A foundational concept in twentieth century sociology is the cohort, a group of people starting something (such as life, employment or college) at the same time. Our view of the human world is based on the idea of cohort-based groups marching (theoretically) in lock-step through life. From grades in school to leagues of increasing skill-levels in sports to career paths within corporations, our world is full of groups navigating the world to the sound of the same drumbeat. If all the world is a stage, the larger drama is a parade of cohorts marching, dancing or straggling along drunkenly, but rarely breaking ranks.
Something very interesting is going on with cohorts today. The parade is ending.
Cohorts used to last as coherent social units all the way from high school to retirement. Now they fall apart within a few years of college. Different patterns of organization take over very quickly. To understand what’s going on, it’s useful to think in terms of the metaphor of racing and what happens to different start-time cohorts. Last year, I ran a series of five 5k races, and observed some very entertaining sorting effects, which I captured at the time in this sketch of my anecdotal observations (and unkind judgments):
Something very similar to this is happening in the human drama. If you’ll forgive some hyperbole, the global drumbeat is faltering. Cogs and mavericks alike are struggling. Cogs are wandering around wondering what to do with themselves. Mavericks are so used to defining their identities in terms of breaking ranks and following the beat of a different drum, that they too are struggling with a world that is not framed by a parade. Increasingly, there is nothing to rebel against.
The end of the parade has profound implications. Cohorts are ubiquitous in social analysis, and in both natural and designed patterns of social organization. When cohort-based social organization fails, civilization itself fails in a sense. Our world is powered by an education system pumping out a new cohort of fresh high-school and college graduates into an impedance-matched global labor economy every year, which then distributes them through a branching network of increasingly specialized career paths into the wider world. What John Hagel and his co-authors call a push economy. The tempo of this economy is dominated by an annual rhythm beating in grade schools.
So to claim that cohort-based social organization is coming to an end is almost like claiming the world is ending. It’s a rather dramatic claim, but there is compelling evidence.
Age Distribution Sprawl
Birth-year cohorts are the most fundamental kind (and strongly drive the composition of other derived cohorts such as graduating classes), so one of the most obvious symptoms is rapidly spreading age distributions in social and work groups, especially online.
My father’s friends throughout his working life, and now into his retirement, have been his cohort members. His best friends graduated high school and college with him, moved with him to the same city a thousand miles away, joined the same company, rose through the ranks together, and have now retired together in their hometown. If you think about it, that was an incredible case of 50 years of marching in lock-step through life. The age distribution in what he considered his peer group has never been wider than a few years.
Me? Most of my social capital is online, resident in groups where the age distribution curve sprawls untidily from 22 to 70. All the work teams I’ve been part of have had similar age-distribution sprawl. And these are not “cross functional” groups. Age in these groups is not a proxy for “seniority,” with attendant privileges. Mentor-mentee relationships do not dominate (such things would be a case of cohort-based groupings sneaking into non-cohort groupings). These are groups where age/birth based cohort identities simply do not matter.
Outside of a couple of college friends, I have no regular contact social or professional dealings with my nominal cohort members. My blog readers seem to range from college students to retirees. And I know I am not alone here. Anyone with a significant online life at the intersection of many illegible Google+ circles is likely to have such a profile.
Categories like Class of 97 or I grew up in the 80s mean far less today.
Weakened Life-Stage Effects
Life-stage effects on lifestyle are rapidly diminishing. This surprises a lot of people who assume that the biological roots of sociological cohorts are very strong. Certainly, there are life-stage effects, but we often forget the extent to which they are arbitrary culture-dependent constructs. Adolescence is only about a century old. It emerged alongside universal high school education in early twentieth century America, and was basically “invented” by psychologist and educator Stanley Hall in 1904, to describe the new 6-7 year epoch between puberty and the start of adult life, which did not exist previously.
Today’s unsettled labor economy, where new workers take up to a decade to find their niche, is creating another new life stage that is being called emerging adulthood. Like adolescence, it is also a discovery-invention rather than a natural construct, this time due to Jeffrey Arnett, in 2000. But it is a much weaker and less universal candidate than adolescence (for instance, I’d wager that it is more a Blue American life stage than a Red American one; see this piece)
What about female biological clocks? Again, child-bearing practices only loosely relate to biology. The last link mentions differences between Red and Blue America. If you look around the world, the differences are even more dramatic. In the West, more and more women wait till they are nearly 40 to have children, whereas in Bali, most traditional women are grandmothers by 36 (as I learned from an Australian expat in Bali, who remarked that the children in a village she frequented called her “grandmother”).
Not only are life stages getting more arbitrary and culturally relative, they are getting weaker as explanatory constructs even where they do apply. Parenthood for instance, is less useful a guide to anything these days. It used to be that having a kid would move you from the singles/childless “scene” to the parental scene. Not anymore. My groups seem to have as many people with kids as without. There isn’t even a strong age correlation there: I have younger friends with kids and mortgages, and older ones who are unmarried and childless.
Unpunctuated Global Affairs
Finally, consider large-scale cohort-based thinking. Demographics researchers like to organize their thinking in terms of generations: Silents, Boomers, X, Millenials. and demographic-trend based (Boomers/X/Millenial) cohorts. We tend to forget how arbitrary these are. Here, the precipitating events tend to be large-scale global disruptions that serve as sources of punctuation marks.
The Depression created the Silents. World War II created the “Greatest Generation” and (as an after-effect) the Boomers. But the lack of such major global cohort-synchronizing events since World War II means that the cohorts that came after are far more weakly defined, and much less global. Generation X is a weakly defined type compared to Boomers, but people at least agree about the name and the rough boundaries (~ 1963 – 1980). Beyond that, people can’t even agree on names and key characteristics, let alone temporal boundaries (I’ve seen everything from Generation Y to Millenials to Netgen, and significant boundaries being proposed anywhere between 1980 and 2000).
The reason is obvious. Before World War II, there used to be at least one huge punctuation mark every couple of decades. In America alone, going backwards, we have: World War II, the Great Depression, World War I, the Robber-Baron/Gilded Age, the settlement of the West and the Civil War.
More recent events, such as the fall of the Berlin Wall, the rise of the Internet and 9/11 simply don’t compare in terms of sudden, life-altering impact (the same is not true for other parts of the world of course; the end of the Cold War was a much bigger deal for Russians, but the parade is ending nearly everywhere).
When the Music Stops
Cohort-based organization is central to the economy. It solves one of the main coordination problems that the money economy does not: allowing central planners to manage the supply and demand of talent collectives. Our world is designed around the assumption that every year a certain group of free agents with varied skills will be released at roughly the same time from high schools and colleges. This is the one major opportunity to stock up on high-potential talent with any mix of capabilities you care to list, to fuel years of mix ‘n match experiments in search of high-performance teams.
The labor market for older people is far less efficient simply because appropriate mixes of people are rarely available together. Things have gotten to the point where it is now often cheaper to buy companies or divisions just to get structurally intact high-performance teams, than to assemble them a la carte from the labor markets and hoping for chemistry to emerge.
Look back at the 5k race picture. I don’t know how accurate my observations and inferences are, but I am convinced of a basic point: without external cues and dictatorially imposed penalties for breaking ranks, cohorts cannot remain coherent (there used to be a time, long long ago, when poaching talent from another company’s pipeline was considered unethical, and when external hires for non-entry positions were as rare as eclipses; a few old-school types like Marshall Goldsmith still swear by internal promotions, but most people have given up).
When the music stops, cohorts dissolve.
What is interesting is not the specific set of new groupings that emerge but the fact that there is a variety of grouping types, and an unpredictable array of variables involved, as suggested by the race graphic. Sorting can happen based on natural personalities, psychological condition, life stage, looks, domestic situation, shared goals, or physical condition.
The same thing is happening in the economy at large, but in a vastly amplified form. People who look for a single new organizing principle to replace cohort-based organization make a fundamental mistake. There is no new drumbeat, just a cacophony. Time — the sorting variable for cohorts — is far too fertile a variable to be replaced by any single substitute variable. It’s not a question of whether personality affinities, interests or social identity takes over. All of them take over. Some groups are defined by who they are, others by what they do, still others based on where they live, how tall they are, what music they prefer or what skills they possess.
Not all at once of course, but one or a few.
Cohorts and geography still matter of course, but far less than they used to, at least within the global creative class. All sorts of other lesser variables start to matter in complex ways. One big result is that self-organization becomes far easier than central planning and talent pipeline management. The talent-collective problem still has to be solved, but when individuals decide to manage their own availability in order to manage collaboration opportunities, the problem gets solved in ways that help individuals and make life harder for long-range planners. This is what Hagel and his co-authors call a pull model.
(aside: my PhD was about precisely this problem; managing schedules to control the rate of churn in teams, using algorithms that trade off tightness of scheduling against frequency of team formation and breakup opportunities)
When individuals start thinking this way, it makes talent-collective markets hell for organizations to navigate. No longer can you expect to stock up on a decent variety of fresh 22-year-old talent and expect to easily solve middle-management casting problems that require experienced candidates, 10 years down the road. You have to go right back to the uncertain, cohort-churning older labor markets.
There are no easy answers. Even leading-edge organizations like the startup incubator, Y-Combinator, organize their activities based on cohort-thinking. By positioning themselves in the more certain upstream portion of the labor economy, right at the exit doors of colleges, they manage to survive.
Tempo for Groups
People sometimes ask me why I did not devote much attention to collective decision-making in Tempo. Besides the obvious constraints of time and available pages, a major reason was simply the sheer lack of non-historicist source material on group dynamics. Researchers who study the subject are like historians in that regard: their work reveals more about extant organizational forms in their own time than about the fundamental nature and behavior of groups. Late twentieth century scholarship on the subject primarily teaches you about corporations of the era. Medieval writings about intrigues in royal courts teach you more about royal courts than about intrigues.
Those are just the more obvious historicist elements. Cohort-based social organization is one of the more subtle elements: I nearly made the mistake of assuming it to be fundamental rather than historicist (in fact I did make the mistake, but fortunately not in the book. I consider an old post, The Crucible Effect, despite its popularity, to be fatally flawed because it did not adequately consider cohort effects). To get at the collective analogues of the concepts and ideas in the book: shared mental models, grand narratives, group beliefs, collective archetypes and common doctrines, you need to reverse engineer a lot of history to extract general insights. Otherwise you end up with a few anemic bits and pieces like game theory and the behaviors of ant swarms.
The idea of tempo — rhythms, emotions and energy — is as powerful for analyzing collective decision-making as it is for analyzing individual decision-making, but you have to do a lot more work to tease apart extrinsic, historicist rhythms and intrinsic, fundamental rhythms. I sketched out preliminary ideas and open questions towards the end of Tempo, in the chapter titled “The Clockless Clock,” but working them out in detail is going to take a lot more work.
But I think it will be worth it. I find it personally annoying that the subject of group dynamics and collective decision-making basically has to be pretty thoroughly rewritten each time a major historical shift in organizational cultures — such as the ongoing stopping of the parade — happens.
Figuring out a non-historicist approach to collectives that is independent of organizational and historical context is a project I am just getting started on. I hope to include the results in a future edition or sequel to Tempo. Or rather what I am starting to call a “loopquel” project. The study of collective decision-making is neither a prequel, nor a sequel to the study of individual decision-making; it is in a chicken-egg relationship with it.
I’ll be beta testing many such ideas on this blog, starting with this look at cohorts. If collective decision-making interests you, stay tuned.