the Federal Reserve Bank of New York, the unemployment rate for the last five classes of college graduates has skyrocketed 40% in two years to 5.8%: a 30-year high and, for the first time ever, well above the national unemployment rate (4%). The Bureau of Labor Statistics is reporting an eye-watering 12% unemployment rate for college graduates in their 20s. And in a New York Times poll last week, half of respondents under 45 say they’re worse off financially than their parents were at their age. Only 13% of young adults think the country is heading in the right direction and 69% of Americans no longer believe in the American Dream.
And in a crucial moment of economic and technological transformation, it’s also bad for companies. In a recent Money Stuff column, Bloomberg’s Matt Levine discussed the increasing concentration of financial returns among big winners. Companies that increase in value 1,000x, for example. For investors, the rest are increasingly irrelevant. So investors should back moonshot businesses and CEOs should allocate resources to moonshot projects. Presumably the same logic should apply to the hiring that powers these companies. Whether or not the 10x or 100x software developer or engineer is a myth, there’s no question that technology can dramatically boost employee productivity and increase value-creation variance within a cohort of new hires. Ability to effectively leverage AI makes the variance even larger. So how would a hiring manager go about trying to find 10x or 100x employees? As they’re likely to have unique characteristics, it might be best to cast as wide a net as possible instead of continuing to recruit in the same small places.
***
How to explain the hiring regression? Campus recruiting is expensive, both in terms of direct cost and opportunity cost (employees recruiting rather than doing their jobs). Also, if 99%+ of applicants are going to be rejected anyway, the bait-and-switch brand hit from soliciting students from hundreds of colleges to apply might outweigh the benefits. And when companies are flooded with AI-generated résumés that look identical, a candidate’s college can be a distinguishing factor; focusing on key schools is a readily available filter. Then there’s the impact of AI on entry-level positions; as companies reduce the number of these jobs, there’s less need to take risks: go with what has worked in the past.
But it’s likely something more is at play. Back in 2022, diverse recruiting was a priority for 60% of employers. In those rainbow-hued days, it was important to recruit beyond a small set of schools. But the proportion of employers who still feel that way has been cut in half. With the demise of DEI, in a few short years we’ve gone from skills-based hiring nirvana to hiring revanchism. Was it all just a dream or was skills-based hiring the employment manifestation of a “woke” ideology that provoked a hiring backlash?
***
The last thing America needs is for hiring, employment, and economic mobility to swing as wildly as our pendulum politics. And that may be what’s happened. Why not hire from a small number of colleges? My department or company will do just fine. Moreover, new graduates from my alma mater will do fine because I’m going to keep recruiting from there. Taking care of #1 is all that matters. As for everyone else... It’s reminiscent of a number of new domestic policies as well as our dismaying approach to international affairs. Why bother maintaining a complex network of security alliances when we can play at might-makes-right, spheres of influence? We’ll get ours and shouldn’t care what happens to others. It’s short-term, survival-of-the-fittest thinking. And it’s dangerous. For the world, for hiring, and for economic mobility.
There has to be a middle ground between pie-in-the-sky skills-based hiring and Making Hiring Great Again. It starts with respecting both candidates and hiring managers. Applicants deserve meaningful consideration and communication; the alternative is bad for business. At the same time, while hiring managers don’t get the full benefit of finding and making 10x or 100x hires, they bear the brunt of bad hires. Without the right support, they’ll turn a blind eye to hiring upside while remaining exquisitely sensitive to the downside. So they’re naturally inclined to be conservative, recruiting from the same old same old.
The good news is that we can accommodate both. All applicants who go to the trouble of applying – even if it means only clicking a button – deserve the opportunity to demonstrate they can do the job. Companies can do this by providing all candidates with a quick job-based assessment so they can demonstrate skills as well as interest. The answer isn’t AI filters attempting to evaluate non-existent work experience of aspiring career launchers. It’s harnessing AI to design assessments that suss out the capabilities most predictive of performance in each role. Then assess results and make recommendations to managers on who should be interviewed. Better yet, shift recruiting to a talent marketplace like FutureFit AI where candidates are shown all available roles in the organization and provided with clear pathways to qualification, including assessments and training opportunities. There’s no better way to ascertain fit and interest before allocating hiring manager time.
Why hasn’t this hiring model become the norm? As a CEO in the sector told me, “HR is notorious for hanging its hat on platitudes and ‘thou shalt’ belief statements, but is usually too lazy to do the hard work of making a paradigm shift in hiring. That would require thought and attention – not following the newest fad, reviving former practices, or bending to the political wind.”
***
Without a middle ground, we shouldn’t be surprised to see companies revert to primordial hiring instincts while candidates attempt to “vigilante their way into the role.” But as we learned from Hamilton, where the aspiring bus driver was charged with theft, obstruction, and driving without the appropriate license, that probably won’t lead to the job.
If “the most polite bus thief” really did steal the bus because it was the only way left to prove his qualifications, that should make us queasier than a passenger on a bus driven by a reckless, incompetent thief. Skills-based hiring failed not because skills don’t matter, but because we haven’t done the work of building out an operating model. The fix isn’t to return to some hiring yesteryear that was never great to begin with. That’s false hiring nostalgia. Rather, it’s using technology to let candidates show what they can do, early, fairly, and at scale. In a world of digital hiring it makes no sense to limit recruitment to any number of schools. Cast the widest possible net and leverage AI or AI-powered talent marketplaces to assess and respond to candidates at the top of the funnel. Give every applicant a fair shot at landing an interview.
Employers that don’t take the lead on this new standard hiring process will miss out on candidates who could have driven their organizations into the future – and thereby miss the bus entirely.