63% time bound loss of productivity due to multiculturalism

From The Authors of The Org Answer Your Questions by Stephen J. Dubner

The question asked is whether there are productivity related costs to organizations/departments working together who have distinctly different cultures. While this is in the context of business, it does relate to the much larger issue of immigration and assimilation - what is the optimum or maximum degree of cultural variability you can accommodate within a society and still function productively?

A couple of professors conducted a behavioral lab experiment to begin to quantify the nature and magnitude of some of the issues associated with merging two organizations. Their paper is Cultural Conflict and Merger Failure: An Experimental Approach by Roberto A. Weber and Colin F. Camerer.

It is a behavioral lab experiment with all the caveats and skepticism that appropriately go with that.

The findings? Read the article for the details of the experimental structure.

As they repeated the task, the pairs found that IMing produced a shorthand that helped them identify pictures more quickly and accurately. Camerer and Weber give the example of one manager who, in the first round, describes a picture as “The one with three people: two men and one woman. The woman is sitting on the left. They’re all looking at two computers that look like they have some PowerPoint graphs or charts. The two men are wearing ties and the woman has short, blond hair. One guy is pointing at one of the charts.” A few rounds later, the description is abbreviated to “PowerPoint.” It took just a few rounds to get vastly more efficient, and after twenty rounds, pairs were able to reduce their time from over four minutes to less than fifty seconds.

You can think of that shorthand as a kind of culture. Then Camerer and Weber disrupted these nascent cultures by adding a second employee under each manager. Each group then repeated the task with the established pair plus the newbie. Managers were paid based on the average speed of their two charges, so most stuck to their “PowerPoint” shorthands to communicate with their old partners. Why not? It had worked before.

But the old methods didn’t work with the new addition. The new management groups argued about how to identify scenes. Predictably, frustration followed (“I don’t care if they’re wearing ties – just tell me if you see the PowerPoint!”). It wasn’t pretty.

But, after playing a half-dozen times or so, the new groups became accustomed to a merged culture, or shorthand, and most got their times back under a minute.

This ties a lot of interests that I have posted about in the past together. Are there productivity differences resulting from the structure and nature of different languages? (See the post, The English language hasn't got where it is by being pure) What are the costs associated with integrating two cultures? To what degree is culture a mechanism for distilling complex learned experiences into heuristic (thus lowering epistemological costs) instructions for increased productivity (and therefore survival)? What are the productivity returns to increased practice?

The numbers from this experiment are at best indicative but interesting none-the-less. So twenty cycles of practice led to an 80% improvement in productivity (over four minutes down to forty-eight seconds). When merged (two thirds old group and one third new group), they went from forty-eight seconds to 130 seconds for the first round post merger. As with the first firm, they were able with multiple rounds of practice to regain their former productivity. The implication, for business at least, is that there is potentially an exceptionally high loss of productivity at the time of a merger, which given exogenous circumstances, might imperil the enterprise. Anybody who has been through a merger knows this to be true experientially and the fact that the majority of mergers fail provides additional evidence.

Of course the experiment misses all sorts of critical nuances. When you have a merger of complimentary capabilities, sometimes the whole is greater than the sum. Sometimes, even though different, the cultures have enough in common to ease the transition and mitigates the loss of productivity. Not infrequently, a merger of cultures introduces enough variability to strengthen the enterprise (variation leading to learning and adaptation) without threatening it. Ad infinitum.

Still, 63% lost productivity is an interesting benchmark and provides at least some sort of ballpark figure for assessing the likely impact of a merger of two cultures.

When you are speaking of nations rather than businesses, and reality rather than lab experiments, 20 rounds is probably 2 generations and the loss of tactical productivity is paid before the possible strategic benefits become apparent. Hence the drama around multiculturalism and immigration.