Thursday, September 18, 2014

Corporate Culture

What does culture, corporate culture no less, have to do with game theory? At first blush the two worlds could not seem further apart. Game theory, with its bizarre concern for turning every single social interaction, from exchanged smiles in the morning to (at least in Mad Men world) exchanged gropings after hours, into a series of tables, trees, and math, seems a faintly ridiculous place to look for insights about something as amorphous, fluid, and nuanced as culture. Yet, perhaps in spite of itself, game theory has something (maybe a lot) to say about this topic.

Ask an economist about culture in the workplace and he or she (mostly he) will respond with stories about the importance of getting incentives right. Such a response seems faintly insulting as culture is obviously much more than a series of carrots and sticks used to induce corporate donkeys to trod some dreary path carrying their pack.

Yet this is correct, to an extent, for cultures, however noble, meaningful or even godly, founder quickly in the face of bad pecuniary incentives. Happily for us (though not necessarily for the individuals so described), tens of thousands of individuals in the US during the first half of the 19th century became unwitting test subjects for examining this hypothesis. At that time, as apparently at all times, people were sure that civilization was going to hell in a handbasket, that godliness, respect for others, politeness, manners, and the work ethic were all pale shadows of their former selves. In short, many were convinced that civilized society was breaking down or in crisis. They were absolutely convinced that their sons and daughters, high on the freedom (or perhaps stronger spirits) of America's rapidly advancing frontiers, were in the process of sending taking a giant step backwards in terms of civilized society.

(A sidenote: These sons and daughters were, in all likelihood, drunk a great deal of the time. The combination of perishable grains, long winters, bad roads, and knowledge of distilling proved a potent "cocktail" for individuals living in such wild frontier places as Ohio, western New York or western Pennsylvania, far away from the Land that Time Forgot, which lies to the east in the Keystone state. Having a grog--or several--before breakfast, at meals, and at breaks was considered perfectly normal. Rather than having a coffee break, workers would stop for "dram breaks" several times a day. This no doubt contributed to the violence, especially domestic violence, of frontier life, as well as the prevalence of workplace accidents, and possibly, to some degree, the remarkable fecundity of the population, which grew at 4% per annum without any considerable influx of immigrants.)

Anyway, back to our story. Rather than merely bemoaning the sad state of civilization, many people formed utopias--societies set apart and usually dedicated to some prescription or other for the well-led life or for a right and just society. Often these prescriptions were religious as their was a tremendous religious revival occurring at the time, mainly consisting of the formation of new Christian sects. Sometimes the prescriptions were the product of reasoning and "science" by (mainly pseudo) intellectuals . Many of these utopias saw money and, more broadly, property as the root of the problem and banished it from their communities. All property was joint. All production was to be shared. Individuals should take as they need and work according to their ability and in whatever field they thought best. There were also stringent social rules tight control over sex, drinking, profane language, and other behaviors deemed as societal ills. Incentives, so far as any existed, relied purely on social and godly rewards and punishments. Such societies would have little use for our typical economist above, except possibly as a strong back to contribute to crop growing.

Overall, the results of these utopias were...awful. Most fell apart within less than five years of their founding, often ending in bankruptcy, numerous legal battles, and sharp acrimony. Utopias frequently ended up starving since most members preferred to write manifestos about the great deeds of their utopia rather than engaging in (still labor intensive) food production, house production, animal husbandry or any other task likely to produce the necessities of life. Laziness in general was a constant problem as many people would happily do nothing whatever whenever possible.People ate too much from what little food was produced. And worst of all people bickered constantly about what (or in some cases who) was theirs. The absence of currency or formal property rights did not mean that individuals gave up the pursuit of "stuff." Quite the contrary as individuals spent a great deal of time scheming and conniving to acquire squatting rights more and better shelter, food, furnishings, and so on.

There were, however, some successes. In New England, one utopia was, in fact if not in name, nothing more than a company mill town run entirely from the labor of young women. Their parents entrusted them to the manager/mayor/chief priest of the utopia (an older and richer man, of course), to keep them out of trouble and earn some money for their families until they married. To be precise, these women worked hard--very hard--and were given food and board in the utopia in exchange for their labor. They also had to adhere to rules on pain of being "sent home." The most important rule was that, other than the manager/mayor/priest, there were to be no men in the community, either as members or visitors. A cynic might see this "utopia" as little more than a scheme to take advantage of cheap, unskilled labor under a mere facade of societal improvement.  Nonetheless, it clearly met some sort of need for New England families to make some money and not have to worry about protecting their daughter's "virtue."

Another notable success were the Shakers. They were a more traditional religious utopia founded by a woman who believed herself to be the second incarnation of Jesus Christ. Her charisma, combined with a strong practical streak---well spelled out rules and the ultimate sanction, banishment and damnation by eternal hellfire---caused the place to run pretty effectively. Ultimately, it was undone by a central rule--celibacy--no individual, under any circumstance, could reproduce. Punishment for doing so, or even performing certain activities that might, if the stars were aligned, lead to reproduction, was exile from the community, stringently enforced. The Shakers lasted surprisingly long given this stricture. They also left us with a nice style of furniture.

But back to the main plot---ask an economist about culture and hear about incentives. Well, it seems not to be entirely nonsense. Incentives do matter a great deal to culture, despite the claims of many intellectuals even today. .

Ask a specialist in organizational behavior (usually trained as a psychologist) about culture and you'll receive a much different answer. She will talk about the importance of empathy, transparency, safety in the frank exchange of feelings, trust, and other such behaviors. This is not to say that these people do not believe in incentives, they do, but tend to call them by different names than economists. To give but one example, most social psychologists believe in a theory called "relational equity" and argue that good cultures are marked by relative balance in relational equity accounts of the key actors. According to this theory, each side keeps an account of the good deeds done for the other (and presumably offsets these debits with credits for bad deeds though this is not much talked about). A relational equity account is balanced when goodness, measured somehow, is approximately equal. Things break down when inequalities persist or grow. It seems intuitive that I might stop being such close friends with someone to whom I grant a string of gifts and favors while receiving nothing back in return. But psychologists think things run the other way as well. I may wish to divorce a friend who is "too nice," someone who endlessly does me good turns at such a rate that I cannot possibly keep up in reciprocation. Thus, both not nice enough and too nice present problems. At any rate, some version of incentives runs through much of the literature on leadership and culture though the incentives tend to be of a more amorphous and personal character rather than the rough and ready dollar variety that economists like.

So where does this leave game theory?  One central insight of game theory is that the same set of external and internal incentives can produce wildly different cultures depending on beliefs about the choices (and feelings) of others. Put differently, the same game (i.e. set of actions, payoffs, preferences, and information) can give rise to multiple cultures/equilibria depending on beliefs. Moreover, many things influence these beliefs. Past history, the shared set of company values, the types of individuals recruited, social norms, and so on are all influencers. In game theory, these features are sometimes referred to by the shorthand of focality, circumstances that make one set of beliefs about the actions others will take more likely than some other set of beliefs. While the list of focality influencers offered above are all concepts our OB specialist would readily recognize and endorse, our economist can readily support the idea of changing the game, or even just changing how the game is presented.

Let's make this concrete: Consider the archetypal game Stag Hunt: Two parties (or perhaps many parties) must choose between an ambitious but risky action that requires them to work together to pull off or a safe, but lower payoff action that does not. Since innovation lies at the heart of the long-term success of nearly any company, most leaders would probably wish to encourage their employees and business units to choose ambitious actions at least some of the time.  One school of managerial thought suggests that we treat employees as "owner/operators" and hold them closely accountable for their financial performance, typically measured at quarterly time intervals. Thus, if an employee or group/division/branch, etc. does well---meets its numbers---it is rewarded, if it "falls down" it suffers some sanction, and if it does something really great, perhaps an extra award might accrue. Trouble arises in determining how much this extra award might be. While firms have a good sense of the value of normal operations and can reward accordingly, the value of an innovation is rarely immediately apparent and, consequently, a firm might, with good reason, be hesitant to reward it lavishly. Returning to our stag hunt, this implies that gap between the payoff from meeting the numbers versus missing is likely larger than the gap between the payoff from successful innovation versus meeting the numbers, at least in the short run.

But note what our firm, following "best practices" has done---they've unwittingly made it very risky to undertake that ambitious project. If the project requires cooperation across several managers, then the firm has, in effect, given veto power to their most risk averse manager. Not exactly a prescription for innovation.

It needn't be this way and, in fact, probably was not this way back in the firm's formative years. At that time, the firm was small, everyone knew one another closely, working together nearly all the time in the firm's startup phase, and, moreover, ambitious projects were undertaken and were successful, perhaps those projects are the reason the firm is now big and faces this problem in the first place. One could simply write off the two situations as a difference in the personalities of the managers and leave it at that, arguing that those early manager/founders were extraordinary and so they pulled the ambitious projects off whereas the current crop are not made from the same stern stuff. This might be true, but is hardly useful to a firm that wishes to be innovative.

Finally, we are to the heart of the matter. There seems nothing wrong with the incentives of the firm as success is rewarded to differing degrees and failure punished. There may be something wrong elsewhere in the culture, the wrong people, the wrong way of fostering interaction on intergroup projects, and so on, but we have little guide where to look. Treating the cultural situation as a game, however, offers some prospects. From this analysis, we might discover that our incentive system is set up to make safe play "risk dominant" and endeavor to fix this. Curiously, one fix would be to make the incentives less high-powered, to make it okay to fail, perhaps even rewarding failure . We might discover that managers across groups are simply too busy, too productive, to get to know each other so as to form sufficient confidence that promises made by the other will be carried through. This suggests a different set of fixes including retreats, high performer training, or even something as simple as social events for managers.

The point is that, by thinking of key challenges in terms of a game, we gain deeper insights into the pathology of the problem and hence a much better idea of which of the many solutions on offer-- monetary, social, coaching, trust, transparency, and so on--to choose. In short, game theory provides a lens through which to understand our culture better.

No comments: