The conventional wisdom - the wisdom that I bought into - has been that we are getting ever more mobile in our careers. Long-term employment is dead, we’re told. The career of the future will be one where we constantly hop from one job to another. Even being employed is old-fashioned; the future belongs to the freelancer, the contractor, the gig-worker.

It’s a popular account that fits with everything we hear about how the world is going. It’s surprisingly hard, though, to get the facts to line up with that story. And I think that discrepancy is important. If we spend all of our time preparing for a high mobility future that isn’t coming, it means we are not developing the strategies that will actually help us succeed in the real world.

I’ve got a book coming out (today!!) that lays out how we build successful careers inside organizations rather than across them. In this post, I want to go through how work isn’t changing, and why that matters.

Subscribe now

Employee Mobility - Myth vs Reality

My students love to tell me that their generation moves jobs much more than the prior generation. But do they? The US government has been collecting data on quit rates for around 25 years. As the graph below shows, quitting is highly cyclical. Quit rates spiking when the economy booms (as it did in the great resignation) and plummet during downturns (see the aftermath of the great recession). Outside these cyclical effects, there isn’t evidence of an overall shift towards higher mobility. Indeed, US quit rates today are lower than they were 25 years ago.

Nor has the amount of time that we tend to stay in employers fallen in the way that expected (or indeed in the way that everybody now assumes it did). The government only surveys people about tenure every 2 years. Drawing on the latest data, we see that average tenure (time you have been at your employer) went up following the great recession when it was hard to move jobs and fell a little after that. Overall, it hasn’t changed markedly over the last 25 years.

Nor are our jobs becoming noticeably less secure. I’ve graphed the layoff rate from the government’s JOLTS survey using a log scale to make the Covid spike manageable. As you can see, the overall trend over the last 25 years has been for layoffs and firings to become less common:

How about the growth of freelancing and gig work? We have surprisingly poor data on such fashionable free agent arrangements, but what evidence we have suggests that they aren’t taking over the economy. The government runs an irregular “contingent work supplement” that aims to count the numbers of temp workers and independent contractors (among others). Their numbers have been curiously stable for the last 30 years:

We have even less data about “gig” or “platform” workers who work in the app-based economy. Back in 2015, I was interviewed for an article in The Economist that claimed that “Freelance workers available at a moment’s notice will reshape the nature of companies and the structure of careers.” (I didn’t end up being quoted; I need to work on my soundbites). I asked Claude to do a rough and ready calculation of the size of the gig economy by estimating the total paid to gig workers by the biggest gig firms (Uber, Lyft, DoorDash, Instacart and Lyft).

$70 billion is a tidy sum, but it is less than 0.5% of the $12.3 trillion that workers earned last year overall: here that is in graph form, in case you find it easier to understand:

The tiny proportion reflects the way that the gig economy revolutionized taxi services and takeout, but pretty much nothing else.

Why Did We Get It Wrong?

Why are the stories that we tell about the changing nature of careers so divorced from the statistics? I think that there are three reasons (yes, I did spend time working as a consultant; why do you ask?):

1. We forget how long it has been since we saw a genuine rise in mobility. Some people did get more mobile in the 1980s and early 1990s as companies rethought how they managed their people. But those shifts were largely complete by the mid-1990s, some thirty years ago. At the time, we thought this was the start of a larger breakdown in long-term employment and a move towards ever greater mobility. In hindsight, it wasn’t.

2. We confuse generational differences with secular trends; put another way, young people have always moved jobs a lot. One study found that between 1957 and 1972 - the heyday of long-term employment relationships - young men moved firms an average of seven times in the first ten years of their careers. We move around when we are young, but tend to settle down a little as we get older. Each new generation expects, though, that they will continue to move around when they are older, just as they did when they were young. And so the myth of accelerating mobility is perpetuated.

3. We have over-emphasized the importance of flexibility, while underestimating the importance of context. We’ve lived out the last three decades against a constant drumbeat of demands for flexibility. Rapid technological change, geopolitical turmoil and intense competition highlight the value that companies get from flexibility. They talk about the need for talent on demand or a liquid workforce (and not the kind you used to get after 3pm on a Friday afternoon). And yet a central thread of my research has been all of the ways that people work better in familiar organizations. You might have learned this last time you changed firms. Companies know it - they complain all of the time about the cost of attrition, and how hard it is when they have to replace departing workers. When you do your best work in familiar surroundings, it turns out that the kind of fast-moving, free agent, gig-led work arrangements we were promised don’t work for anyone.

Why This Matters for You

Stories about how careers are changing aren’t just descriptive. They also carry a strong normative component - that you should be moving regularly to keep your career on track, that staying in the same job is somehow old-fashioned or doing it wrong. That’s not the case. In fact, in my research I’ve found that is often the people that stick around in organizations that end up advancing faster in their careers.

More broadly, this focus on mobility is symptomatic of our failure to take seriously what it means to become effective in organizations, and how the collaborative nature of our work rewards familiarity. I don’t want to underplay the problems that uncertainty cause for maintaining long term employment relationships. But I think we often lose sight of the value of familiarity when it comes to doing good work and advancing in our careers.

I’ve developed this theme extensively in my new book, where I really dig into what it means for how we build successful careers. If you’re interested in going into more detail on how people actually build successful careers, drawing on the data rather than the latest hype, I encourage you to check it out. You can find it here.

Subscribe now

I’ve recreated a version of their data here (thanks, Claude), changing their comparison slightly to look at people who move employers versus those who don’t.

The data tells a pretty clear story: it’s job switchers for the win! So am I giving bad advice? Should we all start hopping our way to increased dough?

This chart bugged me enough that I thought I should dive a little deeper into the data. Does the evidence suggest that moving firms is an important step to increase your pay? Let’s find out.

Lies, Damn Lies, and Job Hoppers

A key part of every statistics class is learning to chant the mantra “correlation is not causation” to the point that the observation starts to feel really trite. I think that the interesting question in any set of statistics is “what is the reason that isn’t correlation causation? What else might be going on here?” Those are particularly important for interpreting the evidence on job switching.

Here, the question we need to ask ourselves is: “who moves and who stays?” If the kinds of people who move firms are also the kinds of people who you would expect to get bigger pay raises anyway, it might explain the above chart.

Think about how movers and stayers might differ. Two reasons leap out at me.

1. Movers are younger than stayers. When I teach about careers, I try to drum into my students that people move much more when they are young. It usually takes a little while to figure out where you fit in at work, and people shift firms to find the place that is right for them. Younger people also see faster general wage growth, because they are learning new things much faster than oldsters like me. Part of the reason that movers get bigger pay raises than stayers on average is that movers are disproportionately young people who get bigger raises anyway.

2. Movers earned less in the first place. This is the big one (some might even call it a load-bearing reason, but I’m not going to). Although pay isn’t the only reason people switch firms, it is important. We know that people who are paid less are more likely to quit their jobs. But people who are initially paid less are also more likely to get bigger pay raises because there is more ground for them to make up.

So if we wanted to get a cleaner comparison of pay changes for movers and stayers, we would want to be comparing people of similar ages who earned similar amounts. This, it turns out, makes a big difference. I corrected the chart above for differences in age and starting pay (specifically, I calculated a “predicted pay change” based just on age and initial pay, then subtracted that from the actual pay change; I also added a constant so that the median would replicate the median in the original data). When I do that, the chart looks very different:

Switchers no longer consistently earn more. Switchers’ pay growth is more responsive to economic conditions, so they got higher pay increases during the heights of the great resignation, and substantially lower pay increases as hiring leveled off in 2024/25. For most of the last five years, though, it has been stayers who did better. Once we account for the way that switchers tend to be younger and worse paid than stayers, we don’t see a consistent financial advantage to changing jobs.

In the Long Run (But Before We’re Dead)

I’ll admit that I was quite pleased to see that the usual data on pay gains to movers versus stayers didn’t point to big gains for moving jobs. But I had already been doubtful about the whole idea of comparing pay gains over a single year.

Most people, I think, care a little bit about the immediate payoff to changing jobs, but much more about the longer-run effects. The important question, therefore, is not: “Can I get a pay raise now by moving firms?” But instead, “Will moving firms set me up to earn more in the coming years?”

I’ve often worried that looking at immediate pay gains may overstate the returns to switching employers. After all, the deal you get when you join the firm can be the best it’s going to get: Employers will often pay over the odds to persuade you to sign up, and will then spend the next few years trying to bring your pay back down to the norm, usually through lower pay raises. There’s also a risk that moving firms increases the time it takes you to get your next promotion. If so, moving firms could be gaining a pay raise today at the cost of a potentially larger raise next year.

To get a better picture of how moving jobs affects long run earnings, I took a look at data from the National Longitudinal Survey of Youth (NLSY). The NLSY is a really neat government program that has followed the same group of people for nearly thirty years, interviewing them every 1-2 years to ask about what is going on in their lives. The resulting dataset contains details on all of the jobs the respondents have ever held, allowing us to see how the number of moves somebody has made in their career relates to how much they are paid.

Below I’ve plotted respondents’ pay in 2022 (when they were 35 or 36) by the number of job switches that people had made in the past. I did some light data cleaning to try to make things comparable. I only counted new full-time jobs (i.e. jobs with more than 30 hours that lasted for more than 3 months), and I started the clock at 24 (by which time normal people have left full time education). I also corrected pay for differences based on education, sex and race. The box plot gives the pay in 2022, broken out by the number of job changes a respondent had had. It’s a box plot, so the middle line is the median, and the boxes span the range from the 25th to 75th percentile.

What does the chart show (other than that I have no future as a graphic designer)?

First, it shows that people who had moved jobs more ended up earning less. We see a smooth decline in average pay as people switch jobs more (this decline stops at 8 job changes, but since that number is based on only 6 individuals out of 2700, I am comfortable ignoring it). Moving employers isn’t generally associated with increased pay. The people who have moved employers the most are actually earning the least.

Second, though, it shows that the association between pay and moving jobs is generally pretty weak. There is much more variability among people with the same number of moves (this is the range indicated by the top and bottom of the boxes) than there is between the averages of those different groups.

Is There a Takeaway?

All of the usual caveats apply here (okay, maybe not the one about stocks going down as well as up). These are correlational analyses. They don’t tell us what would happen if we randomly assigned job moves to some people and not others (although layoffs come pretty close to that kind of randomly assigned job switch, and they definitely lead to lower pay).

That said, it’s very hard to find evidence in the data that moving firms is an important strategy for pay growth. I’m sure it is helpful for some, particularly those that start their careers in low paying firms and low paying industries. And I don’t think that moving firms hurts pay much either; you tend to see fairly small differences between those that have moved more and moved less. But the next time that somebody tells you that it is important to move around to boost your pay, it might be worth taking them through the data, one chart at a time. Feel free to direct the questions to me.

Subscribe now

On Insider CEOs

First, I wrote an op-ed for Fortune about Apple’s new CEO, John Ternus. You can find it here. The piece comments on the surprisingly old-fashioned nature of John Ternus’s career. Ternus joined Apple in 2001, a mere four years after graduating the University of Pennsylvania, then spent 25 years working his way up through the company. Perhaps even more surprising, this kind of internal path to the top is not uncommon, even within the tech industry which has long trumpeted the value of mobility: both Andy Jassy, the Amazon CEO, and Satya Nadella at Microsoft followed very similar routes to the top, working at their employers for more than twenty years before being nominated to the CEO job.

After writing that op-ed, I have been asked by a couple of people about survivor-bias. Sure, sticking around in the company might work for those people that eventually become CEO, but very few of us will actually become CEO of a Fortune 500 company (I’m still hopeful). Is there much that the rest of us can learn from their career paths?

It’s obviously true that staying at the same company doesn’t guarantee you a path to the C-suite. But the studies that I cite in the op-ed point out that you are more likely to advance within companies compared to between them. When we studied Wharton alums, internal moves were much more likely to entail big increases in responsibility compared moves across firms; a study of Finns found that big promotions were far more common in internal moves than shifts across firms.

Sooner or later, most people who are not C-suite-bound reach a stage where they are no longer able to move up by staying at the company; that’s a great time for them to think about switching firms to continue progressing their careers. But when you still have advancement opportunities inside your company, those internal moves will usually progress your career faster than looking for something outside.

A Very Alternative CEO Career Path

There may still be some people who fancy becoming a CEO but don’t like the idea of having to work their way up from within the company. There are other options, as I discovered while researching my book. One is to try to imitate James Daunt, CEO of Barnes and Noble.

Daunt started his career as an investment banker, before his partner persuaded him to quit and try something different. Although he decided to go into a very different line of work - setting up a travel bookstore in North London - he still approached the problem as a banker might. He was acutely aware of the need to make a healthy margin if the venture was going to survive the ups and downs of the business cycle, and spent two years figuring out how to do so before setting up his shop.

Fast forward a few years, and Daunt had successfully established “Daunt Books” (you may well have seen one of their tote bags, even if you’ve never visited a store) and grown it to 4 branches. But he had bigger plans. Having worked out how to make money in the bookstore business, he figured he could apply his approach at a much larger scale. And so he persuaded a backer to buy Waterstones, Britain’s largest bookseller, and install him as CEO. He proved so successful that when one of his investors bought Barnes & Noble in 2019, he was named their CEO as well.

So, yes, there are multiple paths to the top. If you don’t like the idea of working your way up to the C-suite, maybe you can persuade somebody to buy a large corporation for you. Maybe. But you’d better be seriously persuasive. For most of us, the internal route remains the most likely.

Which Brings Me to My Shameless Plug…

Of course, if you are planning to go the buyout route, you will need to be able to demonstrate the kind of serious business acumen that Daunt so ably displays. Which brings me to my final point:

Barnes and Noble currently have a 25% off sale on pre-orders, the kind of innovation that could only result from a career combining investment banking, entrepreneurship and bookselling. And that means that you have a reduced price option to pre-order my upcoming book, “The Insider Advantage.”

Here’s the link:

https://www.barnesandnoble.com/w/the-insider-advantage-matthew-bidwell/1149525861?ean=9781668079171

Go on, you deserve it! Buy two!

Subscribe now

Subscribe now

If we want to talk about how we develop people for an AI future, we need to understand how the use of AI changes skills across most jobs, not just those in technology. Recently, I’ve been working on a project that has been exploring exactly that question. Our early results suggest that we need to start thinking about how we train people to be managers in every job.

Another dubious AI illustration. It captures my main point, but I think she needs to rethink how she chooses where to hike.

The Story So Far

A common theme in most research on AI is that has been reducing demand for younger white-collar workers. The argument is that AI systems are best suited to automating the kinds of less complex tasks that would normally be done by junior workers (i.e. why hire an entry-level worker when you can build an agent?). At the same time, AI may raise the productivity and value of more senior employees doing more complex work.

There is decent evidence in support of this story. An influential early study of payroll data found that employment of young workers has fallen since the arrival of ChatGPT, and that the declines were concentrated in the kinds of jobs most exposed to AI. Although some of those declines may simply reflect the effects of rising interest rates, other more detailed studies also link AI to declining entry-level hiring. A particularly impressive study used LinkedIn data to show that firms adopting AI technology reduce hiring into entry-level and junior positions relative to more senior roles.

Breaking white collar work into simple versus complex tasks is a straightforward way to understand AI’s impact and fits the slowdown in entry-level hiring. That said, I’m not sure it fits all the facts. Many tasks that AI excels in, from reading X-ray scans to hacking into Hugging Face seem pretty complex - to me, at least. Unless we revert to tautology - complex tasks are the ones that AI can’t do - I think we probably need new ways of thinking about what gets done by AI in the workplace, and what kinds of skills will help people add further value.

An Alternative Perspective: Doing versus Deciding

Together with my highly talented colleagues Manav Raj and Susan Wang, I’ve been playing with this question. We’ve been analyzing a huge dataset of pretty much every job-posting put online between 2021 and the beginning of this year, and looking at “within occupation” changes in job-posting - i.e. how job postings for the same occupation have evolved as companies make bigger investments in AI (our imperfect measure of those investments is how much AI skills appear in some of their job-postings).

Our early findings suggest that focusing on the difference between doing and deciding can help us understand how AI is changing jobs. As AI systems grow increasingly powerful, they have become capable of doing a wide range of different stuff. Where they don’t yet have an edge (we think) is in making complex decisions about what to do. Yes, context windows have gotten much bigger, and LLMs encode a mind-blowing amount of knowledge through their training data. But good decisions in real organizations require a deep understanding of the organization’s goals and the full range of activities going on within the firm. That breadth of context is not yet available to AI systems. Deciding, then, currently remains the province of humans.

That difference between doing and deciding provides a decent explanation for how job requirements are changing in companies that adopt AI. Specifically, we’ve found the following changes in job postings among companies that are more active in hiring for AI:

• More demands for coordination and problem-solving skills. As workers spend less time on individual tasks, they start to spend more time on the more strategic aspects of the work, figuring out what they should be doing. That kind of deciding emphasizes working with others to set goals and priorities, managing entire processes rather than isolated tasks, and solving the more complex problems that will inevitably arise. These skills are appearing in more job postings as companies adopt more AI.

• Increased value of breadth. Another effect of the switch from doing to deciding is that breadth of experience becomes more valuable. To perform one task, you just need to be able to perform that task. To make decisions about what to do, though, you need to have a sense of how that task fits into the broader enterprise. Several studies have found that firms value breadth when they hire managers. As AI allows everyday workers to do more managerial work, breadth could become more valuable across a lot of different kinds of work. That, at least, is what we find: AI adoption is associated with companies asking for more skills and more varied skills in their job postings.

• Experience. Of course, building breadth takes time. You don’t just need to master one set of skills, you need to learn a range. The ability to solve complex problems requires being able to carry out the basic tasks first. If employers want generalists who can do more of the deciding, they probably are looking for more experienced employees as well. Consistent with earlier work, we find that experience demands are going up in AI-forward firms.

What It Might Mean For You

These results are very preliminary in two important sense. First, we haven’t been through peer review, and I’m sure that reviewers will point out many opportunities to develop our ideas and evidence as the paper progresses. Second, AI is changing very fast and companies are still in the earliest phases of figuring out how to work with it. The kinds of skills that AI-pilled companies will be looking for in five years’ time may be very different.

That said, I do want to offer some brief, poorly thought-through ideas about the possible implications of a shift from doing to deciding. To my mind, they include:

1. Build breadth. Everybody faces decisions about how much they should build focused expertise or try instead to move across different areas in search of new perspectives. As I’ve noted elsewhere, breadth seems to be important for those people looking to move up the managerial ranks. Our analysis suggests that the balance between generalists and specialists may be shifting for individual contributors too. Being too focused increasingly looks like a career risk. Emphasizing new opportunities that enable you to broaden out your skill set may be increasingly important as AI takes hold.

2. Focus on learning why, not how. If we’re moving from doing to deciding, focusing on technique may be a waste of time. We need to know enough about how the work is done to second-guess the LLMs; I suspect it may be a while before they are fully reliable. Understanding how to do the job perfectly may be less valuable. Learning about the full context of the work, where it fits into the organization and the market, how it will be used and how it is most useful, may be what really differentiates you. One of the big lessons about becoming a manager is that people often overestimate the importance of managing down and underweight the need to manage sideways and up. That may be becoming true of more and more roles.

3. Umm, you need to be older? Okay, that’s probably not very helpful advice. But all of these studies showing growing demands for experience raise questions about how young people will first get their feet on the job ladder. You can’t speed up aging, and nor should you want to (time, that relentless little bastard, takes care of that all on its own). But if companies are going to place more value on experienced workers, it’s going to be really important for young people to figure out where they can get the kinds of experience that will help them build breadth and learn why. Maybe that will involve taking on less well-paid jobs as a kind of apprenticeship. Maybe companies will rethink their career pathways; but that early experience may be becoming increasingly important, and increasingly hard to find.

Subscribe now

More specifically, the final assignment in my “Understanding Careers” class for MBAs requires each student to go out and interview five people about their careers. The interviewees are supposed to be at least 20 years older than the students, and some of them should be in fields they want to go into. Because my students aren’t dumb, they try to interview the kinds of senior, successful people that they hope to emulate. They then write up a 7-10 page summary of what they’ve learned. In total, I reckon my students have completed more than 8,000 interviews over the last decade. That’s a true crowd of people, so I figured that there ought to be a lot of wisdom hiding out in there, if only I could go in and find it.

Enter the CareerBot

Even reading about 8,000 interviews sounds like a lot of work to me; the assignments total 16,000 pages, and I’m pretty lazy. So, like a lot of people faced with a task they don’t really want to do, I turned to AI. After a lot of false starts and general faffing, I created an RAG (Retrieval Augmented Generation) program that I call “CareerBot.” I feed the program a question, it searches through the body of assignments to find the most relevant passages, organizes and summarizes them into different themes, and gives me a count of how many times each theme occurs. It’s not perfect, but that didn’t stop me from feeling unreasonably proud of myself when I got it to (kind of) work.

I figured I would start off by asking the most basic question: what were the main lessons that the interviewees offered about careers? In true listicle tradition, here are the eight top themes, in reverse order of how often they were discussed:

8. Adaptability is a Critical Skill in Any Career

One of the first things that my students hear from the people they interview is that you can’t plan a career, or perhaps more accurately, that even if you do plan your career, it won’t play out in the way that you expect. Plans got derailed by unexpected events. Unanticipated family demands led people to reevaluate what they wanted, or a bad boss stopped them from advancing at their firm. Often, unexpected opportunities would lead them to pivot into a path that they had never considered.

When I hear people tell me that you can’t plan a career, I’m reminded of Eisenhower’s saying that “plans are worthless, but planning is everything.” Careers don’t work as planned, but it’s still worth spending time figuring out what we want, where opportunities might arise, and what sorts of jobs are more likely to lead to the positions we want. A clearer idea of where we might end up and how we might get there can help us make better informed decisions. We always need to be ready, though, to seize unexpected opportunities when they arrive, and adapt to the inevitable setbacks that will come our way.

7. You Shouldn’t Be Afraid of Taking A Risk

Another key them was the value of taking a risk. The interviewees had often benefited from seizing interesting opportunities, even when the outcomes of those opportunities had been highly uncertain at the time. They knew that they were taking on a lot of risk, but had the courage to take those opportunities nonetheless. They recommended my students do the same.

I’ve always been a bit nervous about this one, to be honest. My students go out and interview people who have been successful in their careers — the people, in other words, for whom the risks have generally worked out. Is there another set of people, too unsuccessful to be interviewed, who would tell them with equal conviction never to take any risks, since it had worked out very badly for them?

Probably not. Although we want to be careful in our decisions, there aren’t that many career risks that we can’t come back from (with the exception of major legal or ethical infractions). Taking risks that work out can often turbo-charge our careers; risks that don’t work out may slow our advancement, but they rarely block it altogether. Research suggests, for example, that starting your own company is a pretty good financial bet despite the high risk of failure. If you succeed, the payoffs are often very high. If you fail, you can always go back to your earlier career with little long term penalty. I suspect that the same is true for many risky career decisions.

6. Luck Plays An Important Role in Every Career

Continuing with the uncertainty theme, luck came up a lot as people talked about their careers. Although we hope that our talents will bring us career success, luck usually plays a big role — the random conversation that leads to a transformative job, or the unexpected opportunity that takes us in a completely different direction.

Although the interviewees saw that luck had played an important part in shaping their careers, some of them also reflected that their luck wasn’t totally random. Yes, we often end up benefiting from completely unexpected, unpredictable opportunities. But often, the people who find that sort of luck have put themselves in positions where those opportunities are more likely to occur. Maybe they have worked hard to build out their network (see lesson #5), or maybe they have built a reputation for professionalism that leads other people to seek them out. Maybe they have been thoughtful about finding the kinds of growing fields where opportunities are more likely. They have also been ready to seize those opportunities when they arise, taking the risks that so many people advocate. Luck does play a role in all of our careers, but some people are probably more successful when it comes to making their own luck.

5. Broad Networks Are a Critical Resource

No discussion of careers is complete without somebody underlining the importance of your network, and it seems that my students’ interviewees were happy to oblige. Networks, they emphasized, matter. Your career is never just your career; it’s something you accomplish with the help of countless other people. The more people who can give you that help, the better off you will be.

One common observation was that broad networks are particularly helpful for finding new jobs and opportunities, either within or across firms. Networks can also be a source of advice and support, keep you up to date on the latest trends, and provide the inside information that helps you do your job. The broader your network, the more that you can gain from it.

4. Mentors matter

Some relationships are more important than others. Mentors - senior figures who invest in your development and help you succeed - played a valuable role in many of our interviewees’ careers. Mentors can help increase your visibility, get you access to new opportunities, expand your network, and provide invaluable coaching.

It’s one thing, of course, to know that mentors are important; it’s quite another to secure mentoring from somebody in a position to turbo-charge your career. A lot of the lessons from the interviews surrounded how to find and manage mentors. The key insight? Ask not what your mentor can do for you, but what you can do for your mentor. The most valuable mentors often have limited time and energy. They are more likely to invest in the people who will be most helpful to them in turn.

3. Pursue Your Passions (within reason)

Money and promotions feature prominently within discussions of careers, but my students heard a lot about the importance of finding something that would make them happy. Interviewees stressed the importance of finding things that you want to work on day in day out, of doing work that provides purpose, of finding things that give you meaning.

Like much good advice, this comes with a few caveats. Not everyone can turn their passion into a viable career (no professional cycling contract for me, then), and some interviewees warned about the need to balance passions with pragmatism. Not everyone has their passion figured out, either. One of the reasons that people take my careers class is that they haven’t figured out what is right for them. Nonetheless, the importance of looking beyond material rewards to think about what makes you happy came through loud and clear in the interviews.

2. Be Careful About Moving Firms too Frequently

The value of staying with your employer was a surprising theme in the interviews. Although we keep being told about the value of agility and free agency, that wasn’t how interviewees had experienced their careers. Yes, people did sometimes benefit from seizing the right opportunities to move; but they talked about the real downsides of moving companies too often. Instead, the interviewees noted the value of growing inside the right company, building responsibility and showing senior management that you are there for the long run. I was pretty gratified to see this theme come out so strongly, seeing as how it is also the theme of my forthcoming book (still available for preorder!!).

[Cue drumroll for number 1…]

1. Find a Balance Between Your Career and Your Family

The most common lesson that we heard from the interviewees was about the need to find a balance between career and family. My students are pursuing careers in finance, consulting, management and entrepreneurship. The people they interviewed had extremely demanding careers. The importance of balancing those work demands with the equally important demands of the family was the most common lesson my students took away from their interviews.

It is, of course, easier to say “you need to find balance” than it is to find it. Some people had found balance through seeking out jobs that would give them flexibility. Others had only realized the importance of balancing work and family once it was already too late. From both groups, we heard that work-life balance is a real challenge in managing a career.

So there you have it. 1600 assignments, 8000 interviews, one half-baked attempt to analyze them using the magic of AI. It’s not the most rigorous piece of work I’ve ever done — there is a certain art to interpreting the themes, and I’m not sure how much the conclusions were swayed by how I taught the class — but a lot of the lessons ring true to me, and they seem a pretty good a way of summarizing many of the most important things to consider in managing a career.

In my coming posts, I’ll be working with CareerBot to unpack some of these ideas a bit further. Please leave a comment about specific questions that you want me to run past CareerBot.

Leave a comment

Together we can learn more from the experiences of our 8000 interviewees.

The fundamental idea behind the Peter Principle goes something like this: performing well will generally earn you a promotion, so that the better you are at a job, the less time you are likely to stay in it. You’ll stay longer in the jobs that you can’t do - those that are at the level of your own incompetence - because those are the ones that you won’t get promoted out of. As promotions are used to move people out of jobs that they can do but not the jobs that they can’t, roles gradually get filled up with people who are not up to their requirements.

Now, it’s easy to pick holes in this idea. Underperformers may not get promoted out of their jobs, but they are often fired or demoted instead. I don’t think that most organizations are filled up with people who can’t do their jobs (although a number of my experiences with American Airlines has left me wondering). And the Peter Principle was always intended to be satirical. That doesn’t mean, though, that there isn’t a lot we can learn from it about building a successful career.

What the Peter Principle Gets Right

The Peter Principle rests on a core insight: the kinds of skills and behaviors we need to succeed in our job often change when we get promoted. The things that made us a high performer in our old job (and helped us get promoted) won’t necessarily continue to make us successful in the next role.

Take salespeople. Suppose you want to get promoted as a salesperson into a management position. How do you go about doing that? Easy: you sell more than your colleagues. But is it the best salespeople who go on to make the best sales managers? Apparently not. A classic study of sales managers across 121 different firms found that the better those managers had performed as salespeople prior to promotion, the worse they went on to perform as sales managers (as measured by their effects on subordinates’ sales).

Why would good salespeople make bad managers? Sales professionals will tell you that the best salespeople are often driven by an incredible drive for personal accomplishment. Yet the best managers work through their teams, stepping back and delegating, operating through subtle influence rather than personal heroics. They don’t close the sales themselves; instead they focus on making their teams better. Those are very different kinds of behaviors to the ones needed to be a good salesperson, behaviors that are generally exhibited by different kinds of people. Nonetheless, companies persist in promoting people based on their prior performance. This means that being a good salesperson makes you both more likely to become a sales manager, and less effective once you do so.

The Peter Principle and You

Okay, so the Peter Principle is a cute idea that seems to work in the real world. But why is it important for your career? Because the Peter Principle reminds us that what it takes to be successful often changes over the course of the career, as skills and behaviors that help you in lower level jobs can set you back later. This has two implications: first, you need to be very careful to rethink your approach each time you get promoted into a new role; and second, you need to balance building skills for your current job versus building skills for future jobs. Let’s talk about each of those in turn.

Managing Transitions

Why do successful individual contributors fail when they step up into management? Sometimes it’s because their personalities and skills are a poor fit for the demands of the new job. They’re just the wrong person for the role. But a second reason that they fail is because they don’t understand the need to adapt to the new job. When you’ve been successful all the way along the line, it’s easy to believe that you have everything figured out, and hard to appreciate that you might need an entirely new playbook for your new role.

This challenge is beautifully described by Linda Hill in her classic book on first-time managers. She shows how hard it can be for new managers to learn what is expected of them and adjust their behavior accordingly. It turns out that little in these freshly promoted managers’ experience has prepared them for their new roles, and their employers rarely explain their new expectations to them. Nor are these challenges restricted to first time managers, as Michael Watkins explains. Throughout your career, you are asked to step up into new roles that require very different approaches, but often with very little guidance as to what you should do differently. It falls on you to figure out what is now expected of you. Fail to adapt effectively and you will, well, fail.

Your promotion will often be the culmination of years of hard work, a recognition of your skills, an endorsement of your working style. Yet it is also the moment for you to throw away those laurels, not rest on them. As Watkins and Hill emphasize, continued success requires a high degree of openness, a willingness to reconsider your approach and behaviors, and a flexibility to learn the behaviors that are going to make you successful in the new job. You cannot continue to rely on the habits that helped you succeed in the old job. If there is one thing that you should take away from the Peter Principle, it is the importance of asking “what do I now need to do differently?” each time you move up a rung on the career ladder.

Training for the Job You Want

If there is a second thing you should take away from the Peter Principle, it is the difference between building the skills that will serve you today versus those you will need in the future. As jobs change along the career, so do the skills they require. The problem is that we build skills slowly, through years of experience. While it might make sense to focus on sharpening the skills needed for our current role, speeding our time to promotion, that short-term focus may leave us unprepared if we do make it to the next level.

I explored this dilemma in a recent paper on the careers of corporate scientists, written with Shinjinee Chattopadhyay and Shinjae Won. As with salespeople, we found that the characteristics that predicted promotion into management for scientists were different from the characteristics that predicted success once they became managers. This time, though, those differences revolved around the kinds of skills those scientists built.

The scientists who got promoted faster were the ones who were more specialized. Academics are more productive when they are more specialized, building expertise in one particular area rather than trying to master multiple topics. Among our scientists, initial success also seemed to come through focus: those that patented in a narrower set of technologies were promoted faster.

After becoming a manager, though, the pattern reversed. Now, it was the managers who had patented in multiple areas who were more likely to continue to be promoted. Breadth may not help people to be more productive scientists, but it gives them the diverse perspectives and broad networks that help them to be more effective as managers. Once again, the kinds of behaviors that helped somebody become a manager were different from the ones that helped them succeed once they were a manager.

What really interested me was the lingering effect of the projects people had worked on while they were still individual contributors. Because it takes time to build a portfolio of projects, work done as an individual scientist represented a substantial part of managers’ overall experience and continued to shape promotion rates even once they entered those management roles. Hence the dilemma for the scientists. Those who worked on a broader set of technologies in the early stages of their career went on to more rapid advancement once they were managers; but that breadth made it harder to get that initial promotion into management.

Now, you’re probably not a scientist wondering what kinds of projects to pursue, so the exact lessons about breadth versus depth may not ring true within your field. But I do think that there is a general lesson in these results. The kinds of skills that you will need to draw on are going to change over your career, as you continue to encounter new kinds of challenges as you move up the ladder. As you prepare yourself to move up, it’s worth trying to build an understanding of those demands, and seeking out the kinds of experiences that will prepare you for them. To bastardize a common cliché, the Peter Principle suggests that we shouldn’t be training to do the job we have, but rather for the job that we want.

Subscribe now

My first serious work experience came through an internship that I did when I was seventeen. At the time, I was studying sciences and planning to major in chemistry at university. For four weeks during the summer, I trekked out by bus and train to a chemistry lab on the outskirts of London, where I investigated new catalysts for the petrochemical industry. The internship taught me that I wasn’t interested in being a career scientist. Working in the sciences, I learned, meant spending all day alone in front of various machines. That kind of life wasn’t for me, I realized. I wanted to work with people.

Roll forward a few years. After another false start, this time in consulting, I found myself in academia. How did I spend my days? Sitting alone in front of a computer screen. Oh, well.

I found myself reflecting on those questionable choices as I was reading the latest research on remote work. The shift to remote working has been dramatic, with around 25% of working days in the US now being worked from home. It has also been controversial. For much of the last five years, commentators have been feuding over whether remote work is good or bad for productivity (well-designed studies support both viewpoints). There seems to be a much clearer consensus among CEOs, who view remote work much as I do frozen yogurt - as an abomination. A recent study, though, suggests that we might have been arguing about the wrong question. Rather than asking whether remote work is bad for companies, the study asks whether it is bad for employees.

Obligatory dubious AI illustration

Up until now, the evidence has strongly suggested that remote work benefits employees; indeed, their strong desire to work from home is what has allowed remote work to persist in the face of managerial backlash. Surveys consistently report that employees would prefer to work from home for at least two days a week and often more. They similarly report that employees believe that working from home for at least some of the week would be better for their mental health. Working from home helps people avoid a commute that we know weighs on life satisfaction. It also provides people with the flexibility to manage work-life conflict. Estimates suggest that people will give up anywhere from 4% to 25% of their pay to be able to work remotely.

Then along comes a study in Science arguing that remote work is bad for people’s mental health. The evidence that the study provides is indirect (they don’t have data on mental health for people who are working remotely versus at home) but thought-provoking. The core of the paper shows that:

1. Overall rates of mental distress have risen substantially over the last six years (given the six years we’ve had, you might worry if they hadn’t).

2. That increase in distress is significantly stronger among people who are in “remotable” jobs, in which people tend to spend more time working from home.

The evidence isn’t bulletproof. Remote work expert Nick Bloom points out that people in stressful life situations could be using remote work as a coping mechanism. We know that depression leads people to want to retreat from social contact (and not just our own; I recently had a manager complain about a subordinate who couldn’t come into the office because her cat was depressed!). It seems very plausible - even likely - that people who are suffering anxiety, depression and other mental health issues would choose jobs that allowed them to work remotely.

One Simple Trick to Guarantee Happiness.

The study did prompt me to start thinking about my own life choices, though. That’s because it echoes a lot of other evidence that suggests that our workplace relationships (the ones I gave up to do a PhD) are unusually important to our overall well-being.

I had learned during my internship that time spent with other people is an important part of our workplace experience. It was only while researching my book on careers (available now for preorder!! :) ) that I gained an appreciation for just how important those working relationships can be. For example:

• The General Social Survey details evaluations of job satisfaction and working conditions for thousands of workers. When I explored the main determinants of job satisfaction, I found that the strength of working relationships were much more important than other working conditions like stress, physical conditions, or variety. They were also far more powerful in shaping job satisfaction than how much people were paid.

• A very cool study by Krueger and Schkade asked people how they had felt at different times during the prior day. The study found that people were happier when they were interacting with coworkers than when working alone, and that people whose job involved frequent interactions with coworkers reported being happier and more satisfied with their jobs.

• Another study, this time of Canadian data, found that trust in coworkers and managers had a profound effect on life satisfaction.

  

Results of analysis of GSS job satisfaction data. Main variables come from factor analysis of working condition questions.

So there you go. The one simple trick to guarantee your happiness at work and in life is to find a job where you work with people you like (simples. can I have my TED talk now?). Again, this is not to say that remote work is necessarily making people unhappy. But from everything we know about what makes us happy, it wouldn’t be surprising if it had some negative effects.

Do We Know What We Need?

As well as highlighting the critical importance of workplace relationships, the latest study on remote work resurfaces the old age question, beloved of parents and despots alike, of whether people really know what’s best for them. On the one hand, we have plausible evidence that the isolation of working from home may reduce mental wellness. On the other, we have the mountain of data showing that people overwhelmingly prefer having that opportunity to work from home.

The obvious way to reconcile those facts is to concede that, as I noted in my opening, people are idiots. To put it more politely, there is a lot of evidence that we are not very good at figuring out what will make us happy (which would also account for the inexplicable persistence of frozen yogurt). Some of the starkest evidence for this comes in our attitudes towards money.

Now, I like money; I think most people do. Quite a lot of my time is spent doing things that will get me more money. It plays an important role in my decisions. I’m pretty sure I’m not alone in that. On some level this is fully rational. My economist colleagues would tell me that I can use money to purchase whatever it is that would make me happiest. But if you look at the research on happiness, it suggests that paying so much attention to money is a big mistake.

That’s not to say that more money won’t make us happier, on average. It does. It’s just that the effects that in studies of life satisfaction are really, really small. Imagine that you doubled your income. How much happier do you think it would make you with your life? You would think it would have a big effect; yet one of the best studies on the subject estimates that doubling your income would, on average, raise your happiness by less than 0.1 of a standard deviation. That’s a really small number, equivalent to going from the 50th percentile in the distribution of happy people to the 54th percentile. Is it an improvement? Yes. Is it likely to be a noticeable improvement? No.

Hence my question: would you take a 50% pay cut to have better colleagues? I have a hard time imagining that I would. I suspect that most of you wouldn’t. But the research on life satisfaction makes this look like a no-brainer. The people we work with have a big impact on our happiness; our income doesn’t.

My Takeaway

I don’t want to take a strong stand on whether remote work is good or bad for you based on just this one study. But the results are consistent with other research suggesting large gaps between the things we tend to prioritize in our decision making versus the things that actually make us happy. That makes me wonder where else we are making these mistakes, and what we can do about them. Can greater introspection help us to reprioritize? Can we break our fixation on the things that are easy to count (dollars earned; minutes spent on the commute) and learn to better appreciate the more nebulous rewards in life (time spent with good colleagues)? I’m not sure. But if we want to make decisions that will help us live a more rewarding life, finding ways to reevaluate our decision-making might be helpful.

Working in different countries can certainly be psychologically enriching. I’ve held jobs on three continents over the course of my career; including shorter visits of 2-3 months, I’ve worked in five countries. I’ve thoroughly enjoyed each move. But are international moves professionally rewarding? Do they advance our careers, make us more likely to get the jobs we want, or help us to earn more money? Is international experience necessary to advance in the corporate world? The evidence is surprisingly complicated.

Subscribe now

The Case For International Mobility…

The case for international mobility is pretty self-evident. We work in a global economy, and many of the largest and most successful employers are multinational companies. If you aspire to a senior role in one of those companies, it behooves you to establish your international chops through stints in other countries. Working in other countries teaches you the importance of international differences in beliefs, institutions, and working styles. Somebody who has only ever worked in one country will often assume that what works here will automatically work elsewhere. Someone who has experienced the culture shock of working abroad will be much more thoughtful about transferring lessons across geographies.

Evidence for the value of international experience also comes from the resumes of some of the world’s most successful CEOs. When you look at the CEOs of America’s most valuable companies, a truly surprising number are immigrants; Many of the rest, people like John Furner at Walmart and Dave Ricks at Eli Lilly, worked international assignments on their way to the top. If you’re ambitious for corporate success, time spent overseas looks like a sensible investment.

… And the Case Against

Conventional wisdom might argue in favor of international assignments, but academic research has often painted them in a more negative light. Some surveys find that people who have held international assignments believe that those assignments adversely affected their careers. Interview studies emphasize that “expatriate” executives who return to their home countries are often dissatisfied with their career progress, frequently quitting their jobs soon after their return. Those studies highlight how the return home can be very difficult to manage. Being away from their home country often takes people out of the networks that will help them find their next role. With less visibility, it is harder for them to seek out the best opportunities for further advancement. To quote one review of the literature:

“It is not atypical for returning expatriates be put into a ‘holding pattern’ relative to the new position upon return to the domestic organization. ‘Out of sight out of mind’ is a common problem, so that expatriate managers find themselves without an appropriate position on return. Finding a position that is a reasonable career step for expatriates may take up to six months, leaving them questioning the wisdom of the expatriate assignment. At the same time, cohorts of expatriates appear to be ahead of them, because they have ‘been around’ when positions came open in the organization.”

Employers may also question the value of the experience acquired overseas. Sure, you might have built a new appreciation of how to do business in a foreign market. But how is that going to help you be a better executive in your home market? There’s a risk that the skills and relationships developed overseas end up being less valuable than you expect.

Lessons from MBA alums

In a recent study, we tried to gather some more quantitative evidence to assessthe relative strengths of the costs and benefits of international mobility. Along with my friends and colleagues Martine Haas, Isabel Fernandez-Mateo and Giovanni Capponi, we surveyed alums of one of the top European business schools, asking them about the different jobs that they had held post-MBA. We used this data to explore whether people who moved internationally ended up being paid more. The results highlight the complexity surrounding international assignments.

We did find that people who had spent time working in different countries received bigger pay packets. It turns out, though, that this was because people who were already well-paid were more likely to make those international moves. High-fliers, it seems, are more likely to be asked to take on an international assignment.

When we looked instead at how people’s pay changed from one job to the next, we found that moving internationally was actually associated with slower subsequent pay growth. On average, people seemed to advance more slowly in their careers following an international move. All of the complaints that we had read in previous studies about international moves holding people back were reflected in how our alums were being paid.

There was, though, a very small group of people who seemed to benefit from those international moves. Although the vast majority of the alums in our sample made, zero, one, or at most two international moves, there was a small group of people who made even more moves. For the elite few that made four international moves or more, those moves were associated with increased pay. I’ve reproduced the plot of the effects below for people who like to see things visually (for the real nerds, the y axis is predictions of log pay; that’s just the way we roll).

The vertical lines in the chart are error bars around our estimates. They grow as the number of moves increases because there are fewer and fewer people to base our estimates on. There are too few people with four or five moves to get a really precise estimate of how they are paid, but the inflection point around two moves is pretty robust. Beyond that point, further international moves seem more likely to increase pay.

It’s hard to be completely sure what is going on here, but our interpretation is that you need a lot of international experience to gain value as a global manager. Work in two countries, and you know those two countries. That experience may not be enough to position you as a seasoned global executive. It takes multiple international stints to be valued for your global perspective. Some of the people that we spoke to echoed this idea. One of them told us:

“People in international companies are looking for a track record of success in more than one country, so just getting lucky and turning around a company in one country wouldn’t necessarily mean that you’d be successful in another one … Being successful in three or four different countries definitely proves it’s not a fluke. You prove that you can adapt, that you can deliver a performance and contribute to different scenarios and different successes.”

My Takeaways

The research on international mobility has tempered the advice that I give my students. I tell them that working overseas can be immensely psychologically rewarding. I thoroughly enjoy spending time in a new country. I love the new perspectives that it gives. But is it professionally and financially rewarding? That’s less clear.

International moves are a part of many successful careers. The better that you are doing at work, the more likely you are to be asked to take a role in a different country. You shouldn’t assume, though, that your international experience will be valued when you return home, or that your employer will work hard to find the right role for you to return to.

If you’re planning to return home at some point, you need to network relentlessly to remain visible in your home country and ensure you are being considered for the roles you want. You should think carefully about how to tell the story about what you learned overseas, and how it will make you better at your job in a different country. If you are planning to return, you probably shouldn’t stay away for too long.

If you are thinking of spending time working abroad, I would recommend it. But be very aware that it’s not working in a different country that will be the biggest career challenge. It’s figuring out how to continue to advance when you return home again.

Subscribe now

I am curious about this trend. In part, that reflects my personal investment in middle management roles. I teach MBAs, and many of them plan to spend much of their careers working in middle management roles. I also teach a lot of executives who are in middle management. I like these people. I respect the work they do, and I want them to have a future.

In part, though, my curiosity about the future of middle management stems from my own confusion. Yes, AI tools are becoming much more effective, and they are starting to transform a number of jobs. But why middle management in particular? When I think about the kinds of work that middle managers do, it’s not at all obvious that it is particularly automatable by current or even near-future AI tools.

Middle Management is Not an Obvious Candidate for Automation

I decided to quantify my confusion by going to some data. Back in 2023, my colleague Daniel Rock worked with co-authors at OpenAI to assess how different occupations might be affected by LLMs. Their highly influential measure estimates the overall share of tasks in each occupation that could be substantially sped up using AI. I used their estimates to compare the AI exposure of middle management jobs compared to other roles in the tech industry.

The below chart maps the biggest occupations in the software publishing industry (since the charge against middle management seems to be led by tech, I figured I would start there). The Y axis charts the size of the occupation (number of employees in the software industry). AI exposure measure is on the X axis. I’ve colored the management occupations in red to distinguish them from other roles.

Data from OEWS and Eloundou et al, 2023

The graph backs up my intuition. The management occupations are among the least exposed to AI in the entire industry. If you were going to build a list of jobs in the tech industry that are most likely to be affected by AI, management occupations would be at the bottom of your list - not near the top, as companies are suggesting.

So what’s going on? It’s possible that the AI exposure estimates are wrong. Our attempts to estimate occupational exposure to AI are pretty rudimentary, and this may be one of the places where they are failing. That said, I think that the AI exposure measures pass the smell test.

Think about what middle managers actually, you know, do. I’m a huge fan of Michael Watkins’ book “Your Next Move.” In it, he lays out some of the critical competencies in different roles on the managerial ladder. The tasks that he highlights revolve around include goal setting, delegation, selecting and developing subordinates, developing strategy, and managing tradeoffs.

The O*NET database, a Department of Labor resource that catalogs the key activities and requirements of 1,016 occupations, agrees. I used the database to identify the tasks that are most common among middle managers (specifically “Operations Specialties Managers”) relative to the other common occupations in Tech. The top five were as follows:

Maybe you could make an argument that AI can monitor resources effectively. But beyond that, these don’t seem like terribly good tasks for Claude. So why is it middle managers who seem to be taking the fall as firms go all in on AI?

We’ve Seen This Movie Before

Consider the following lines:

“We don’t need the questioners and checkers, the nitpickers who bog down the process… Today, each staff person has to ask, ‘How do I add value? How do I make people on the line more effective and competitive?’”

“We used to have department managers, sector managers, subsector managers, unit managers, supervisors. We’re driving those titles out… We used to go from the CEO to sectors to groups to businesses. Now we go from the CEO to businesses. There is nothing else. Zero.”

Give or take a couple of word choices, these quotes could have come from another Silicon Valley titan lauding their latest restructuring. They don’t, though; they come from the late Jack Welch, describing his 1980s transformation of GE. Nor was Welch alone in his desire to take out middle management: “delayering” - restructuring in order to eliminate layers of middle management - was a major corporate trend during the end of the twentieth century and the beginning of this one. Companies have, it seems, been looking to cut back on middle managers since long before the advent of AI.

The current desire to pare back the ranks of middle managers therefore seems to be part of a much longer-term trend of companies dialing back bureaucracy and hierarchy for the information age. But even that story is a little difficult to square with the facts. Because we may have spent thirty plus years talking about cutting back middle management, but the statistics don’t seem to have been paying attention. As a number of researchers have pointed out, the proportion of managers in the workforce has actually been growing over time.

It’s difficult to get consistent data going far enough back, but the American Community Survey gives good information on the sizes of occupations going back to 2004. In the chart below, I’ve plotted the size of the Operations and Specialties Managers (again, the group that best fits the middle management definition) (left axis) and the overall workforce (right axis). We see that middle management has grown at an even faster proportional rate than the overall workforce over the last 20 years.

Data from the American Community Survey

So I don’t think that middle managers are under particular threat because of AI. Instead, middle managers seem to be in the middle of some kind of forever war, where organizations continually declare the need to cut back on those middle management ranks, all while the actual numbers of middle managers continue to grow relentlessly. And that seems like an interesting dynamic to understand in its own right.

The Parkinsonian Dialectic

One of the most influential perspectives on the growth of middle management is more than sixty years old. Back in the 1950s, Cyril Northcote Parksinson noticed a stark disconnect between the growth rate of the British Navy’s fleet versus the civil servants managing it. In 1914, Britain had 62 capital ships, and 2,000 officials overseeing the navy. Fourteen years later, the number of capital ships had shrunk to just 20. The number of officials had, by contrast, grown to 3,569. Through the twentieth century, the number of Admiralty officials continued to grow as the number of ships shrank. Parkinson wrote about this phenomenon in a wildly popular satirical article and book titled “Parkinson’s Law.”(if you haven’t read it, you should. After you’ve finished reading this post, obviously). At the heart of Parkinson’s description were two axioms that he claimed accounted for the tendency of bureaucracies to expand:

Factor I - An official wants to multiply subordinates, not rivals; and

Factor II - Officials make work for each other

Parkinson argued that managers are often overworked and will therefore lobby to hire subordinates to help them out. Unfortunately, those subordinates do not actually lessen the load. Instead, they just make work for one another as they add to and review one another’s work. As Parkinson explained, the end result of that hiring can look something like this:

“Seven officials are now doing what one did before... For these seven make so much work for each other that all are fully occupied and A [the original official]is actually working harder than ever. An incoming document may well come before each of them in turn. Official E decides that it falls within the province of F, who places a draft reply before C, who amends it drastically before consulting D, who asks G to deal with it. But G goes on leave at this point, handing the file over to H, who drafts a minute, which is signed by D and returned to C, who revises his draft accordingly and lays the new version before A…

A is a conscientious man. Beset as he is with problems created by his colleagues for themselves and for him—created by the mere fact of these officials’ existence—he is not the man to shirk his duty. He reads through the draft with care, deletes the fussy paragraphs added by C and H and restores the thing back to the form preferred in the first instance by the able (if quarrelsome) F. He corrects the English—none of these young men can write grammatically—and finally produces the same reply he would have written if officials C to H had never been born. Far more people have taken far longer to produce the same result. No one has been idle. All have done their best.”

I’m not sure if Welch ever read Parkinson’s article, but he expressed a remarkably similar sentiment when he noted that: “We were hiring people just to read reports of people who had been hired to write reports”

As I keep saying, I’m a fan of middle management. I believe that the work that they do is critical to organizations. Strategies need to be developed and adapted, priorities set, units aligned, and people managed. Compliance needs to be monitored. As LT Zhang has noted, modern managers play a central role in shaping the collaboration that allows large organizations to work. The problem, though, is that those processes often identify new organizational needs, creating new work to be done.

Parkinson’s fundamental observation was that work is elastic, so that there may be little relationship between the work to be done and the size of the staff to do it. And because the output of middle managers is particularly hard to measure rigorously, coming as it does in the form of strategies enhanced, coordination improved, and goals clarified, it is particularly hard for organizations to figure out what the “right” number of middle managers should be. As new managers create new work for new managers to do, Parkinson’s law dictates an inexorable tendency for middle management to continue to expand.

I think Parkinson’s law provides a credible explanation for both the continued growth in middle management and for the hostility towards it. At the most senior levels, organizations are conscious of the tendency of middle managers to find work to justify hiring more middle managers. Each individual decision looks sensible, meeting a need that has been clearly identified. And so the ranks of middle managers begin to grow. But sooner or later, the senior executives will look in the mirror and see the organization looking particularly unkempt. That’s when they decide it needs a haircut. Middle managers are laid off and the virtues of leanness and frugality are trumpeted. And then the cycle starts again.

Implications for Middle Managers

So what does this all mean for careers, and for careers in middle management in particular? I disagree with the commentators who say that middle management jobs are under particular threat from AI. I think that those jobs will change, but so will most white-collar jobs. I don’t buy the case that middle management roles will suffer particularly. Indeed, the need for judgment and strategic thinking that lies at the heart of middle management should grow. I think that middle management will continue be a very viable career path into the foreseeable future (albeit that the foreseeable future keeps getting closer and closer).

Looking at the tradition of middle management layoffs in a longer run context highlights that these have always been somewhat precarious roles, though. The Parkinsonian dialectic adds volatility into the world of middle management roles as the managerial ranks go through their cycles of expansion and contraction. Middle management roles are often being created as work continues to expand. But because the output of these roles is harder to measure than others (despite being no less valuable) these also tend to be roles that face regular trimming. Those in middle management roles will always need to be nimble to ensure that their career aren’t derailed in one of those periodic restructuring.

That’s my reading of the evidence on middle management layoffs. Think I’m wrong? Please let me know!

Leave a comment

It’s clear that there are several factors reshaping the consulting industry today. But when I look at how it’s changing, I keep coming back to a thirty-five-year-old book on the development of the legal industry. If its thesis is correct, consulting firms - and our graduates - may be facing some profound, structural challenges in the coming years.

Subscribe now

Lessons from Law Firms

In their book “Tournament of Lawyers,” Marc Galanter and Thomas Palay sought to resolve a puzzle: why had corporate law firms grown so rapidly in the second half of the twentieth century? Back in the 1950s, a leading corporate law firm might have employed 10 or 20 lawyers. By the late 1980s, that number had grown to anywhere between 150 and 300 lawyers. Why? Conventional explanations revolved around the growth of the law firms’ clients and the increasing complexity of their legal needs. Galanter and Palay proposed a provocative alternative: perhaps law firms kept growing because it was the only way they could manage their associates effectively.

Making money in corporate law is a matter of effectively combining senior and junior lawyers. Ultimately, work is sold by senior lawyers - partners - whose expertise and reputation enable them to bring in new clients. Those partners then pass off as much of the work as they can to more junior professionals (associates), freeing up their own time to sell more work. Those associates have the legal training to do much of that work competently, although they lack the reputation or depth of expertise to sell work on their own. Their efforts allow the partners to more effectively “leverage” their expertise, driving up firm profitability.

The obvious challenge for law firms, Galanter and Palay argued, is persuading the associates to put in the long hours and hard work necessary to make the firm a success. The partners’ reputation depends on the work of those associates; the firm needs the associates to give of their very best. The solution that law firms hit on in the early twentieth century was what is sometimes described as the “Cravath system“. Associates were told that if they worked hard and demonstrated excellence, then they had a chance of being promoted to partner themselves after a period of 6-10 years. Those associates that didn’t make the grade would be asked to leave. That “up or out” system induced associates to work hard in the hopes of one day reaping the rewards of partnership.

There’s no doubt that this system helped to align everyone’s incentives; corporate lawyers work phenomenally hard. But it also meant that over time, the law firms needed to promote increasing numbers of their associates to partner. Each time the firm gained a new partner, of course, it would need to hire several new associates in order to retain its leverage ratio; and eventually, some of those associates might need to be promoted to partner in order to fulfill the implicit contract. Embedded in this incentive system, therefore, was a requirement for continual organizational growth. Just like sharks have to keep swimming to stay alive (I’m told), so law firms needed to keep growing if they were going to keep their personnel system working. It was this dynamic, that Galanter and Palay held responsible for the growth of the large law firm.

Is this really why law firms grew so much? To be honest, I’m not entirely sure. But I do think that Galanter and Palay identified a really important issue with the up-or-out partnership model: when promotion to partnership is an important part of how you motivate people, you had better keep minting new partners. It is also striking how much law firms have continued to grow since Galanter and Palay wrote their book. Way back in 1990, they were trying to explain why Latham and Watkins had grown from a handful of lawyers in 1950 to over 250 in the 1980s. Today, Wikipedia reports that Latham and Watkins has over 3,500 lawyers.

The Growth Imperative in Consulting Firms

Although the “Cravath system” of up or out promotion to partner originated in the legal industry, it is also central to the culture of consulting firms. Maybe that’s because it’s the best way to manage professional services; or maybe it’s a historical accident. Marvin Bower, an early employee at McKinsey, did more to shape the consulting industry than pretty much anyone else. He had also worked as a lawyer before moving into consulting, and drew on that legal experience in how he shaped the firm.

And consulting firms have also followed the path that Galanter and Palay laid out at the beginning of the 1990s, demonstrating consistent, exponential growth over the last three decades. It’s not easy to get reliable figures on consulting firm employment as many of the biggest firms are private companies. Accenture, though is publicly traded, allowing us (well, Gemini) to plot out its growth rate. Since it’s IPO in 2001, the firm has grown something like ten times, from 75,000 to almost 800,000.

The evidence is that the other big consulting firms have also grown rapidly over the same time period. Take Deloitte, for example. In 2000, Deloitte had 30,000 employees in the US. By 2025, that number had risen five-fold, to 154,000. Back at the end of the 1990s, McKinsey had 7,700 staff. Today, they have 38,000. Their competitors have delivered similar growth rates.

Now this rapid growth might just reflect client’s dawning realization that the route to untold riches was hiring more consultants. Maybe. Or perhaps the growing complexity of business meant that firms had more need of outside counsel. It’s interesting, though, to note that total industry revenues for management consulting don’t seem to have grown any faster than the overall economy. Much of the growth of large consulting firms has instead been through expansion into related areas.

So no, we can’t be sure that the exponential growth of consulting firms was a direct result of the up-or-out system. But I, for one, am struck by how closely those firms’ development has tracked the arguments that Galanter and Palay put forth nearly forty years ago.

Exponential Growth and the Future of Consulting

Okay, so “Tournament of Lawyers” made a big impact on me when I read it. But what does it mean for the job prospects of my students, or the role that consulting will continue to play in people’s careers? First off, I think that it is valuable to see consulting’s recent struggles within the context of their spectacular, multi-decade expansion. Consulting firms have been growing exponentially for a long time, and there’s one thing that we know about exponential growth: it can’t continue forever. That doesn’t mean that consulting firms’ growth has to slow down now, or next year, or even within the next ten years. But I’m pretty sure that our future economy won’t consist solely of consulting firms selling services to one another. Sooner or later, consulting firm growth has to converge on the overall economic growth rate.

The second lesson is that that slowdown is going to be surprisingly painful for many of the consulting firms. For generations, those firms have offered young employees a clear bargain: work hard, and you have a decent shot at becoming a partner. Not everybody made it, of course, but enough people got promoted that the up-or-out system didn’t feel too punitive. That system relied on (and drove) continued organizational growth, though. If consulting firms stop growing, each new associate’s odds of making it to partner are going to shrink, and a demanding culture risks morphing into something more akin to the Hunger Games. I personally doubt that that would be a sustainable model - the rewards of being a partner are unlikely to be attractive enough for talented graduates to want to run the kind of gauntlet that promotion tournaments in a steady-state firm would imply.

So how might careers in consulting adjust to slower growth? I’m not sure that there’s a clear playbook. One option is to lower leverage ratios (the ratio of associates to partners), so that fewer associates compete for each partnership. We see such low ratios in academia, and they allow us to sustain an up-or-out system with very little growth. It may also be that lower leverage ratios are the future of consulting anyway, with commentators suggesting that AI will replace much of the work currently done by associates. It’s worth noting, though, that such a reduction in leverage will require a major change in consulting firms’ economic models, which currently rely on billing out large number of junior employees at high margins.

A second alternative is to move more and more people off the partner track, so that most of the associate work is being done by people who are not expecting a promotion. This alleviates growth pressure while maintaining leverage. It also looks to be the direction that a lot of professional service firms are going in. The obvious challenge is that these off-partner tracks are a lot less attractive to ambitious young graduates. I doubt that consulting firms would be able to continue hiring the cream of the crop if they aren’t able to also offer a credible path to the top. Without the promise of a shot at partnership in the future, employees may also demand more money today.

Most likely, consulting firms that face slowing growth will opt for a little of column A and a little of column B. Both approaches are likely to involve substantial, often wrenching, cultural change. The one thing that’s clear, though, is that firms can’t continue indefinitely with an employment model that requires endless, exponential growth.

On Strategy and Careers

I feel very invested in what happens in consulting, both because of my own experiences in the industry and because it is where so many of my students end up. But even if consulting is not your thing, I think that there are still lessons to take away from the tensions facing the industry.

We often think about careers as being downstream from organizational strategy - companies figure out their business model, which then shapes the available jobs and the career paths between them. Sometimes, though, the relationship flows the other way, as a desire to meet commitments to employees shapes firm goals. The more that firms rely on promotions as a means of rewarding their employees, the more pressures that they face to keep growing and provide more opportunities to their employees.

And, of course, the relationships run in the other direction. The traditional career model within consulting works well when there is a lot of scope to grow, but will come under stress if and when growth slows down in a meaningful way. The up or out model makes this link between strategy and careers particularly visible. Still, shifting industry fortunes are likely to reshape careers in a whole lot of other sectors over the next decade.

Leave a comment

It’s graduation season. Every morning for the last couple of weeks, I’ve been passing students dressed in their graduation regalia, taking photos in front of college landmarks. It’s a major milestone in their lives; for most of them it marks an end to education and the point at which they enter the labor market full time. So what does that labor market have in store for them?

Subscribe now

To be honest, the news isn’t great. Although overall unemployment within the US remains low by historical standards, unemployment among college graduates has been trending up for the last four years, and is weirdly high compared to other groups (the trends can be seen clearly in this chart, using data taken from the Fed).

As you can see, the 5.6% rate of recent graduate unemployment is still some way below the 7.8% peak achieved in early 2011, and way below the 2020 COVID spike. Pandemics aside, though, recent graduate unemployment is at it’s highest level since 2013, and it seems to be trending in the wrong direction. There’s no doubt that recent college graduates are finding it harder to get a job than they have for most of the last decade

It’s got something to do with your Vorsprung Durch Technik, you know…

Why are recent college graduates having such a tough time compared to other adults? A lot of people think it is the first sign of an impending AI jobs apocalypse. New AI tools are designed to replace the kinds of white collar work that college graduates usually go into, and LLMs are particularly effective at the kinds of routine tasks that those recent grads are often asked to do, things like summarizing documents, editing Powerpoints, and writing first drafts. Why would companies hire recent grads, when they could build an agent instead? Early studies provide some support for these AI stories, showing that the slowdown in employment of young workers has been concentrated among the kinds of occupations that are more exposed to AI, and in the companies that have been first to adopt it.

I think that the evidence that AI has made life harder for recent graduates is looking increasingly strong. I’m less sure, though, that it is the whole story. If you look at the unemployment trends posted above, the growth in graduate unemployment seems to be part of a much longer trend. Between 1990 and 2010, recent graduate unemployment was reliably lower than overall unemployment (the grey line in the chart). Yet, recent graduate unemployment started to increase relative to overall unemployment in the mid 2010s. The gap between them has declined steadily, and then reversed, over the course of a decade. This process seems to have started well before anybody was talking about LLMs.

Second, as Ben Casselman at the New York Times likes to point out, many of the biggest post ChatGPT employment declines have been in industries that seem least exposed to AI, industries like household appliance manufacturing and furniture and home furnishings. It’s clear that other things are also going on in the economy that are making life hard for employees.

In particular, it looks like many of the challenges facing current grads are grounded in the more prosaic dynamics of the business cycle. Although the overall unemployment rate remains relatively low, data from the US government Job Openings and Labor Turnover Survey (JOLTS) shows that employers just aren’t hiring across all jobs.

Now some of the hiring slowdown could also reflect companies’ reluctance to hire workers when they think that they might soon be replacing them with AI. But there are many other reasons that hiring is slow. Some employers may have been suffering from a prolonged hangover, following their post-pandemic hiring binges, and don’t want to add yet more employees. The massive uncertainty of the past few years, driven by interest rate rises, tariffs, wars, more wars, inflation, and so on have also left many companies reluctant to take on additional costs by expanding employment. All of these pressures mean that hiring rates in recent months have been at their lowest level since 2012, a time when we were still recovering from the great recession.

Now a low-hiring economy isn’t fun for anyone: it makes it harder for employees to find better jobs, and reduces pressure on employers to raise pay and improve working conditions. But a low-hiring economy is particularly bad for new graduates. They don’t have the option of staying where they are; they have to find a new job. And that’s particularly challenging when nobody wants to hire.

That’s the bad news. The good news is that business cycles are, well, cyclical. And that means that hiring may well pick up soon. The after effects of the post-pandemic hiring binge are surely fading. If we can avoid other major shocks to the economy (yes, that’s a spectacularly big if), employers are likely to want to start hiring again. Certainly, a recent survey by the National Association of Colleges and Employers found that 3 times as many employers expect to increase hiring of college graduates for the class of 2026 as expected to decrease hires. So there is hope. But as of today, graduates should be prepared to enter a tough job market.

Some Advice for New Graduates

What are the implications of graduating into a tough job market? How should it change new grads’ strategies? My main advice would be to get ready to work harder to find a job in the short-term, but also see their job search as a much more long-term process than it might have been.

In the short-term, new graduates will need to work hard at all of the things that are important to job searching, figuring out what jobs are out there, and carefully researching each application. A particular challenge in today’s labor market is that AI makes it so easy to submit tailored applications that employers are drowning in candidates. Making your application stand out, even just making sure it is read, is getting harder and harder. Probably the most effective way to stand out from the pile of other applicants is through networking. I suspect that there are higher benefits than ever before to figuring out who you know who knows somebody working where you want to work. This is a good time to look for fellow alums who might work at companies you are targeting or to try to get somebody to recommend you personally for the role. Those personal connections may well be the thing that makes the difference.

In the longer-term, new graduates will need to be more intentional about moving up in the labor market compared to those who graduated when life was easier. Some pretty grim economic studies have found that labor market conditions at the time when people graduate can have long-term effects on their careers. Not only do they find that people who graduate at times of high unemployment earn less in their first job; they also show that those unlucky graduates can continue to earn less many years later, even after the economy improves. Those costs of graduating at the wrong times are particularly large for people who graduate from colleges and majors that tend to pay less anyway.

Why does graduating into a difficult job market have such persistent effects? The lack of opportunities forces graduates to take worse jobs than they otherwise would have done, ending up at firms that pay less than their peers and working in lower paying occupations. Those jobs then risk putting them on a permanently lower paid trajectory, as they build careers in those poorly paying firms and occupations. In order to catch up with where they should have been, people need to move back towards higher paying employers and occupations as the economy improves. Those graduates who make that leap recover quickly. Those that cannot see their wages continue to stagnate.

Now, it’s worth noting that the studies I’m referring to looked at graduating in a recession, and we’re not in a recession today. But it is a tough jobs market for recent college graduates, and that may well mean taking a job that is less rewarding than those graduates hoped for or are qualified for. As a result, I think that it is important to take the lessons of this prior research on board.

If you, or somebody that you know has been struggling to find a good job and has had to settle for something worse than they expected, I have two pieces of advice. My first piece of advice is “don’t panic!” (to be fair, that’s my first piece of advice in most situations). This happens when job markets are tough. Think of it as a temporary setback, not a permanent state of affairs.

My second piece of advice, though, is to keep up the search. You’ve not got what you wanted today, but the job market will turn around. Make sure you are ready when it does. Keep networking. Keep learning about the jobs that really interest you. Start building the skills, in and out of work, that will help position you for something better as new opportunities become available. Unfortunately, launching your career effectively will take longer than it should. But if you put that work in, your initial struggles will be no more than a temporary setback.

The more I study careers and labor markets, the harder it is to escape the conclusion that life is desperately unfair. Yes, what we achieve partly affects our abilities, talents, and hard work. But our careers are also shaped by a myriad of factors outside our control. It’s not fair that graduating into a tough job market should put people at an unfair disadvantage compared to prior cohorts. But that’s the hand that our freshly minted grads are being dealt. How they play it will shape the rewards that they can reap over the early years of their career.

In this newsletter, I plan to lay out the other things that we can learn from research about how careers work today, and what it means for us and our employers. Please subscribe to learn more!