There has been a lot of discussion in the blogosphere about a research paper by Lauren Rivera that describes how elite professional service firms (top investment banks, law firms, and management consulting firms) go about hiring. The argument is that it is basically a self-perpetuating old boys’ network. Reading the reactions of smart, well-intentioned people with no first-hand experience of this process, and who therefore take her paper at face value, this seems to be feeding into a meme of “the capitalist game is all rigged.”
I think that there is an element of truth to Rivera’s description, but it is mostly pretty misleading. I’ll focus my comments on management consulting, where I worked for about ten years. I participated in every stage of the process from job candidate to new junior consultant to hiring partner.
Start with some quick industry background. You can divide management consulting into “strategy consulting” and “other.” Strategy consulting is the elite end of the consulting business. Most of strategy consulting can be subdivided into two tribes: McKinsey and “Bruce Henderson’s children.” McKinsey is the industry leader. Bruce Henderson founded the Boston Consulting Group (BCG) in the 1960s. A number of BCG spin-offs have occurred since (Bain, SPA, LEK, etc.), and some of these have created further spin-offs. By far the largest and most important is Bain. Together, McKinsey, Bain, and BCG (“MBB”) are the dominant elite recruiters for consulting, though a swarm of smaller strategy firms can compete successfully for the best talent.
There is a lot wrong with Rivera’s picture of how recruiting works, but I’ll focus on three important issues.
Rivera grossly exaggerates the degree to which access is limited to graduates of four super-elite schools.
Employers formally restricted competition to students at the nation’s most prestigious campuses and, contrary to common sociological assumptions about the role of institutional prestige in occupational attainment, having attended a highly selective school (e.g., top twenty five) was typically not sufficient for access to elite labor markets.
Here are about 40 schools in America where BCG and/or Bain are doing on-campus recruiting this year (meaning not just that they will accept resumes, but that they are doing things like on-campus presentations to get students interested, and then doing on-campus interviews): Duke University; Amherst College; Brigham Young University; Brown University; California Institute of Technology; Claremont Colleges; Columbia University; Cornell University; Dartmouth College; Harvard University; The University of Virginia; Princeton University; Yale University; UCLA; University of Michigan; Northwestern University; University of Chicago; Massachusetts Institute of Technology; New York University; University of Notre Dame; University of Pennsylvania; Emory University; Rice University; Southern Methodist University; Stanford University; University of California at Berkeley; University of Southern California; University of Texas; Georgia Tech; Morehouse College; UNC Chapel Hill; Washington University, St. Louis; University of California at San Francisco; Vanderbilt University; Baylor University; Texas A&M University; Georgetown University; Davidson College.
They also hire a lot of people from these campuses. Consider MBA hiring into the associate position, which is the most common starting point for the climb to partner. The top MBA programs are usually considered to be Harvard and Stanford (please save your e-mails, as the point will hold if you move a couple of schools up or down a notch). In the next tier, MBB hired the following numbers of MBAs by campus this year: Northwestern (88), Wharton (85), Chicago (76), Columbia (60), MIT (55), Michigan (40), Duke (36), Dartmouth (24), Berkeley (23), and UVA (20). That’s over 500 hires, or more MBAs than these firms typically hire in a year from Harvard and Stanford.
This is nothing like a random sample of American universities, but it is a much bigger pool than Rivera implies. It’s not the top four schools, but more like 40 or 50 highly competitive schools and 10 or 15 highly competitive MBA programs, that get you access to these firms. Further, while the odds of getting an offer are higher from the most highly ranked schools, a large fraction of each incoming class normally comes from schools not ranked 1–4.
If you get into, for example, Michigan, UCLA, Emory, or the University of Texas, work hard to get good grades in a difficult major, and score very well on standardized tests, you will likely be able to get an interview with one of the leading strategy firms.
Which brings me to the next point.
Rivera grossly overestimates the importance of extracurriculars, and grossly underemphasizes the importance of standardized test scores, and especially, case interviews.
Rivera says of elite employers that:
They restricted competition to students with elite affiliations and attributed superior abilities to candidates who had been admitted to super-elite institutions, regardless of their actual performance once there. However, a super-elite university affiliation was insufficient on its own. Importing the logic of university admissions, firms performed a strong secondary screen on candidates’ extracurricular accomplishments, favoring high status, resource-intensive activities that resonated with white, upper-middle class culture.
There will be exceptions to everything I say, but the way the actual selection process usually works is like this.
Interview slots with the best strategy consulting firms are a scarce commodity. Your resume gets you selected for an interview (though in rare cases, the cover letter and any interaction you may have had with the firm at campus recruiting events can matter a little). Take the example of undergraduate resume screening. Candidates are required to submit resumes, complete transcripts, and SAT scores. As an operational matter, the pile of resumes plus supporting materials submitted for all the candidates for school X is assembled, and each of these candidates is scored independently by three members of a school recruiting team (typically consultants who went to that school). These scores are then combined to create an aggregate rank for all candidates in that school. So, school prestige, which looms large in Rivera’s artificial mock resume screens, matters in selecting target schools, but is not usually relevant in real resume screening because you are screening resumes within a school.
In my experience as a resume screener, the logic normally goes something like this.
If you don’t have at least 750 on the math SAT, you’re out. The most common score is 800. Math plus verbal scores should be well over 1500, and typically over 1550. GRE, GMAT, and other scores should be scaled similarly.
Then, your degree should be in something hard: math, physics, electrical engineering, analytical philosophy, computer science, and so on. It’s okay to major in history or literature, but you better have some really tough quantitative or analytical classes on your transcript, and have done very well in them. If your GPA is below about 3.5, you’re out unless there is some really compelling rationale for why. The average successful candidate has a GPA above 3.7. Everyone understands how bad grade inflation is, and that it’s worst in the most elite schools. Any reasonably smart person with good instincts about course selection can figure out how to get a decent GPA at one of these schools. A GPA-plus-major screen is not about IQ, as much as it is a quick screen to see who is capable of figuring out how to succeed in a new environment, and of doing at least some sustained work. Screeners and interviewers will typically look at the transcript to make judgments about raw candlepower; for example, checking which calculus sequence the candidate completed, and if it was the most difficult track, what grades were achieved.
Prior summer internships at an MBB firm, Goldman, etc., are a strong positive. The reason is not that this means somebody is “clubbable.” The recruiting process for internships has similar resume screening/interviewing steps, and there are even fewer internship slots available at each than there are slots for full-time jobs. Therefore, it means that this candidate has succeeded already in a similar, very difficult selection process.
Finally, extracurriculars matter; but they are marginal compared to these other factors. They are mostly relevant if they show incredible drive. For example, working your way through school in some crap job where you have to deal with people is a big plus.
If you get an interview slot, there are then typically three rounds of interviews. Rivera’s paper claims directly that these are pretty loosey-goosey affairs which people are trying to get a sense of how polished the candidates are:
Interviews — which followed screens — were reported to be highly subjective assessments, where abstract notions of “fit” and “chemistry” routinely drove hiring decisions
Here’s how interviews really work, in my experience.
Interviews begin with first-rounds on campus. Each candidate is given a 45 minute interview, about 44 minutes of which is devoted to presenting the candidate with analytical challenges, and seeing how he or she works through them. The goal is to understand how the candidate can reason analytically — translating an unrehearsed real world problem to a mathematical representation, doing the math, and then translating this back to a real world solution, with awareness of all the simplifications that were necessary — under pressure. This is an attempt to recreate the most challenging part of the job, and is termed a “case” interview. Based on these, a majority of first-round candidates are cut. Second-rounds normally happen the next day on campus. It is a repeat, except that each candidate is normally interviewed twice that day, and each interview is longer. Based on these, a majority of second-round candidates are cut. The remaining candidates are invited back for third-rounds. This takes place at the company’s offices over a full day. Each candidate is subjected to ten or more additional interviews, most of which are case format, but some of which are more typical “let’s talk about you and us” discussions.
Each candidate is then considered holistically, but because so few people in this final selection group have anything but extremely strong grades and test scores, the interviews are usually the crucial way to try identify the highest potential performers. A small fraction of those who started the interview process are offered jobs.
Which brings me to my third point.
Firms sell into a competitive marketplace.
Almost everybody would love to live in the cozy club atmosphere that Rivera described, and it is always creeping into hiring, promotion, and compensation decisions at these firms, as in any human institution. Discipline is maintained only because you have to sell into a competitive commercial market. Unlike how it works in the movies, it is very rare that the COO of a Fortune 500 company hires your case team to do a six month project at $375K per month so that you can sit around and reminisce about rowing crew together at Old Eli. The more vibrant, competitive, and open the overall corporate environment is, the more you will see continued emphasis on finding people who can produce, because the corporations will only pay for what creates commercial value for them. The more the corporate environment becomes dominated by crony capitalism, the more you will see the reverse. And to be more precise, you will see the characteristics that define a competitive consultant become more the kind that Rivera describes: fit, smoothness, and familiarity with the ruling class.
In an earlier post, Steve Hsu made a useful distinction between what he calls the “soft” elite firms that Rivera profiles (investment banks, law firms and management consultancies) versus “hard” elite firms such as hedge/venture funds, startups and technology companies. He argues that the hard elite firms produce something more measurable, and therefore rely less on prestige in selecting people. This distinction is a useful starting point, but what has been happening over the past 20 years or so is the increasing migration of value from soft to hard; basically, to math, technology and analytics-intensive work. This is happening within firms and industries — the emphasis on math ability was growing within consulting in the period I worked in it, as it was within banking — and across sectors as technology start-ups and math-intensive finance became the most obvious ways to make real money in America. This isn’t random, but is happening because these are huge opportunities to create value. This is in part why I left consulting to start an analytics software company. It became obvious that this was the way to create value for clients. This won’t last forever, but has been true for some time.
This should emphasize that the people selected for these jobs are in no sense a “meritocracy.” Most obviously, it has nothing to do with moral worth, as people don’t earn their genes or parents. But more broadly, what I have described is not a hunt for the “best and brightest” in some general sense, but rather for people who have an incredibly arcane bent of mind and set of ambitions that fit into an environmental niche that they didn’t create. This is as true of “hard” elite firms as “soft” elite firms. Despite all the chest-pounding, 50 years ago most of these people would have clerks; a couple hundred years ago, not especially successful farmers. Fifty years from now, for all we know, these will no longer be especially valuable talents.
Ironically, moving away from the idea that firms are looking for “the best” or “most worthy” in some general sense, and simply recognizing that they should look for predictors of success within a given business model relevant to a specific point in economic history, allows not only more effective hiring decisions, but also a little much-needed humility.