Originally, I had intended the 9th part to be the one in which I solve this damn thing for good. Unfortunately, as that “9th post” crossed into 12 kiloword territory, I realized that I needed to break it up a bit. Even for me, that’s long. So I had to tear some stuff out and split this “final post” yet again. So here’s the 9th chapter in this ongoing lurid saga. (See: Part 1, Part 2, Part 3, Part 4, Part 5, Part 6, Part 7, Part 8). I am really trying to wrap this fucker up so I can get my life back, but I’m not going to solve it just yet… there are a few more core concepts I must address. Today’s topic, however, is convexity.

What is convexity?

Convexity pertains to the question: which is more important, the difference between excellent work and mediocre work, or that between mediocre work and noncompliance (zero)? For concave work, the latter is more important: just getting the job done is what matters, and qualitative concerns are minimal. For convex work, the difference between excellence and mediocrity is substantial and that between mediocrity and failure is small.

The theoretical basis for this is the logistic function, or “S-curve”, which is convex at its left side (looking exponential at y << 0.5) and concave on the right, as it approaches a horizontal asymptote (saturation point). Model input as a numerical variable i pertaining to resources, skill, talent, or effort. Then model task output as having a maximal yield Y, and the function Y * p(i) where p is a logistic function with range (0, 1) representing the proportion of maximum possible yield that is captured. Now, the inflection point (switch-over from convexity to concavity) is exactly where p(i) = 0.5. Taken in full, this logistic function is neither concave or convex. Yet, for most economic problems, the relevant band of the input range is narrow and is mostly on one side of the inflection point or the other. We can classify tasks as convex or concave based on what we know about the performance of the average.

To get concrete about it, consider exams in most schools. A failing student might be able to answer 60% of the questions right; an average one gets 80%, and the good ones get 90%. That’s a concave world. The questions are easy, and one needs to get almost all of them right to distinguish oneself. On the other hand, a math researcher would be thrilled to solve 50% of the (much harder) problems she confronts. With concave work, the success or completion rates tend to be high because the tasks are easy. With convex work, they’re low because the tasks are hard. What makes convex work worth doing is that, often, the potential yield is much higher. If the task is concave, it’s been commoditized, and it’ll be hard to make a profit by doing it. Even 100% completion will yield only marginal profits. If the work is convex, the most successful players can generate outsized returns. It may not be clear even what the upper limit (the definition of “100%”) is.

Convexity and management

Consider the following payoff structures for two tasks, A and B. A is concave; B is convex, and 0 to 4 represent degrees of success.

| Performance | A Payoff | B Payoff | Q dist | R dist |
-------------------------------------------------------
| 4 (Superb)  |      125 |      500 |  20.0% |   0.0% |
| 3 (Good)    |      120 |      250 |  20.0% |  20.0% |
| 2 (Fair)    |      100 |      100 |  20.0% |  60.0% |
| 1 (Poor)    |       60 |       25 |  20.0% |  20.0% |
| 0 (Awful)   |        0 |        0 |  20.0% |   0.0% |
-------------------------------------------------------


Further, let’s assume there are two management strategies, Q and R. Under Q, the workforce will be uniformly distributed among the five tiers of performance: 20% in each. Under R, 20% each of the workforce will fall into Good and Poor, 60% into the Fair tier, and none into the Superb or Awful tiers. R is variance-reducing managerial strategy. It brings people in toward the middle. The goal, here is to maximize bulk productivity, and we assume we have enough workers that we can use the expected payoff as a proxy for that.

For Job A, which is concave, management strategy Q produces an output-per-worker of 81, while R yields 96. The variance-reducing strategy, R, is the right one, yielding 15 points more. For example, bringing up the worst slackers (from 0 to 60) delivers more benefit than pulling down the top players (from 125 to 120).

For Job B, which is convex, strategy Q gives us an average yield of 165, while R delivers only 105– 60 points less. The variance-reducing strategy fails. We see more of a drop in pulling down the best people (from 500 to 250) than we gain in hauling the laggards (o to 25).

In short, when the work is concave, variance is your enemy and reducing it increases expected yield. When the work is convex, variance is your friend; more risk means more yield. 

The above may seem disconnected from the problems of the MacLeod organization, but it’s not. MacLeod organizations are based on variance-reduction management strategies, which have worked overwhelmingly well over the past 200 years of concave, industrial-era labor. MacLeod Losers naturally desire familiarity, uniformity, and stability. They want variance to be reduced and will give up autonomy to have that. MacLeod Clueless (middle managers) take on the job of reducing variance in conditions for the Losers below them, and reducing volatility in performance for the Sociopaths above. Their job is to homogenize and control, and they do it well. It doesn’t require vision or strategy. MacLeod Sociopaths start out as the “heroic” risk-takers (entrepreneurs) but that caste often evolves (as especially as transplant executives come in) into a cushy rent-seeking class as the organization matures (necessitating the obfuscation enabled by the Clueless and the Effort Thermocline). The Sociopath category itself becomes risk-averse, out of each established individual’s desire to protect organizational position. The result is an organization that is very good at reducing variance and stifling individuality, but incapable of innovation. How do we come back from that?

In the concave world, the failures of the MacLeod organization were tolerable. Businesses didn’t need to generate new ideas. They needed to turn existing, semi-fresh ones into repeatable processes, and motivate large groups of people to carry out difficult but mostly simple functions. Variance-reduction was desired and encouraged. Only in the past few decades, with the industrial era fading and the technological one coming on, has there been a need for business to have in-house creative capacity.

Old-style organizations: the optimization model

The convex/concave discussion above assumes one dimension of input (pertaining to how good an individual is at a job) and one of output (observed productivity). In truth, a more accurate model of an organization’s performance would have a interconnected network of such “S-curve” functions for the relationships between various variables, many of which are hidden. There’d be a few input variables (“business variables”) and the things the company cares about (profit, reputation, organizational health) would be outputs, but most of the cause-and-effect relationships are hidden. Wages affect morale, which affects performance, which affects productivity, which affects the firm’s profits, which is its performance function. With all of the dimensions that could be considered, this function might be very convoluted, and while it is held to exist “platonically” it is not known in its entirety. The actual function relating controllable business variables to performance is illegible (due to hidden variables) and certainly not perfectly concave or convex.

So how does the firm find an optimal solution for a problem it faces?

This gets into an area of math called optimization, and I’m not going to be able to do it justice, so I’ll just address it in a hand-wavy way. First, imagine a two-dimensional space (if only because it’s hard to visualize more) where each point has an associated value, creating a 3-dimensional graph surface. We want to find the “highest point”. If that surface is globally concave, like an inverted bowl, that’s very easy, because there can only be one maximum. We can start from any point and “hill climb”: assess the local gradient, step in the most favorable dimension. We’ll end up at the highest point. However, the more convoluted our surface is, the harder the optimization problem. If we pick a bad starting point on a convoluted surface, we might end up somewhere sub-optimal. Thinking of it in topographical terms, a “hill climb” from most places won’t lead to the top of Mount Everest, but to the neighborhood’s highest hill. In other words, the “starting point” matters if the surface is convoluted.

Real optimization problems usually involve more than two dimensions. It is obviously not the case that organizations perform optimizations over all possible values of all possible business variables (of which there are an infinite number). Additionally, the performance function changes over time. As a metaphor, however, for the profit-maximizing industrial corporation, it’s surprisingly useful. One part of what it must do is pick a good “starting point” for the state of the business, a question of “What should this company be?” That requires non-local insight. Another part is the iterative process of refinement and hill-climbing once that initial point is selected.

This leads to the three-tier organization. People who are needed for commodity labor but not trusted, at all, to affect business variables are mere workers. Managers perform the iterative hill-climb and find the highest point in the neighborhood. In startup terms, they “iterate”. Executives, whose job is to choose starting points, also have the right to make non-local “jumps” in business state if needed. In startup terminology, they “pivot”.

The MacLeod organization gets along well with this computational model of the organization. MacLeod Losers are mediocre in dedication, but that’s fine. That aspect of them is treated as a hidden variable that can be modulated via compensation (carrots) or managerial attention (sticks). In the optimization model, they’re just infrastructure– human resources in the true sense of the word. The fault of MacLeod Clueless is that they aren’t strategic, but they don’t need to be. Since their job is just to climb a hill, they don’t need to worry about non-local “vision” concerns such as whether they’re climbing the right hill. That’s for someone else to worry about. They just assess the local gradient and move in the steepest upward direction. Finally, there are the MacLeod Sociopaths, whose goal is to be strategic and have non-local insight. Being successful at that usually requires a high quality of information, and people don’t get that stuff by following the rules. The source could be illegal (industrial espionage) or chaotic (experimental approaches to social interaction) or merely insubordinate (a programmer learning new technologies on “company time”) but it’s almost always transgressive. The MacLeod Sociopath’s ability to get information confers more benefit, in an executive position, than the negatives associated with that category.

Why the optimization model breaks down

In the model above, there is some finite and well-specified set of business variables. The real world is much more unruly. In truth, there are an infinite number of dimensions. Two things make this more tractable. The first is sparsity. Most dimensions don’t matter. For example, model “product concern” as a vector representing the products that a company might make (1.0 meaning “it’s the only thing we care about”, 0 meaning “not interested at all”). Assume there are 387 trillion possible conceivable products that a firm could create. That’s 387 trillion business variables; 386.99999…+ trillion of those entries are always going to be zero (excepting a major pivot) and can be thrown out of the analysis. Second is aggregation. For personnel, one could have a variable for each of the world’s 7.1 billion people (again, most being zero for ‘not working here’) but most companies just care about a few things, like how many people they employ and how much they cost. Headcount and budget are the important business variables. Whether John A. Smith, 35, of Flint, Michigan is employed at the company (i.e. one of those 7.1 billion personal variables) isn’t that relevant for most values of John A. Smith, so executives need not concern themselves with it.

Even still, modern companies have thousands to millions of business variables that matter to them. That’s more information than a single person can process. Then there is the matter of what variables might matter (unknown unknowns). If the optimization problem were simple, the company would only need one executive to call out starting points, but these information issues mandate a larger team. The computation that the organization exists to perform must be distributed. It can’t fit on one “node” (i.e. one person’s head). That also mandates that this massively high-dimensional optimization problem be broken down as well. (I’m ignoring the reality, which is that most people in business don’t “compute” at all, and that many decisions are made on hunches rather than data.)

As far as many dimensions are separable (that is, they aren’t expected to interact, so the best values for each can be found in separation) the problem can be decomposed by splitting it into subproblems and solving each in isolation. Executives take the most important business variables where it is most likely that non-local jumps will be needed, such as whether to lay off 15% of the workforce. The less important ones (like whether to fire John A. Smith) are tackled by managers. Workers don’t participate in the problem-solving; they’re just machines.

This evolves from an optimization model where business variables and performance functions are presumed to exist platonically, to a distributed agent-based model operated by local problem-solving agents. This is a more accurate model of what actually happens in the corporation, made further amusing by the fact that the agents often have diverging personal objective functions. Centralized computation is no longer the most important force in the company; it’s communication between the nodes (people).

Here’s where MacLeod comes in to the agent-based model. MacLeod Losers consume information and turn it into work. That’s all they’re expected to do, and ideally the only thing that should be coming back is one word: done. MacLeod Clueless furnish information up and down the food chain, non-editorially because they aren’t strategic enough to turn it into power. They tell the Losers what to do, and the Sociopaths what was done, and they aren’t much of a filter. The only information they take in, in general, is what information others want from them. The MacLeod Sociopaths are strategic givers and takers of information, and (having their own agendas) they are selective in what they transmit. Organizations actually need such people in order to protect the top decision-makers from “information overload”. It’s largely the bottom-tier Sociopaths who participate in dimensionality reduction and aggregation, so they’re absolutely vital; however, they make sure to use whatever editorial sway they have toward their own benefit.

Optimization and convexity

The actual performance function of a company, in terms of its business variables, is quite convoluted. It’s generally concave in a neighborhood (enabling managers to find the “local hill”) but its global structure is not, necessitating the non-local jumping afforded to executives. The underlying structure, as I said earlier, is driven by an inordinate number of hidden variables. It might be best thought of as a neural network of S-curve functions (“perceptrons”) wherein there are elements of concavity and convexity, often interacting in strange ways. It’s not possible for anyone to ascertain what a specific organization’s underlying network looks like exactly. The overall relationship between business variables (inputs) and performance (output) is not going to be purely concave or convex. The best one can hope for is a well-chosen initial point in which the neighborhood is concave.

For typical organizations and most people in them, concavity has a lot of nice properties. For one thing, it tends toward fairness. Allocation of resources to varying parties, if the input/output relationships are concave, is likely to favor an “interior” solution– everyone gets some– because marginal returns diminish as the resource is allocated to richer people. If the input/output relationship is convex, resource allocation can favor an “edge” solution where one party gets everything and the rest get none, which tends to exacerbate the “power law” distributions associated with social inequalities: a few (who seem the most capable of turning those resources into value) get much, most get little or nothing. Another benefit of concavity is that performance relative to a standard can be measured. At concave work, the maximum sustainable output is typically well-studied and known, and acceptable error rates can be defined. With convex work, no one knows what’s possible. Once a maximum is established and can be reliably attained, the task is likely to become concave (as people develop the skills to perform it successfully more than 50% of the time). Research is inherently convex: most things explored don’t pan out, but those that do deliver major benefit. When those explorations lead to repeatable processes that can be carried out by people of average motivation and talent, that’s concave, commodity work.

MacLeod organizations exist as a risk trade between those who want to be protected from the vagaries of the market, so creating islands of concavity and easiness is kind of what they do. The Big World Out There is a place with many pockets of convexity, plenty of bad local maxima, and a difficult and mostly unexplored landscape. The MacLeod organization provides its lower-level workers with access to an already-explored and safe concave hill neighborhood. Executives read maps and find start points. Managers just follow the steepest gradient up to the top, and workers just have to follow and carry the manager’s stuff.

Technology and change

There’s a problem with concave work. It tends to be commoditized, making it hard to get substantial profits on it. If “100% performance” can truly be defined and specified, then it can also be achieved mechanically. Not only are the margins low, but machines are just better at it than humans: they’re cheaper, don’t need breaks, and don’t make as many mistakes. Robots are taking over the concave work, leaving us with convexity.

We cannot compete with robots on concave work. We’ll need to let them have it.

Software engineering is notoriously convex. First of all, excellent software engineers are 5 to 100 times more productive than average ones, a phenomenon known as the “10X engineer”. As is typical of convex projects, most software projects fail– probably over 80 percent deliver negative net value– but the payoffs of the successes are massive. This is even more the case in the VC-funded startup ecosystem, where companies that seem to show potential for runaway success are sped along, the laggards being shot and killed. In a convex world, that’s how you allocate resources and attention: focus on the winners, starve the losers.

Convexity actually makes for a very frustrating ecosystem. While convex work is a lot more “fun” because of the upside potential and the challenge, it’s not a great way to make a living. Most software engineers get to age 30 with no successful projects (most of this being not their fault) under their belt, no credibility on account of that series of ill-fated projects, and mediocre financial results (even if they had successes). Managing convex work, and compensating it fairly, are not things that we have a society have figured out how to do. For the past 200 years– the industrial era, as opposed to the technological one that is coming on– we haven’t needed to do it. Almost all human labor was concave. What little convex work existed was generally oriented toward standardizing processes so as to make 99.9% of the organization’s labor pool concave. We are now moving toward an economy where enormous amounts of work are done by machines, practically for free, leaving only convex, creatively taxing work.

The fate of the MacLeod organization

MacLeod organizations, over the past 200 years, could perform well. They weren’t great at innovation; they didn’t need it. They got the job done. One of the virtues of the corporation was its ability to function as a machine made out of people. It would render services and products not only at more scale, but much more reliably, than individual people could do. The industrial corporation co-evolved with the failings of each tier of the MacLeod organization, hence converging on the “optimization model” above that uses the traits of each to its benefit. Of course, I do not mean to suggest that these “computations” are performed in reality, but the metaphor works quite well. 

The modern technological economy has created problems for that style of organization, however. Microeconomic models tend to focus on a small number of business variables– price points, quantity produced, wages. Current-day challenges require thousands, often ill-defined. What product is one going to build? What kind of people should be hired? What kind of culture should the company strive toward, and how will it enforce those ideals? Those things matter a lot more in the technological economy. Hidden variables that could once be ignored are now crucial, and industrial-era management is falling flat. Combine this with the convexity of input/output relationships regarding individual talent, effort, and motivation, and we have a dramatically different world.

The islands of concavity that MacLeod organizations can create for their Losers and Clueless are getting smaller by the year. The ability to protect the risk-averse from the Big World Out There is diminishing. MacLeod Sociopaths were never especially scrupulous about keeping that implicit promise, but now they can’t.

Even individuals now have to make non-local (formerly executive-level) decisions. For a concrete example, consider education. The generalist education implicitly assumes that most people will have a concave relationship between their amount of knowledge in an area and utility they get from it. It’s vitally important, as most educational institutions see it, for one to first get a mediocre amount of knowledge about a lot of subjects. (I agree, for non-economic reasons. How can a person know what is interesting without having a wide survey of knowledge? A mediocre knowledge gives you enough to determine if you want to know more; with no knowledge, you have no clue.) However, there’s no such thing as a Generalist job description. The market doesn’t reward a breadth of mediocre knowledge. People need to specialize. In 1950, having a college degree bought a person credibility as someone capable of learning quickly, thus entry into the managerial, professional, or academic ranks. (Specialization could begin on the job.) By 1985, one needed a marketable major: preferably, math, CS, or a physical science. In 2013, what classes a person took (compilers? machine learning?) is highly relevant. The convex valuation of a knowledge base makes deep knowledge in one area more valuable than a broader, shallower knowledge. Choosing and changing specialties is also a non-local process. A well-rounded generalist can move about the interior by gradually shifting attention. The changing specialist must jump from one “pointy” position to another– and hope it’s a good place to be.

In technology especially, we’re seeing an explosion of dimensionality. General competence doesn’t cut it anymore. Firms aren’t willing to hire “overall good” people who might take 6 months to learn their technology stacks, and the most credible job candidates don’t want to pin their careers on companies that don’t strongly correspond with their (sometimes idiosyncratic) preferences. When there’s a bilateral matching problem (e.g. dating) it usually has something to do with dimensionality. Both sides of the market are “purple squirrel hunting”.

This proliferation of dimensionality isn’t sustainable, of course. One thing I’ve come to believe is that it has an onerous effect on real estate prices. That might seem bizarre, but the “star cities” are the only places that tolerate purple squirrel hunting. If you’re a startup that wants a Python/Scala/C++ expert with production experience in 4 NoSQL products and two PhDs, you can find her in the Bay Area. For some price, she’s out there. That’s not because the people in the Bay Area are better; it’s that, with more of them, you get a continuous (it’s available at some price) rather than discrete (you might wait intolerably long and not find it) market for talent– and also for jobs, from a candidate’s perspective– even if you’re trying to fill some ridiculous purple squirrel specification. That’s what makes “tech hubs” (e.g. Bay Area, New York, Boston) so attractive– to candidates and companies both– and a major part of what keeps them so expensive. The continuous markets make high-risk business and job-hopping careers– that aren’t viable in smaller cities unless one wants to move or tolerate remote work– possible. Since real estate in these areas is reaching the point of being unaffordable for technology workers, I think it’s a fair call to say that this dimensionality explosion in technology won’t continue forever. However, convexity and high dimensionality in general are here to stay, and about to become the norm for the greater economy. The convexity introduced by an economic arrangement where an increasing bulk of commodity labor is dumped directly on machines has incredible upsides, and is very attractive. Now, in the late-industrial era, global economic growth is about 4-5 percent per year. In the thick of the technological era– a few decades from now– it could be over 10% per year.

If MacLeod rank cultures are going to become obsolete, what will replace them? That I do not know for sure, but I have some thoughts. The “optimization model” paints a world where the relevant business variables are known. Executives call out initial values (based on non-local knowledge) for a gradient ascent performed by managers. As the business world becomes high-dimensional– too many dimensions for any one person to handle them all– it begins to break down the problem and distribute the “computation” (again, solely in metaphor). High-ranking executives handle important dimensions (sub-problems) where tricky non-local jumps might be in order. Managers handle less-important ones where continuous modulation will do. Getting the communication topology right is tricky. Often the conceptual hierarchy that is created will look suspiciously like the organizational hierarchy (Conway’s Law?) This leads to an interesting question: is this hierarchy of people– which will limit the firm’s capacity to form proper conceptual hierarchies and solve its own problems– even necessary? Or is it better to have all eyes open on non-local, “visionary” questions? Is that a good idea? Organizations claim to want their employees to “act like owners”. Is that really true? With the immense complexity of the technological economy, and the increasing inability of centralized management to tackle convexity (one cannot force creative excellence or innovation by managerial fiat) it might have to be true.

Enter the self-executive. Self-executive people don’t think of themselves as subordinate employees, but as free agents. They don’t want to be told what to do. They want to excel. A manager who will guide one (mentorship) gets loyalty. However, typical exploitative managers get ignored, sabotaged, or humiliated. Self-executive employees are the ones who can handle convexity, and enjoy the risk and challenge of hard problems. They strongly toward chaos on the civil alignment spectrum. These are the people one will need in order to navigate a convex technological economy, and the self-executive culture is the one that will unleash their capabilities.

That said, the guild culture has a lot to add as well and should not be ignored. There’s a lot of lost work in exploration that can be eliminated by advice from a wise mentor (although if things change, as they do more rapidly these days, that “don’t go there” advice might sometimes be best discarded). The valuation of knowledge and skill are so strongly convex that there’s immense value generation in teaching. Not only should that not be ignored, but it’s going to become a critical component of the working culture. Companies that want loyalty are going to have to start teaching people again. Self-executives don’t work hard unless they believe they’re learning more on the work given to them than they would on their own– and these people tend to be fiercely autodidactic.

This brings us to the old quip. A VP tells his CEO that the company should invest more in its people, and he says, “What if we spend all that money training them and they leave?” The VP’s response: “what happens if we don’t and they stay?” That ends up looking like MacLeod rank culture over time. There’s a lot to be learned from guild culture, and when I finally Solve This Fucking Thing (Part 11? 12? 5764+23i?) I won’t be able to afford to overlook it.

Well, this is yet another “second-to-last post” in the Gervais/MacLeod series (See: Part 1, Part 2, Part 3, Part 4, Part 5, Part 6, Part 7) as I’ve realized that I need to cover one more topic: human nature, especially in the context of the corporate organization (e.g. Theories X and Y). What is it? Is it inherently good, or evil? Is it natural for people to be altruistic, or selfish? I addressed the morality and civility spectra and it should be obvious that I am neither committed to the idea of an inherently bad or good human nature. Mostly, I think people are localistic. We are altruistic to people we consider near to us in genetic, tribal, cultural, or emotional terms. We’re generally indifferent to those we regard at the periphery, favoring the needs of our tribe. Good and evil don’t escape from this localism; they just handle it differently. Good attempts to transcend this localism and (perhaps cautiously) grow the neighborhood of concern: expanding it to all citizens of a polity, then all humans, then all living beings. Evil, not always being egoistic, turns this localism into militancy. Both involve an outside-the-system comprehension of localism that is somewhat rare, leaving most people in an alignment considered neutral. Morally neutral people are best described as weakly good. Assuming they have a strong sense of what good and evil are, they’d prefer to be good, but this preference is not strong and they do not have a burning desire to seek good at personal or localistic risk.

The civility spectrum, between law and chaos, reflects peoples’ biases toward organizations and those who lead them. While lawful good people will oppose an evil society and chaotic good will support a good one, the truth about most societies and organizations is that they are themselves morally neutral, so a person’s civility (bias in favor or against establishment) will influence her tendency to oppose or support power more than the sign-comparison of her and its moral alignments. Lawful people think organizations tend to be better than the people who comprise them; chaotic people think they tend to be worse than the people who make them up. For my part, I’m chaotic, but just slightly. I think that individual people average a C+ on the moral scale (A being good, F being evil) and organizations tend to average a C-. Chaotic bias makes it natural to see corporations as “evil”; in reality, most of them are indifferent profit maximizers.

Interesting enough, software engineering is intrinsically chaotic. Because software requires exact precision, while human communication is inherently ambiguous, large software teams do not perform well. The per-person productivity of a large development team is substantially lower than that of an individual engineer. A team of 10 might be 2-2.5 times as productive as a single engineer. This leads us, as technologists, toward the (chaotic, possibly faulty) assumption that organizations are inherently less than the sum of their parts, because that is clearly true of software engineering teams.

Management theorists have questioned human nature, generating two opposite sets of assumptions about the typical employee of a corporation.

Theory X (presumed egoism): employees are intrinsically lazy, selfish, and amoral. If they are not watched, they will steal. If they are not prodded, they will slack. They are not to be trusted. The manager’s job is to intimidate people into getting their work done and not doing things that hurt the company.

Theory Y (presumed altruism): employees are intrinsically motivated and inclined to help the organization. If they are given appropriate work, they’ll do well. The manager’s job is to nurture talent and then get out of people’s way, so they can get work done.

Theory X is socially unacceptable, but a better representative than Y of how business executives actually think. Theory Y is how executives and organizations present their mentality, because it’s more socially acceptable. So which is right? Neither entirely. Theory X is ugly, but it has some virtues. First, it can be, perversely, more egalitarian than Theory Y. Theory X distrusts everyone, including the most talented and best positioned. Executives are no better than worker bees; everyone must be monitored and a bit scared. Theory Y, which is focused on talent and development, requires (non-egalitarian) decisions about whom to develop. Second, Theory X is more tolerant of scaling, because large-scale societies run (by necessity) on X-ish assumptions. To keep a Theory-Y organization intact, you cannot hire before you trust. Only in the technological era (where small groups can deliver massive returns) has it been possible for growth-oriented organizations to hire so selectively as to make Theory-Y organizational policies tenable.

My ideology (e.g. open allocation) might be seen as “extreme Theory Y”, but that’s not because I believe Theory Y is inerrant. It’s not. Reality is somewhere between X and Y. I believe that organizations ought to take the Y-ward direction largely (on this spectrum) for the same reason that archers aim slightly above their targets. With the actual leadership of most organizations tending toward egoism and X-ness, an organization that doesn’t set inflexible, constitutional Theory-Y pillars (for some concerns) is going to suffer a severe X-ward bias. X-ism is tolerable for concave industrial work, but in the convex world, organizations need to be somewhat Theory Y. How X (or Y) should an organization be? There’s actually a very simple and absolutely correct answer here: trust employees with their own time and energy, distrust those who want to control others’. It really is that simple– a rarity in human affairs– and to continue with anything else is moronic. Employees who volunteer to use their own energies toward something they believe will benefit the organization should be trusted to do so; those who exhibit a desire for dominance over others should be deeply distrusted.

There’s one thing I haven’t addressed, which is which Theory is actually more in force. Theory X was the industrial norm from antiquity to about 1925, when Henry Ford discovered that being a jerk (which almost all industrialists at the time were) was bad for business. High wages for employees meant a strong consumer base. Eight-hour work days were just as productive as longer ones, with fewer accidents. While there were some severe bumps in the road (Great Depression, World War II) the following 50 years saw the emergence of a large middle class, and a changing workforce. Theory Y, at least in aspiration, set in, along with the growth of positive psychology and even the 1950s-70s countercultures, which were more of a reaction against perceived hypocrisy (in organizations claiming to be Theory Y) .

With Theory-Y organizations– especially in research and development– we cracked the German Enigma, sent people to the moon, advanced science more in one half-century than had ever been thought possible, invented the Internet, and grew the global economy at an astounding 5.7 percent per year. Theory Y was the dominant organizational culture from 1925 to about 1975. Then something happened in the counterculture. The 1950s counterculture was mild, liberal, and cautious about the potential for organizational overreach, but tame by modern standards. The 1960s took these seeds of dissent to their logical (civil rights, Great Society) and illogical (Tim Leary, Weathermen) conclusions. The 1970s counterculture was transitional, meek, and reactive to the failed aspirations of the 1960s. In the 1980s, the counterculture was: Let’s Be Dickheads Again. Thus emerged the golden age of private equity, rampant cocaine use (exacerbating its already-present tendency toward context-free arrogance and vacuous superiority) among the upper class, and pro-corporate “greed is good” mentalities. The yuppie generation disgusted their (cautiously liberal, as befit the 1940s-60s) parents with how illiberal and materialistic they were.

Theory Y failed in the 1980s. If your employees are coming into work looking to steal your secrets and launch their private equity careers, you actually can’t trust them. This decade of betrayal, greed, and organizational dissolution proved Theory Y inadequate. Bad people exist at all levels. Some people will try to steal from their employers, employees, and colleagues.

If the Gilded Age nightmare of Pinkertons and company towns was the height of Theory X, and the mid-20th century United States was that of Theory Y, what came after? The chaos of the 1980s settled down, and I think what emerged in its wake can be called Theory Z. By 1995, corporations had been looted at bottom and top (mostly, at the top) and had ceased to inspire. Technology startups were taking on corporate behemoths of much greater size. People at the bottoms of corporations (MacLeod Losers) were beginning to recognize that presumed upward mobility could no longer be believed in. The arrogant egoism of the coked-up 1980s ubermenschen had faded somewhat, but the bilateral altruism existing between the paternalistic corporation and employee was forever gone as well. People returned to localism in personal alignment: trying to do right by the people they care about, and the people near them.

Theory Z (prevailing localism): a few employees will be unusually egoistic or altruistic, but most are going to be localist. Interpersonal loyalty will bind them together, and growing affinity within the group will encourage “pro-social” behavior. People who feel excluded by the group will defect; those who feel included will cooperate. The manager’s job is to build a great team– to use an intuition for human localism to direct that tendency toward pro-organizational behaviors– and to marginalize or separate from (i.e. fire) those whom it excludes.

Theory Z is the most accurate of the 3 “human nature” calculi put forward thus far, insofar as it covers most of an organization. One might also note that these 3 theories correspond neatly to the MacLeod hierarchy. The executive suite (MacLeod Sociopaths) tends to be dominated by Theory-X mentality. These people know that they shouldn’t be trusted, so they aren’t inclined to trust anyone else. Clueless middle-managers tend to overestimate human nature and have a Y-ish bias. MacLeod Losers want to be socially acceptable and get along well with the group. The Loser world, driven by interpersonal and team affinity, is a Theory-Z one. They want to get along, and will manage their effort level to the exact point that keeps in the best social standing– the Socially Acceptable Middling Effort (SAME).

Theory Z may be the most accurate model of the MacLeod Loser class that does most of the work in an organization. This said, Theory Z also has some severe defects, having generated a cargo cult of teamism. Organizations waste time and money on pointless “team-building” paraphernalia: “mandatory fun” retreats that no one enjoys, in-office perks that adolescentize the workforce but detract from actually getting stuff done. A person is judged not on her individual merits, but based on (a) the social status, outside of her control, of the team on which she has landed and (b) as a tiebreaker, her performance on that team. The top people in the organization (rather disgustingly) call themselves “the leadership team”. Teamism also creates closed allocation, of which I’m not a fan. People who attempt to serve the organization directly by moving to more appropriate teams (which their native teams and managers view negatively as attempts to swing to higher-status teams) are viewed as “not team players” and, instead of being allowed transfer, are discarded. Teamism is especially defective in software, where large teams are almost never productive. Theory Z conformity actually solves the industrial problem: what’s the best way to manage concave work? Concave work is that in which the difference between mediocrity and excellence is minimal in comparison to that between mediocrity and noncompliance (zero) and variance reduction (at which management excels) is desirable. It doesn’t solve the technological problem that emerges when we confront convex work, in which the difference between excellence and mediocrity is critical and that between mediocrity and nonperformance is negligible.

The industrial paradigm is heavily oriented toward concave work. To see that, consider educational testing. Students are given very easy problems (most of the difficulty being in artificial resource limits– timed, closed-book exams) so that an average performer will get 85 percent right. The pass/fail line is then set at 70%– in other words, no more than twice the defect rate of the average. If we wanted to re-orient exams toward a convex world, we’d give students very hard problems so that average performers only get 20% (the pass rate might be 10-15%) and call excellence 40%. I’m not actually saying that’s a good idea– I’m out of my depth on these sorts of educational issues– but this is just one way in which in the presumed concavity of industrial work is visible in the pedagogical training people get before entering it.

Why did Theories X, Y, and Z exist? What will replace them? To answer this, it’s useful to look at humanity in several stages– agrarian, industrial, and technological– based on the prevailing rate of economic growth. In the agrarian era, from 10000 BC to about 1750, economic growth was slow (0.01 to 1.0 percent per year) and generally imperceptible in a human lifetime, especially in comparison to the local rises and falls of empires. Most people who wanted to get rich had to steal or kill. Mercantilism was the predominant economic theory, slavery was he most common form of organizational labor, and Malthus was right– not in his modeling of food production growth as linear, it being a slow-growing exponential function; but in his assumption that human population growth exceeded agrarian economic growth. (England didn’t have a Malthusian catastrophe, the Industrial Revolution intervening, but an overwhelming number of societies have had them. Some have argued that England, in the 19th century, outsourced its Malthusian problem to Ireland.) Economics in the agrarian era could be approximated as zero-sum; with population growing as fast the economy did, the average human’s standard of living didn’t improve much. Machiavelli probably wrote The Prince as satire, but it was apropos of the political climate of the time, and any time before or up to about 250 years after that.

The industrial world came into being gradually, with the advent of science and, later, rational government. It started in the late 17th century, and by the 18th, progress was (while slow) visible. Malthus, despite his pessimistic projections, acknowledged that growth existed: it just wasn’t happening very fast in 1798– about 0.9% per year. This rate being too slow to sustain human population growth, economics truly earned its name of “the dismal science”. Personally, I define the industrial threshold (very arbitrarily) as the point (early 19th-century) at which global economic growth reached 1.0% per year. Since I define the technological threshold at 10% per year, we haven’t gotten there yet. (More on this here.) But the most interesting companies (technology firms oriented toward convex work) have that capability.

The architects of the industrial world were quick to realize that coercive labor wouldn’t suit their needs: the jobs were too complex and variable to leave to people who’d been deprived of all autonomy (slaves). This had to be replaced with a semi-coercive model in which employees had some freedom: they’d need to have a boss to survive, but they could choose which one. Industrialists studied sailors (pirates, privateers, explorers and merchants– all different in how ships were run) to learn about group sociology apart from the agrarian state. They studied militaries, large organizations which had left important duties to non-coercive labor (and less important ones to semi-coercive conscripted labor) for centuries. They looked at prisons to see how free people handled the temporary loss of liberty that would be similar to a merchant’s conscription into a middle-management office role. (Slaves were rarely put into prisons, but beaten or killed.) As most complex organizations of the time were semi-coercive, vicious, and prone to violence (that was often a part of the business) this naturally led into a Theory-X mindset: bring ‘em in, and don’t beat ‘em so hard they can’t work, but don’t trust ‘em either.

The zero-sum world of agrarian humanity suffered a major blow in the mid-19th century when the industrialized nations began abolishing slavery, but human behavior is slow to change. Progressive mentalities began to form within nation states, but the old ways of interaction still existed between them, and also between advanced nations and the colonized people. That blew up spectacularly in the World Wars. By 1945, it was evident that being a jerk was not going to work anymore. Racism, for one example, lost all intellectual respectability after what Hitler did. Militant localism (jingoism) had to be replaced by a climate of prevailing respect and positive-sum thinking. The U.S. rebuilt the economies of nations it had defeated at war, instead of inflicting further economic penalty as occurred after World War I. The corporate analogue of these changes came out of positive psychology and political progressivism: Theory Y.

Unfortunately, while Theory Y built good organizations, it left them unable to defend themselves against bad people, as the 1980s showed us. Academia and basic research, in the U.S., still haven’t recovered from the barbarian attacks. Rather, it’s ongoing. Global economic growth dropped– from 5.7% per year to about 3.5– due to society’s disinvestment in progress and science. (It has recovered somewhat, to 4.8%, largely because of the declining relevance of the gutted U.S.) The chaos of the 1980s left working Americans bereft of faith in institutions and in the people they worked with. This led to the more cautious and accurate Theory Z, which correctly models human localism but prescribes a managerial style based on conformity and mediocrity– solving the concave/industrial problems, but failing at the convex/technological ones.

So, what is human nature? Are people inherently altruistic, egoistic, or localistic? We’ve seen a tendency toward localism– somewhere between altruism and egoism– as a default. Does this mean that “human nature is localistic”? Can we say that human nature is morally neutral (rather than good or evil, as some philosophers have suggested)? For my part, I don’t. I’m not convinced that it’s anything, because I don’t hold strong beliefs in human nature. I’m not sure that there’s a there there. We can understand biophysics mathematically, observe sociality, and experience spirituality, but a complete understanding of ourselves eludes us. “Human nature” is a “God of the gaps”.

Personally, my philosophical and religious beliefs are most in line with Zen Buddhism. It would be un-Zen to say that I am or am not a Buddhist, so I won’t, but I believe that the Zen approach to reality is among the most accurate. Most phenomena are empty. People tend weakly toward moral good, but circumstances can easily steer normal people toward lawful evil (Milgram Experiment) or even the chaotic kind (Stanford Prison Experiment). Theory X presumes a hostile human nature, as a slaveowner might. Theory Y presumes an altruistic one, leaving organizations unable to defend themselves against bad actors. Theory Z correctly concludes the human default to be localism  but settles prematurely for mediocrity and cargo-cult teamism. None of these are well-equipped to tackle the needs of the technological era, in which the fast rate of growth and change necessitate unlocking creative energies, while a certain caution is needed regarding those who might wish to subvert the organization, or gain inappropriate dominance over it.

Theory Z gets what Y did not– that there are “toxic” bad actors out there that the organization must reject– but takes a stupidly teamist approach. People aren’t fired from Theory-Z organizations because they’re harmful, bad people, but because they’re “not a team player”. The effort is almost never exerted to assess alternative possibilities to individual defect, such as (a) a defective or poorly-configured team, (b) bad management, or (c) no-fault lack-of-fit. All of these are more common than the extremely damaging but rare toxic individual.

In the convex world, creative output isn’t going to come from “teams”, at least not in the managerial sense where the teammates have little control over membership and organization, and in which “team” is conflated with “career goals of the manager”. (Note: a manager who says “not a team player” is actually saying, “not a me player”.)  Theory-Z management tries to control human localism, corralling people together and saying, “Be a team, now!” That doesn’t work very well. Rather, the creative energies that can produce technological-era progress come from individuals who sometimes choose to form teams, and sometimes to work alone.

Why is Theory Z just as foolish as X and Y? X and Y inaccurately claim “human nature” to have a strong directional bias toward self-serving egoism or pro-organizational altruism. It does not. Theory Z maintains a belief in “human nature” and assumes it to be inflexibly localist, because that’s an observed default. I maintain that “human nature” is pretty damn empty. People are mutable. Don’t settle for bland localism; you’ll get pointless institutions that way. People can be very good; try to make it happen. They can also be very evil; try not to have that happen. They will sometimes form teams; that is fine. They will sometimes work alone; that is also fine. Judge people on their actions and not assumptions about some “nature” that is illegible at best and nonexistent at worst.

How does one convert this into an actionable management style? Lord Acton said it very well:

Judge talent at its best and character at its worst.

Theory X fails because it allows no room for excellence (talent at its best). Theory Y fails to account for bad actors (character at its worst). Theory Z throws its hands up in the air and mediocritizes: let’s all just get along and be a team. How do we assess talent or character? The truth is that we can’t; we can only look at peoples’ actions. In practice, this usually gives us enough data. If people show even the potential for excellence, that should be explored and encouraged. On the other hand, it should be very rare (if ever?) that a person is presumed to have good character and given more power over others than is absolutely necessary. So I actually nailed it, already, above. Here is an upgrade of Theory Y that is more robust against problematic people:

Theory A: trust employees with their own time and energy; distrust those who want to control others’.

That is where I’ll stop for today.

Over the past couple weeks, I’ve delved into organizational health and the processes that compromise it. (See: Part 1, Part 2, Part 3, Part 4, Part 5, Part 6.) Typically, organizations tend toward a MacLeod hierarchy with three tiers: the Sociopaths, the Clueless, and the Losers, in that order from top to bottom. Sociopaths, who take an up-or-out strategy and either end up at the top or fired, are strategic and dedicated but not subordinate. They’ll never sacrifice their individual career goals for the benefit of the organization, although they may work to improve the company if there is gain for them. These aren’t always bad people. Of the three personalities, I have most in common with the Sociopaths, but I’m not a bad (much less psychopathic) person. Clueless, on the other hand, are dedicated and subordinate but not strategic. They’ll work hard, and take orders, but they don’t have a good intuition for what is worth working on. They’ll eagerly follow or lead pointless, “death march”, projects. They tend to be shunted into middle management where their dedication and eagerness are an asset but their lack of strategy does minimal damage. Losers are subordinate and strategic but not dedicated. They know what’s worth working on and what not, and will follow orders– they’re subordinate because it makes their lives easier– but rarely put forth more effort than is required of them. (As already said, they’re not actually “losers”. More on that, later.) They’re discomfort-minimizers, while the equally strategic Sociopaths are yield-maximizers.

The MacLeod process exists because most organizations make it inconsistent for a strategic person to also be subordinate and dedicated, meaning that the confluence of these three traits won’t occur. Strategic people tend to be rational and selective with their efforts. Some will aim to minimize change and discomfort, willing to sacrifice compensation and career progress in exchange for an easy job where they won’t get fired. That favors subordinacy: make the boss happy, get away with more and reduce risk. If such a person can leave work at 3:00 and collect the same salary, he will. (Why not? If the firm remains happy with his work ethic, is it unethical?) These are the Losers. It’s important to note that they’re not losers in the sense of being disliked, undesirable, or defective. They only “lose” in the first-order economic sense, since their position is one in which they receive less (usually, a steady pittance) from the organization than they put forth. They trade expectancy (long-term average compensation) for the risk-reduction offered by a large, stable company in which they can retain good standing by appeasing a small number of important people. There’s another crowd who are also strategic, but who have more of an appetite for risk and want to capture some of the surplus value generated by the Losers’ trade (of expected value for risk reduction). They make the equally rational decision to put forth a lot of effort, and to take on a lot of risk, in exchange for rapid career growth. They tend to be dedicated, hard workers, but they’re not subordinate. This doesn’t mean that they’re ideologically or obnoxiously insubordinate, but they’ll never buy into the fiction that people must put company goals above their own career needs. Ultimately, such a person will gladly work a 50 or 70-hour week for the right price, but will never sacrifice her own career goals for the company or “the team”. Such people can’t stay in one place for very long. It’s not that they make themselves disliked, but there is a certain “offness” about them. They’re ambitious. They have “an agenda” (as if that were a bad thing). They’re only loyal to companies that treat them well and give their career needs a special value that can’t be given to all. They’re mercenary. It’s up-or-out for them. Companies promote as many of these as they have room for, but must discard the rest. This is actually an impersonal and necessary process– when you have more ambitious people than there is room at the top, you must make a way out for some, even if they are good people that you personal like. However, companies often have a psychological need to create a mythology for hard decisions, and this leads into the “not a team player” epithet that exists to justify this process (that, in my opinion, companies are not required to justify, because it’s not immoral to fire people; it’s just the way human organizations work). These strategic, ambitious people are the Sociopaths. Finally, the non-strategic who are dedicated and subordinate must, almost by definition, be Clueless. Those are the “true believers” who give to the company without expecting much in return.

The MacLeod organization is very stable according to its own internal metric. The Clueless provide a barrier between Losers and Sociopaths on the rank spectrum, while Losers (who give little, and get little) provide one between (value-capturing) Sociopaths and Clueless (who give a lot, and get little) on the hedonic spectrum. The effort thermocline and differential social status hold everything together. It’s an envy-reducing structure. Clueless are oblivious to the effort thermocline and don’t want the jobs immediately above them. Losers don’t want to become Clueless middle-managers– with only token power and slight improvements to compensation, but substantial increases in responsibility and discomfort. Thus, very few people want the jobs immediately above them, and those who clearly do want to ascend (Sociopaths) are fast-tracked either up or out. The MacLeod organization is, on its own terms, fairly peaceful. On the external market, however, such organizations are not always able to compete. MacLeod hierarchies keep peace within the firm, but often make the organization slow to adapt to the world outside.

The problem with the MacLeod organization is that it’s driven by three tiers for which each has a critical defect. Insubordinacy (of the Sociopaths) is not always bad, but it is a defect from an organization’s point of view, and with no one to audit the Sociopath tier here is no way to exert control over the moral character of the leadership. If “good Sociopaths” (Technocrats) are in charge, it will be a good organization. If bad people get into power, they will drive out the good. The problem isn’t that such ambitious people are uniformly or even often bad, but that organizations can almost never self-regulate in such a way as to audit or change the moral character of their leadership– and once an organization goes bad, it will rarely revert, because bad people have an innate competitive advantage; true psychopaths are more agile in social competition than the irreverent-but-decent “good Sociopaths”. In short, MacLeod Sociopaths are not necessarily bad people, but they’re impossible for an organization to control. The Clueless, lacking strategy, are unable to function without good people above and below them. They produce most of the tactical effort, but require management from above and below in order to retain focus on useful stuff. This dependence on others is their shortcoming. Finally, there are the Losers, who do most of the work and will directly affect the company’s ability to execute as well as its reputation for quality of service but who are, rationally, only dedicated enough to remain in good social standing. It’s not true that they “do the minimum not to get fired” (that’s a common misunderstanding of Loser-ism) so much as they manage their performance to the Socially Acceptable Middling Effort (SAME). They want neither the reputation for being a slacker, nor that for working too hard. They aim for the middle. While less important than the strategic competency of the leadership, the SAME level is a major cultural factor in a company’s macroscopic performance.

The organizational issue with the SAME is that it drifts over time. A high SAME will keep the working people of the company, who are often very competent despite their lack of ambition, highly productive. On the other hand, when the SAME falls to zero, the company ends up with a lot of deadwood; Losers stop working, because there’s no reason to do so. In this light, we can understand Marissa Mayer’s recent decision to crack down on work-from-home employees. This effort is about raising the SAME, which is hard to manage in a distributed setting. The problem isn’t that WFH employees are likely to be lazy. That’s often not the case. In fact, the best WFH-ers are far more productive than any in-office employee. The issue is that the good WFH-ers have no effect on the SAME– they get a lot done, but the rest of the office doesn’t see them working hard– and the bad WFH-ers have fallen to zero. An in-office policy will, at least at first, reduce bulk productivity (by hurting the good WFH-ers more than it brings up the bad ones, most of the latter being incorrigible people who will need to be fired) but the expected second-order effect is to raise the SAME. Mayer’s changes may kick Yahoo into the ugly state of a tough culture, but that’s likely to be less dysfunctional than the stodgy rank culture into which MacLeod corporations devolve. In rank culture, getting along well with management is more important than effort itself, which pushes the SAME down as obedient non-performers get a pass and weak managers become entrenched.

If each of the 3 MacLeod tiers is defective in one of these three desired organizational traits, we must ask the question. Can people be strategic, subordinate, and dedicated? It only happens in the context of a mentor/protege relationship. In truth, people who are truly strategic are never fully subordinate or insubordinate. While I originally approached the MacLeod hierarchy and this three-trait decomposition, I assumed the existence of all three traits to be impossible. In reality, those who are strategic are subordinate or insubordinate based on context. Strategic and dedicated people will subordinate, in the short term, if the leadership (management, advisors, investors) shows a long-term interest in their careers. True loyalty is not incompatible with the insubordinate tendency of the highly capable mind. There’s a pay-it-forward mentality and a personal affinity that enables such people to live in harmony with the organization. This is observed in the rarest of the four work cultures: the guild culture, which creates balance in the relative importance of dedication, subordination, and strategy. The other cultures tend focus on one of the 3 traits at the expense of the others: rank culture, on subordinacy; tough culture, on dedication and sacrifice; self-executive culture, on being strategic. Guild cultures encourage balance. Why, then, are guild cultures rare (and dying)? The answer is two-fold. First, they’re hard to maintain, because the leadership must continually refresh its skill base in order to mentor new people, which means that the teachers must “moonlight” as students and study new methods as well as the external market. Second, guild cultures cannot grow fast. If they take on too many new people at subordinate ranks, it becomes clear that the not all of their careers will succeed (there isn’t enough room higher on the ladder) and the whole thing falls to pieces. While guild culture may be “good”, it is not fit (as an evolutionary term) insofar as it does not allow fast growth of its organizations.

A guild culture that collapsed recently is that of large-firm law (“biglaw”). Originally, it was difficult to get an associate position at a “white-shoe” law firm, but one who had one stood a very high chance of making partner. Not making partner (after seven to ten years) was the exception reserved for bottom-5% performers. In the 1980s, this changed, due to the increased workload generated by the private equity boom. The firms began taking on large numbers of associates without allowing the partnership ranks to grow in tandem, the result being that making partner is now the rarity. Now, it’s about 5 percent who get partnership, rather than 5 percent being denied it. The guild culture died horribly, being replaced by a catastrophic tough culture where 70-hour work weeks and “4:30 work drops” (late-day assignments with next-morning deadlines) are the norm. New York attorneys frequently describe the state of their (ex-)profession as “banking without the upside”.

So what defines a healthy working culture? We now have the vocabulary to address this. I’m going to elaborate six statements about a healthy workplace culture, and then arrive at a 7th (The Mike Test) which is the most important of all. I derive these six assertions from the three workplace traits– strategy, subordination, and dedication– in addition to the need for moderation in each.

1. (Dedication, or DED) People can be relied upon to do their jobs well, and to treat their work as important. Ideally, this trust and reliability should exist at all levels of the organization. The company functions best if everyone is willing to do the hard jobs. 
2. (Moderation in dedication, or M-DED) People do not engage in unnecessary sacrifice for social or subordinate reasons. Pain does not become a measure of a person’s work or value.
3. (Subordinacy, or SUB) People take a “pay-it-forward” attitude toward their colleagues and the company. Junior employees take direction from mentors; seniors invest in their reports for the long term. People invest in their relationships with the company, because it is worthwhile to do so. 
   
4. (Moderation in subordinacy, or M-SUB) Doing the right thing is more important, to employees and the company, than following orders. This requires a culture that enables people to speak up without fear of retribution.
5. (Strategy, or STR) Individual employees take ownership of their own work and have the autonomy to place their efforts where they perceive the most value. There’s no need to ask permission or “apply for transfer” to work on something that appears important. This is where open allocation shines.
6. (Moderation in strategy, or M-STR) Allowance is made for exploratory work that might pay off only in the long-term. Work need not deliver short-term dividends to be acceptable. People are allowed, at all ranks, to invest at least some of their time into R&D that is abstractly beneficial to the organization, even if it doesn’t produce an immediate benefit.

If I were to grade each of the four cultures (0 to 8, 8 best) for its typical level of success on each of these traits, here’s how I would assign them (based on their observed performance, not on their purported values).

Trait Rank Tough Guild Self-exec.
DED     2    6     5      8
M-DED   3    0     5      6
SUB     4    1     8      4
M-SUB   0    2     5      7
STR     1    4     4      8
M-STR   2    0     5      3
Avg.   2.0  2.2   5.3    6.0


Rank cultures are across-the-board weak. They excel in superficial subordinacy, but the lack of true loyalty earns them only a ’4′ grade for that trait (SUB). The only good thing about them is their internal stability, which is why they are the eventual state of a hierarchical (MacLeod) organization. Still, they tend increasingly toward macroscopic underperformance due to their corrosive mediocrity. When the only thing that will prevent laziness is aggressive management, effective managers leave (wanting better reports) and the company ends up getting stuck with lazy managers who tolerate crappy employees. Tough cultures are slightly better– people actually give a shit, if only for selfish reasons such as aggressive performance reviews– but still fail in most critical areas, and behavior tends toward moral degradation. In tough cultures, people are as subordinate as they need to be to survive, but deeply disloyal. Two examples of tough-culture degradation are Enron and Google after the introduction of “calibration scores”, but the tracks of the organizations are different. Enron maintained its tough culture by executive fiat and experienced top-to-bottom ethical corrosion. Google’s tough culture (introduced by a play-for-play copy of Enron’s performance review system) reverted quickly to a rank culture, because many managers had no desire to enforce it and began agreeing to “peg” calibration scores in return for loyalty. (Google is an unusual example, having a healthy self-executive culture above the Real Googler Line– enabling it to maintain strong macroscopic performance– and a necrotic tough-turned-rank culture below the RGL.) The latter path is more common. Tough cultures usually return to rank cultures; those who have the power to protect people from the harsh review system become the new holders of rank.

Let’s examine the healthy cultures, in the light of the grades above. Note that self-executive cultures get top marks for dedication and moderation in dedication. When people are trusted to direct their own efforts, and rewarded for good work, they tend to put their effort levels at the right level. They also excel at moderation in subordinacy and (of course) strategy. What self-executive cultures tend to be bad at is rewarding long-term investment, including mentoring new hires. Guild cultures, although not especially fit due to their intolerance of fast growth, are well-rounded and healthy. When they work well, they excel in terms of a “pay-it-forward” attitude that replaces subordinacy (or insubordinacy) with genuine altruism. They’re also, perhaps surprisingly, the most prone to long-term investment (moderation in strategy). The central planning inherent in guild culture can be a weakness, but it can also allow the firm to “future-proof” itself consciously– if its leadership wants to do this.

Both cultures struggle when it comes to growth, but in different ways. Guild cultures can grow vertically– bringing in people of lower or higher skill levels– because they have the machinery for assimilating them into mentor/student relationships, but tend to fail at horizontal growth, because guild cultures require a carefully managed balance of work quality, skill level, and rewards. If that gets out of whack due to rapid growth, some form of scarcity (e.g.,of good reports, of high-quality work, or of room at the top) will turn the firm into a zero-sum slugfest that destroys the mutual trust of the guild culture. (It becomes a tough or rank culture.) On the other hand, self-executive cultures can grow horizontally– taking on more people at the same skill level– but tend to be incapable of assimilating people of superior or inferior skill levels to those who are already there, because self-executive cultures rarely provide the incentives for managed growth (as opposed to going out and putting forth work that brings immediate results). An example is Valve, an ideologically self-executive culture that simply does not hire junior-level people– it has enough self-awareness to know that it can’t maintain its self-executive culture in the face of the skill inequality that vertical growth creates.

In truth, self-executive cultures tend to downplay differences in skill– everyone has basic autonomy, there are no real “bosses”– and can’t hire below the prevailing skill level and maintain their culture. They could hire above the prevailing skill level and survive– that would be desirable– but have a hard time getting such people to work for them without offering some authority to them. People who are used to be entitled leaders (executives) do not usually want to work on equal terms with people they consider of inferior skill. Thus, self-executive cultures can only hire near the prevailing level and, since they only succeed if the prevailing skill level is fairly high, the scarcity of such talent means they cannot grow fast even if they would want to.

Tough and rank cultures, on the other hand, grow much faster. Why is this so? Tough cultures do not focus on growth so much as churn. They hire a lot of people, fire a lot of people, and are responsive to market conditions but generally agnostic on the direction of headcount numbers. Tough culture, being the absence of a work culture, has no growth management. A tough culture can hire a person of low market value, accepting the very high chance (in some cases, over 50% per year) that it will have to fire him. What causes directional growth (i.e. more hires than fires) in tough cultures is the long-term tendency toward inefficiency (necessitating greater headcount) and the emergence of pockets of rank culture as the tough culture decays. Tough cultures tend to grow in spite of themselves, because the cultural corrosion reduces individual performance despite the culture’s intent of doing the opposite. Rank cultures, for their part, actively encourage headcount growth. Hiring binges mean that there are more subordinates to go around, which makes managerial decision-makers happy.

In other words, neither of the desirable cultures (self-executive, guild) excels at horizontal growth and vertical growth. This makes them less fit, at least in terms of the ability to subsume people, than the pathological rank or tough cultures of typical organizations. There might be as many of these desirable cultures (self-executive and guild) as there are pathological cultures, but the latter house more people. It may be that most businesses have good cultures, but most employees work in dysfunctional businesses, if only because the dysfunctional work cultures are most able to grow quickly.

This conclusion is depressing: most people will work in dysfunctional cultures. Can we change that? If we want to do so, we need to answer some questions? What is the best workplace culture? Should one aim for the guild culture, or the self-executive one? I think the answer is obvious: one needs a hybrid, with ideas from both. I would say that one should aim for a culture that is mostly self-executive, and especially so for senior hires, but that creates an internal market for mentoring new employees, and for long-term investments, in which those efforts are equally valued. In doing so, it would manage to capture some of the assets of the older guild cultures, thus enabling some degree of vertical growth.

On the topic of growth: in addition to the six cultural evaluations above, I’ll add a 7th that is, above all, most important. The other six pertain to the present-time cultural health, but this 7th determines whether an organization will improve or decay over time. It’s The Mike Test. It has one criterion. The Mike Test is…

…would you, if no one would know or give you credit for doing so, want to hire someone better than you? Would it be rational to do so?

Assume that there’s no hiring bonus, and she’s not your best friend. You won’t get credit, in any form, for a great hire, and she won’t feel preternaturally loyal to you if she rises fast. The only benefit you get from hiring this great person is that you improve the company. The risk you take on is that your relative status in that company will decline. In a large company, the risk is low– you’re unlikely to compete directly with her– but so is the reward, making the balance zero. (It’s not worth it to hire her even in the large firm, not because there’s a risk of competition, but because you could spend time on other things.) I would argue that most companies fail The Mike Test. In the vast majority of companies, it is not rational to hire a person of superior ability.

This hits on the (slightly altered) Jack Welchism that “A players hire A+ players; B players hire C players”. In my use, however, “A” and “B” orientation pertain to context and security rather than innate competence. A players are secure enough that their interests (at least, in hiring) align with the company’s. They want to work with great people, and it has nothing to do with coarse personal benefits (hiring bonuses, “finder’s” credit). They make their companies better. B players, who are insecure because they are unremarkable, want to be safe and manage themselves to the middle. Hiring incompetents minimizes their risk of drifting out of the middle (and into the bottom). Good cultures discourage this kind of B-ness. They make sure that competent people are secure, and they rid themselves swiftly and fairly (preferably with severance; it reduces drama) of incompetents. Sadly, that’s rare. Most companies fail the Mike Test, and MacLeod hierarchies form because of organizations’ tendencies to hire people inferior to the (semi-insecure) real decision makers and, when superior people are brought in, to try to make them inferior using their control over the division of labor. Those tiers emerge because people in an insecure context have the incentive to hire only inferior copies of themselves.

Let’s discuss each of the four cultures in the context of the Mike Test.

Tough cultures fail the Mike Test in the worst way. In a tough culture, no one is secure– that’s the whole point of tough culture– so everyone is a contextual B-player (i.e. incentivized to hire mediocrities and make oneself look better). No rational person living in a tough culture would hire someone of superior skill– that person becomes an immediate threat. Rank cultures don’t fare much better. In a rank culture, one tends to have one of two kinds of managers. The first is the checked-out (i.e., also lazy and mediocre, as befits rank culture) boss who just doesn’t care. There isn’t a major loss inherent in hiring strong people, but it’s not a worthy use of time. The second is the hard-ass, careerist boss who creates a pocket of tough culture under him. (The company might still have a rank culture, but this boss holds high expectations.) That type of boss only cares about hiring insofar as it builds his team, and the tough-culture rules (don’t hire someone better than you, lest you be thrown under the bus) apply. Bosses in rank cultures might want to hire people of superior skill, but only if they could lock in permanent rank superiority. Rank cultures especially stigmatize reporting inversion (i.e. having a younger, less senior, or less educated boss) to the point that it can be career-ending to be at the butt of one. The benefits of having a strong subordinate are offset, in a rank culture, by the risk of reporting inversion. Therefore, both of the dysfunctional cultures (rank and tough) fail the Mike Test, explaining why they hire worse people with each generation of growth.

Guild cultures pass The Mike Test. If you hire someone better than you are, he might end up in a higher place, but that’s okay. You get a mentor, not a threat. A well-managed guild culture can assimilate someone at a higher skill level without eroding the well-being of the rest of the organization. This is the symbiotic nature of the guild culture. Self-executive cultures also pass, because their “bossless” nature means that the prospective hire’s superiority of skill is not a major concern. It’s a good thing to hire a better person, because you’ll have a strong colleague.

To summarize this briefly for each culture:

• Tough cultures fail the Mike Test catastrophically. You endanger yourself directly if you hire someone better than you are.
• Rank cultures fail the Mike Test by apathy. People only hire strong people if they can guarantee that person’s subordinate status. Rank cultures tend toward inferior hiring– strong people are less likely to be subordinate– but not as fast as tough cultures.
• Self-executive cultures pass the Mike Test, as most people would prefer stronger colleagues, but the equality inherent to a self-executive company is a hard sell to one who might expect (because of a higher skill level) a leadership role.
• Guild cultures pass the Mike Test, if they can convince the superior to take the role of a mentor rather than a manager.

Perhaps surprisingly, most VC-funded startups fail the Mike Test. They pass it for founders and real owners, who will have a more successful company, but they fail for engineers. Founders and investors control the dilution process and, even if they lose relative share, they’ll only make deals that are beneficial to them (50 percent of something is better than 100 percent of nothing.) If you own 20 percent of the company, you’re ecstatic if you just hired someone better than you are. What about engineers, however? What are their incentives?

Let’s consider a typical software engineer at a 50-person technology startup with 25 software engineers. His compensation is $25,000 per year lower than the market level, offset by equity equal to 0.05 percent of the company, vesting over 4 years. (In practice, he’d more likely have options at a low strike price, but I’m going to simplify here.) Salary raises are rare in startups, and equity improvements (without a promotion) are almost unheard-of. When things are going badly, that’s not a time to ask for anything; when they’re going well, the appreciation of equity is the raise. In addition to the $100,000 lost over four years at a low salary, let’s value the lost salary growth at $50,000 over 4 years, plus another $50,000 to account for future income lost to being at a low salary level when he exits. So he’s paying $200,000 for 0.05% of the company, implying a valuation of $400 million. This is, most likely, much higher a valuation than the one at which investors would buy into it. However, I can make a strong case that the engineer’s valuation should be lower than that given by investors, especially in a VC-feeding frenzy. When risk, liquidation preferences, cliffs and the lack of control are included, an engineer doesn’t do well to accept this typical startup offer ($25,000 salary drop, no raises; 0.05% equity) unless he can realistically value the company at approximately $1 billion. With the shoddy IPO climate, not many startups deserve that kind of valuation. (In fact, the companies that do are no longer really startups.) So, the equity offered by a startup is not, for most mere engineers, a good reason to be there.

So why do software engineers work for these VC-funded startups? There are many Clueless in the mix who massively overvalue their equity consideration, but most engineers believe their initial grants to be “teasers”. They know that their starting allocations are low, but expect that they’ll get something closer to a “fair” share when they grab those “inevitable” executive positions at which real equity allotments (0.25 to 0.5 percent for Directors, 0.5 to 1.0 percent for VPs, 2.0 for C-level, 3.0 for CTO/COO and 5.0 for CEO) are common. That is what keeps engineers in VC-istan motivated; the implicit (and often broken) promise of a higher-ranking role (with investor contact, and real equity) as the company grows. The real concern, when engineers participate in hiring decisions, isn’t about equity dilution associated with strong hires. They just don’t have enough equity for that issue to really matter. Their fear is that, if they hire people stronger than them, they’ll lose out on the executive positions implicitly promised to them as early hires at their startups. If they hire engineers better than they are, they’re doomed to languish in a company that has begun hiring above them. One note about typical startup sociology: once your startup hires someone from outside directly above you, it’s over. It will continue doing so, and your career has stalled out. You’re not a real player in the firm.

There’s an inherent conflict of interest in this style of VC-funded startup. The real owners (investors, founders, top management) get a social-climbing mentality and seek to hire stronger technical talent (even if they don’t need it!) for the sake of “scaling”. Engineers, on the other hand, would do well to subvert this. They don’t actually want to hire bad people (terrible software engineers reduce the productivity of the team) so much as they want to hire just-slightly-inferior people who aren’t so bad as to poison the team, but won’t challenge their chances of getting an executive position. This is more of a case of B players hiring B- players than one of them hiring C players. 

However, two to five generations of “just slightly inferior” will turn to bad, and the rapid churn of VC-funded startups means that it doesn’t take that long. VC-istan startups degrade severely and predictably once hiring decisions are made by people with small equity slices, whose concern is not the health of the company (ownership) but the protection of their own inside track to executive positions. Such startups fail the Mike Test, and it doesn’t take long for it to show.

What is VC-istan’s place in the 4-culture taxonomy?

What is the culture of a typical VC-istan startup? Within VC-istan, there are efforts (e.g. Y Combinator) to generate a pay-it-forward guild culture. This is actually quite interesting– entrepreneurship is supposed to be market-oriented, but the most healthy subsector of VC-istan is guild culture– the healthy variety of a command culture. On the other hand, the healthy market culture (self-executive) is shockingly rare. This suggests that VC-istan has become somewhat of a command (not market) culture, and it seems that it would be more of a rank culture than anything else. I tend to believe that this is probably true. VCs talk to each other. They decide as a group who is hot and who is not, and this extortionate power makes them the true holders of rank. VC-istan, a postmodern corporation that manages to transcend mere companies– in VC-istan, companies are disposable– is so amorphous and complex that it cannot be tagged as “only” belonging to one culture, but rank culture is increasingly the dominant force. Unfortunately, I don’t see a solution for this. One possibility would be for the government to interfere with communication among VCs in order to kill off the collusion, herd mentality, and “accept this term sheet or I’ll pick up a phone and no one will fund you” extortions, but a government step-in would probably cause more problems than it would solve.

VC-istan itself may be a rank culture, but an imperfectly formed one. VC-istan exists because a web of socially connected people– reputable investors, and those with welfare-check (err, I mean “acq-hire”) writing ability at large companies– have created a controlled-market zone. Outside of that is regular ol’ business formation: an actual market. By the way, what are free markets? They are like self-executive cultures at best, but tough in their own way. Corporate tough cultures are mean-spirited and driven by intentional human malice (stack ranking, “calibration scores”, transfer blocks). Free markets are tough only because they are indifferent, but self-executive for those who’ve managed to develop a pattern of success. So the “greater world of business” formation is one that contains the market’s mix of self-executive and tough features at its periphery, and becomes more rank-culture-like as one grows closer to the VC-istan power players.

Startups are a reaction against the dysfunctional rank cultures, and would prefer to set themselves up as self-executive enterprises.However, as the company grows, power shifts to the rank culture of the world without (investors, acquirers). Full-on rank culture is usually in force shortly after liquidity, but the phase most typically associated with VC-darling “startups” is a transitional spell of (unplanned) tough culture. 

Startups tend to remain “flat” for a long time, but this is usually not out of executive altruism. Self-executive cultures are naturally flat, but so are tough cultures. Tough cultures want everyone in competition with everyone else and use a (pathological) flat model to maximize internal competition. The self-executive model of the VC-istan startup dies as soon as the company stops handing out real equity slices (often immediately after the Series A). At this point, the “colony” mentality (live or die as a group) ends and the true goal of the new hires is to get into the executive positions that make real equity slices possible. The internal competition that emerges causes alliances to form and break, influence to be peddled, and informal management (people who manage to win credibility through illicit trades and manufacture the appearance of high performance) will emerge. The tough culture’s informal management is actually substantially worse than the rank culture’s rigid system, because the former encourages more influential people (managers in practice, if not in title) to compete with their (again, informal) inferiors. Rank cultures, at least, are rendered stable by the absence of competition between managers and subordinates– leading, over time, to the Gervais hierarchy. Tough cultures are just amoral, vicious messes.

When a startup first gets funding, the cultural neighborhood of the founders switches from the tough one of an indifferent market to a more self-executive one. They now have some autonomy, measured numerically in “runway” (how long they can survive with current capital). Unfortunately, this self-executive culture cannot survive the rapid growth typically expected by investors (especially VCs). Self-executive cultures can only hire near the prevailing skill level. VCs, on the other hand, want the company to indulge in social climbing (hiring above that level, which involves enticing people with executive positions that reduce the autonomy of those within) and rapid growth (which mandates hiring below the prevailing skill level to tackle the grunt work generated by sloppy, fast growth). Guild culture is clearly not an option, because tight deadlines leave no time for mentoring. Nor is rank culture, because it tends rapidly toward the underperformance of the MegaCorps that startups exist to destroy. Thus, VC-funded startups tend to degrade into a tough culture by default. This should explain the churn-and-burn behavior for which they are so infamous.

So, what culture is VC-istan? It depends. If you’re lucky enough to win the attention and mentorship of the (extremely rare) well-connected person who’s not an asshole, it’s a guild culture. If you’re a founder who just got funding, or a very early hire guaranteed a strong position, it’s self-executive. For founders and executives fighting mostly external battles (with acquirers, investors, and others with a million times more in the way of connections) it’s a rank culture, but one that does not (at that point) seep into the organization. For typical engineers facing internal battles (for scarce future executive positions that come with real equity) it’s a tough culture of long hours, harsh deadlines, vicious politics, and fast firing.

In other words, startups go through four cultural phases. Before funding, they live in the prevailing, indifferent tough culture of the market. Once they get their initial funding and have some autonomy, they turn self-executive (second phase). The romance of the “startup ideal” is derived from this phase of organizational life. As they grow (often sloppily) and take on more people than they can really provide startup perks (real equity, autonomy, leadership) for, they turn into tough cultures (third phase). Most VC darlings have tough cultures. By the time the company is getting that much attention, the self-executive era has ended. Finally, once the startup reaches liquidity and is a full-fledged corporation, it tends toward typical rank culture (fourth phase).

Mike Testing VC-istan to predict its future

Let’s see if we can apply the Mike Test to VC-istan, noting that Silicon Valley emerged as a reaction to the stodgy rank cultures associated (at the time) with East Coast corporations. The Mike Test tells us that self-executive and guild cultures can improve as they grow, although these low-entropy work cultures are hard to grow fast. Rank cultures, on the other hand, tend to stagnate, and tough cultures actively devolve, as headcount grows. What is the future, then, of VC-istan?

Most VC-istan companies are in active cultural devolution, being tough cultures where people are prone to hire inferior versions of themselves. In software engineering, this process is slowed somewhat by the (genuine) desire not to hire outright incompetents. Software is structurally cooperative enough that the pain (bugs, bad designs, low code quality) associated with hiring an incompetent does not offset the gain in relative position. As I said, this software-specific trait means that B players don’t try to hire C players, but B- players, which slows the Mike-Test decay normally associated with tough cultures. All of this is, in fact, not a major problem for VC-istan. Companies are disposable! The thing should be sold before it falls that far. VC-istan can tolerate the corrosion of the tough corporate cultures it generates, since these companies are just going to be sold to rank-culture corporate behemoths. VC-istan is not designed to build free-standing companies. They exist to be bought or to die.

The tough cultures of VC darlings will bring incompetents into those companies, but not a rate that is fast enough to matter to the corporations themselves. They’re build-to-flips that have no reason to care about slow cultural corrosion. There might be a “littering effect”, if these tough cultures are (a) bringing undesirable people from outside of VC-istan, and (b) leaving them in the VC-istan ecosystem after the company fails or is bought. If both (a) and (b) are true, then we can hold these cultures responsible for lowering the overall quality of people in it. I don’t know if that’s the case. That subject needs further study. Since the most egregious incompetents in VC-istan are not low-level engineers but the supernumerary, non-technical, executives, I tend to doubt that any engineering-specific littering effect is a problem, either in the short or long term. The executive-specific one is well-documented; VC-istan does have a way of turning shitty, failed bankers into even shittier non-technical startup “executives”.

In other words, I don’t know if the tough culture that typifies the standard VC darling is going to fill VC-istan with incompetents. The company itself will take on undesirable people, but those companies aren’t built to last very long. That tough cultures are so common is a symptom of something pathological, but that’s a topic for another essay. Instead, to project the future of VC-istan, we need to look not at the culture of the typical VC-funded company, but at the culture of VC-istan itself: the postmodern corporation in which the VCs are executives, so-called “CEOs” are glorified project managers, and engineers are clueless chumps who don’t realize they work for a big company. What we see is a very typical rank culture, with VCs and well-connected “serial entrepreneurs” (read: people whose connections entitle them to continued funding no matter what happens) in the Sociopath tier and a swollen Clueless tier. It’s a very effective, internally stable MacLeod rank culture.

As I discussed earlier, the stability of the MacLeod organization comes from the lack of envy between Losers and Clueless; and between Clueless and Sociopaths. Differential social status (DSS) and the effort thermocline (the level at which jobs become easier, rather than harder, as one ascends) ensure this stability. Losers prefer things of genuine value over the non-transferrable DSS coveted by the Clueless. Clueless are oblivious to the effort thermocline and consider the Sociopaths above them to be harder working and more capable than they are. Losers who want to become Sociopaths are fast-tracked up or out of the organization. How does this work out in VC-istan? The effort thermocline is blindingly obvious: it’s the distinction between startup employees and powerful investors. Above the thermocline are the reputable VC partners, whose social connections entitle them to an easy life; below it are founders and the people they hire. The founders are the upper-tier Clueless who (just as in a MacLeod organization) don’t want the jobs above them, on the assumption that the Sociopaths “just shake hands and push paper around; we do the real work”. Differential social status, in VC-istan, is access to funding and publicity– something that a generation of idealistic idiots has spent decades of hard work to get, while their peers “sold out” and made actual money that could be used to buy actual houses and vacations.

For all this, I used to think that VC-istan didn’t have MacLeod Losers (with the losses borne by the swollen Clueless tier) but I realized that I wasn’t looking hard enough. Clueless are true believers, while Losers are rationally disengaged people who work hard, but minimize discomfort. (Again, they aren’t actually losers per se. Unlike the Clueless, they know the trade they make.) In VC-istan, MacLeod Losers tend to be consultants and freelancers. They cherry-pick the work that’s interesting to them, can lead quite a nice lifestyle if they’re good, and never work for equity. Since these people can often command $250 per hour or more, it’s hard to call them Losers! What they are is checked out of the Clueless game of VC-istan. They will never get rich– it’s hard to get more than 800 hours of work per year, with a decent rate, as a freelancer– but they will never lose a vacation or a relationship to a get-big-or-die company either. They sell tools and water to the idiots who dig for gold in the 120-degree heat. The MacLeod Losers of VC-istan are the mercenary consultants who manage to get by, contribute some work, but never hitch their fortunes or make undue sacrifices for a specific company.

If VC-istan is a MacLeod organization, then what is its future? MacLeod rank cultures decline, but they do it slowly. That is also what will happen to VC-istan. I couldn’t possibly say whether it will happen over the next 2 years or the next 20, but VC-istan is already in a MacLeod state and, while its decline is likely to be gradual, it’s also inexorable. I’ve started calling many of these companies “ad-banking”; VC-istan is what investment banking was in 1995. That is far from “death”. It does not even mean that it will become impossible to get good people; investment banks are able to get good people even now, but they pay dearly. It only means that good people will become more expensive over time. Only when compensation ceases to grow exponentially (in banking, circa 2009) will the best people start to trickle out and look for something new. VC-istan engineer compensation has quite a few years of 10-20% annual growth before reaching the unsustainable level, so I don’t see its “death” setting in until about 2025. Until then, there’ll be a lot of money to be made by mercenary engineers, so long as they know the market well enough to play it.

What will make VC-istan’s unwinding process interesting is its generation of alternatives. As investment banking grew stodgy, boring, and difficult, banks raised compensation (for genuine talent) into the stratosphere. No one would spend 15 years in New York investment banking for less than $400,000 per year at the end of it. This transfer of wealth generated early retirees and a few hedge funds, but not rival investment banks. People did not take their $6-million nest eggs and launch rivals to Goldman Sachs. VC-istan’s story will be different. As talented engineers wake up to the scam, wages may rise to the same levels that bankers commanded, generating a transfer of wealth toward a set of people not typically associated with richness: hard-working computer programmers. Those among them who have a mind for business will want to participate in investment, and to build something different from VC-istan. Something better. I don’t think anyone knows– yet– what will be built.

So what should be built? At 7.7 kilowords, I can’t hack that now. That’ll be covered in a future essay.

This is probably the second-to-last item in my series on MacLeod’s organizational hierarchy and the Gervais Principle. (See: Part 1, Part 2, Part 3, Part 4, Part 5.) I’m leaving the country for a little over a week, and intended my “tie it all together and solve it” post for today, but that will have to wait until later in the month. I’ve decided to take a detour into certain unanswered moral questions associated with the MacLeod hierarchy. What does it mean, morally, to be a MacLeod Sociopath? Is it necessarily harmful? (Answer: no.) What about Losers and Clueless? Are there not psychopathic Clueless out there? (Answer: there are.)

The Alignment model

As it were, the most useful classification I can come up with, in order to assess the MacLeod organization’s moral bearings, is the two-dimensional alignment system of many role-playing games, such as Dungeons and Dragons. These present a moral spectrum (good vs. evil) and a civil spectrum (law vs. chaos). These are independent of each other: one can be lawful and evil, or chaotic and good, for example. I’ll use that system to analyze the moral correlates of the MacLeod hierarchy.

The Moral Spectrum: good vs. evil

How do we define good and evil? It’s not an exact science, of course, but I think that most peoples’ definitions of “good” come from the so-called Golden Rule: do unto others as you would have them do unto you. As a general ethical guideline, this is a good one, and it’s the spine of almost all major religions. However, it’s simplistic and flawed, as I’ll address later on. Evil is miltant disobedience of that ethical principle, in favor of a “power does as it can” world where the winners gloat and the losers suffer. Most people, of course, are in between. They buy into the values of good, but sometimes indulge in evil behaviors as well. About 80 percent of people’s actual behaviors are in the “neutral” midsection of the spectrum, with 10% on each side being good and evil.

There’s a problem with the Golden Rule, which is that it fails to account for asymmetry, and that it can be warped to justify actions most people would consider evil. Perversions of the Golden Rule could be used to justify rape (asymmetry in sexual desire) and the rule has been (ab)used to justify religious persecution, it being better for people to suffer in this life than in the hereafter.

Trade, on the other hand, is all about asymmetry: comparative advantages, differences in value. It’s an action that seems, on some level, to oppose the Golden Rule, in that the two parties have exactly opposite financial outcomes. What makes trade “good” (not a violation of the Golden Rule) is the unequal value of the traded items for each party. As long as both gain, according to a difficult-to-define pseudo-quantity called “utility”, the trade is good.

At scale, there are very few actions that are good for everyone, resulting in debates over justice and politics, and attempts to resolve massively multilateral disputes through aggregation (voting, markets) that will drive general improvement, although it is impossible to make everyone happy. Ultimately, the Golden Rule falls in favor of the Silver Rule: do less harm to others than you do good. This represents the evolution from an inflexible but absolute good to a more flexible, pragmatic sense of “good”. Societies must favor the Silver Rule over the Golden one, in practice. Murderers must be jailed, and roads must be built.

The Silver Rule, however, is also flawed for computational reasons. How are good and harm measured? Gathering and processing information is an activity that itself imposes a cost (to others, but especially to oneself) which means that at some point, decision makers have to stop hearing all sides and just decide. This leads naturally to the Bronze Rule, which is: do your best, with the information and resources you can reasonably get, to do more good than harm to those you can credibly affect. This tends toward a more local sort of altruism that (inadvertently or intentionally) favors the well-connected. We’ve left the realm of the good and are now in the neutral-aligned territory.

The issue with the Bronze Rule and its tolerance of localism is that it enables selective goodness, because people can modulate how much weight they put into others’ concerns and how much effort they put into discovering what they need, and this leads people to favor those who are close (genetically, culturally, religiously, and geographically) to them.

The natural tendency of most humans is not to be egotistical or to be altruistic, but to be local. People want to confer benefit on those with whom they have personal affinity or similarity. Egoism and altruism are extreme points on a spectrum based on how people define their moral neighborhood. The extreme egoist defines it to contain only him, and the extreme altruist inclues all humans (or, perhaps, all living beings). Yet almost all of us are localist when it comes down to our day-to-day interactions with other people. It’s how we work.

Cognitively, most people know that inflexible or militant localism (which can tend toward racism, elitism, or jingoism) is wrong, but are not unusually energetic in pursuing the right. They give it a try, but it’s not crucial to them. They’re selective in how much effort they’ll put forth in the pursuit of good, depending on the affinity they have for the beneficiaries. That’s how the neutral alignment works. Good and neutral are localistic in practice; the difference seems to be in aspiration and energy. Good people will make serious sacrifices for others’ benefit; neutral will generally not.

If the Golden Rule is the archaic and idealized good, the Silver Rule is the practical good that accounts for asymmetry and massively multilateral decision-making. The Bronze Rule is the constrained, more austere spin that accounts for informational surface areas and human exhaustion, and generates the neutral alignment.

Finally, we have one more metallic ethical rule, the Iron Rule: take whatever you can get. This is the militant or even psychopathic egoism most commonly associated with “evil”. Actually, Iron-Rule psychopathy isn’t actually the extremity of evil. Beyond it are sadistic reaches that I don’t care to explore: people who actively seek harm to others, rather than merely tolerating it in the pursuit of their own needs. For this purpose, the sadists (as a class apart from psychopaths) aren’t important.

The Bronze Rule, I would contend, describes the state of nature. We are not naturally evil, egoistic, or psychopathic. Nor are we naturally good, universalist, or empathetic toward all. We made decisions (most, with an earnest desire to make the right ones) under extreme scarcity of information and with heavy influence (some of which is intractable) from the biological evolution that made us, and that makes us naturally localist. What generates the moral spectrum is where people try to go. The good aim for the Golden and Silver rules in their interaction with other people. The neutral tend toward Bronze Rule localism. The evil celebrate Iron Rule egoism or, worse yet, tend toward sadism.

The Civil Spectrum: law vs. chaos

The second dimension of the alignment system is the civil spectrum, which pertains to one’s approach toward authority and social stability. As with the moral spectrum, about 80 percent of people would be classified as neutral, with 10% on each side being lawful or chaotic.

Good and evil describe the direction that people, personally, try to take from our Bronze Rule state of nature. Of course, there’s another dimension, which is a person’s willingness to cooperate with authority. While good people will ultimately oppose an evil society, the reverse also being true, an overwhelming majority of complex societies are neutral, regressing to the mean as they get large. Thus, most peoples’ attitudes toward authority will often be more of a function of their personal biases toward law or chaos than of the character of the society, predominantly because large societies are not that different from one another in any morally meaningful way.

For personal ethics, the Silver Rule is to do more good than harm to others, with the tacit intent to take in as much information as one can absorb. Lawful people believe in analogous Silver Rule with regard to society, which is that authority will best aggregate the available information and do the right thing. (However, a lawful evil’s person’s definition of “the right thing” may be harmful to those judged not to matter. Lawful evil people place faith in society’s ability to decide who matters.) Lawful people do not believe necessarily that societies or organizations are infallible, but only that they perform far better than individual judgment.

Civilly neutral people believe that societies implement the localist Bronze Rule. Organizations and those who hold authority may try to do their best, but are limited by their informational surface area and limited time and energy. Ultimately, those who are close to those in power enjoy an advantage, also known as corruption. It’s not that organizations tend toward self-serving pathology or even intentional elitism, but that a certain degree of localism is inevitable and mostly tolerable. Civilly neutral people believe that societies tend to be no better or worse than individuals.

Chaotic people believe that those in power, in most societies, will follow the Iron Rule. They distrust authority, believing that power will almost invariably be used toward bad ends, and that those who are in control will take whatever they can get. Neutral people admit that civil authority can tend toward localist corruption, but chaotic people believe that organizations tend toward defectiveness. Chaotic good people believe that authority will, over time, lead to evil. Chaotic evil people view those in civil power as contemptibly incompetent.

Social acceptability

Lawful good represents what people are “supposed to” be, ideally, while the central “true neutral” alignment is what they actually are. Actually, I would argue that neutral may be the wrong term, since people seem to be, by default, weakly good and weakly lawful. Anyway, what all of this means is that people who are lawful or neutral on the civil spectrum, and good or neutral on the moral spectrum, fall into a category that people are familiar and comfortable with. This 81 percent of the population will generally have no difficulty following social norms.

What remains is an L-shaped region (19%) that contains the chaotic or evil. Chaotic people face above-normal rates of social rejection. Evil people are punished and despised– if they are caught. Chaotic evil are the pinnacle of dysfunction, and only succeed amid severe environmental disorder. One example (from Final Fantasy VI) is Kefka. His chaotic evil (as opposed to Emperor Gestahl’s lawful evil) renders him an incompetent nincompoop in the (ordered) World of Balance, but he becomes a demigod in the (disordered) World of Ruin.

In general, the utter social dysfunction of chaotic evil (1%) divides the “L of social unacceptability” into two separate islands, each of which can be socially functional under some circumstances. One contains those who are evil but lawful or civilly neutral (not chaotic). These people can succeed socially as long as they can move faster than the consequences of their actions catch up with them. The other contains those who are chaotic but good or morally neutral (not evil). They can succeed socially as long as they are in environments that recognize the benefits of disruption and that value creativity over uniformity.

I describe the chaotic crowd as “those who wear hats”, using the hacker terminology where good guys wear “white hats”, the neutral wear “gray hats”, and the bad guys wear “black hats”. Wearing a hat (of any color) indoors is, at least traditionally, socially unacceptable. The lawful and neutral take them off. An environment that tolerates hat-wearing is one in which the chaotic can thrive.

This explains my desire to split the MacLeod Sociopath category into two. People use “sociopath” to describe those who live in this “L of social unacceptability”, the chaotic good radicals being “good sociopaths” (after they are recognized as good, the reality being that most people cannot parse chaotic morality in its own time). In my view, this deserves further exploration.

Psychopaths are, as I’ve defined the moral spectrum, evil. That doesn’t mean they participate in evil’s most brutal manifestations, but they devalue others’ needs, gains, and losses outright. Technocrats are chaotic by nature. Rather than gaining power through typical social means (dues paying, credibility, deal-making) they attempt to create radical and new forms of power. The goal is to take superior craftsmanship, art, science, and knowledge (techne) and turn that into influence, wealth, or power (kratos). That is innately disruptive to those who are vested in the old forms of power.

This does not exclude the possibility of “black hat” Technocrats forming an organizational presence, but my experience is that chaotic evil people very rarely move into positions of power or importance. They are just too socially dysfunctional. Complex societies will form subcultures that give chaotic good and chaotic neutral people second chances… but chaotic evil people only seem to acquire power in damaged environments.

The MacLeod pyramid

With this understanding of alignment, it’s possible to approach the MacLeod pyramid in the context of the moral and civil spectra. Perhaps not surprisingly, the civil spectrum is more correlated to it than the moral one.

MacLeod Losers, at the base of the pyramid, tend to be civilly neutral. Whether they are morally good, evil, or neutral doesn’t matter much to the health of the organization, because they have very little power. Since they view the organization as a Bronze Rule localist organization (not a Silver Rule, omnibenevolent meritocracy) they have a take-it-or-leave-it attitude and will show loyalty so far as they’re accorded social status, stability, and comfort. The Clueless, predictably, tend toward lawful alignments, but can be anywhere on the moral spectrum. Organizations actively try to make it this way. They don’t especially care about good versus evil in grunts or middle managers, but they want rules to be blindly enforced when necessary and blindly broken when authority requests it. 

If corporations could consciously choose leaders, they’d generally want people who are morally and civilly neutral, because that’s what most organizations are. Neither an overbearing lawful, chaotic, good or evil bias is beneficial to the organization’s objectives, and all can be harmful. Additionally, 64 percent of the population falls into that “true neutral” category. So it seems like the desirable set is a large one. However, rapid organizational ascendancy is abnormal. It breaks the rules of the on-paper pseudo-meritocracy, and it favors the stand-outs, who tend to fall into one (or two) of four categories:

• those who exert above-normal energy for the benefit of others, the organization, and the world (good).
• those who exhibit an unusual ability to conform and subordinate (law).
• those who will do anything, even harm others, in order to acquire power (evil).
• those who pursue disruptive and possibly anti-authoritarian avenues toward creativity (chaos).

In general, stand-out good people don’t get promoted. They get more responsibility, but not power. Stand-out lawful do, but at a plodding pace through “front door” avenues, and rarely past the effort thermocline. This leaves the evil and the chaotic, who tend toward variability because organizations just don’t know what to do with them. They exhibit an “up-or-out” distribution of organizational success. They’re either promoted or fired. (Sometimes it’s both.) They are the only ones who can pass through the effort thermocline.

A fully self-conscious organization desires neither evil nor chaos, so people judged to exemplify either are usually expelled from it (fired). the only forms of these that survive are those that manage to “trick” the organization enough to go undetected. Of course, this only reinforces the bias toward the promotion of evil or chaos, since deception is usually motivated by one or the other.

The surprising (sociopathic?) result is that the most successful people will come from the “L of social unacceptability”. Organizations, to the extent that they are conscious, try to exclude them. The result is an arms race between such people (as they fight to get as much out of organizations to survive or coexist) and the organizations, as they strive to improve their detection of law, chaos, good and evil. The winners become leaders; the losers get fired.

Eligibility pools

People who are lawful or civilly neutral (90 percent) are eligible for Loser-level roles in organization. Those who are lawful (10 percent) are eligible for Clueless middle-management positions. These numbers correspond roughly with a typical organization’s needs at each level. At the upper-tier, executive level, there’s a surplus. The “L of social unacceptability” (19 percent) is much larger than the organization’s needs for executives, so it can be selective. It can favor chaotic good, chaos, evil, lawful evil, or even chaotic evil. It gets to pick. In theory. In practice, almost no organizations exhibit anything like conscious, rational “thought”, so the selection is likely to be subconscious and by default.

Evil, I would say, is almost never desirable. Even if we were to judge evil to be necessary (with which I don’t agree) the darker shades of the moral neutrality can usually be coerced into it, especially if they have a civil bias. Lawful neutrality will support evil laws, and chaotic neutrality will oppose good rulers. Although many corporations devolve into macroscopic evil behavior and internal strife, and plenty of them are used for evil purposes, I still contend that even the most evil owner or executive would prefer not to have evil lieutenants. (Lawful neutrality is more desirable in a subordinate.)

The forward-thinking leaders that companies (if they are to remain adaptable in a chaotic world) should want, then, are the chaotic good and chaotic neutral– the Technocrats– with a hand-over to civilly neutral people as the organization grows. What remains an open question is which of these two alignments is to be preferred. That one, I would have a hard time to answer. I am (for obvious reasons) in favor of chaotic good, but I tend to think that chaotic neutrality may be more adaptive. In the rare case where a chaotic individual obtains power, the chaotic good person will limit her own power and create a system of checks and balances, creating a tougher job for the next generation of power holder in that institution. Chaotic evil people will abuse power so flagrantly that others will rush in to halt them. So chaotic good and evil both lead toward the reduction of power. It may be that chaotic neutral people (who just don’t know what to do with power) are the best ones for an organization to have hold it, because they are more likely to transmit it untouched to the next generation.

Conclusions

It seems that the worst pathologies of the MacLeod hierarchy come from the tendency to favor psychopathy at the top layer– the one called MacLeod Sociopaths. It is psychopaths who continue the dishonesty that deludes the middle layers (MacLeod Clueless) and the poverty that depletes the workers (MacLeod Losers). However, organizations create such bureaucratic walls that only stand-outs, rule-breakers, and tricksters can get through them. That favors evil or chaos (with those who exemplify both often being too pathological to succeed). It seems that organizations are doomed to have one or the other become prominent within its leadership. Therefore, the best antidote toward psychopathy (evil) might just be an increased tolerance of chaos.