As I read the article about knowledge spillover dynamics, I started thinking about how knowledge spillover works in my dance group. We do international folk dance, and the liveliness of our group depends on whether people know the dances. We have a repertoire of hundreds of dances, but only a handful of people are prepared to lead dances, and that handful typically only leads the dances they like. The rest of us usually follow along in a long line. I've been part of two such communities in the last year, and it has been interesting is to watch what happens when there is a large international dance community as compared to a small one. When I moved to Boston last year I was excited because it is one of the meccas of international folk dance. I thought I could take advantage of high turnout and resulting knowledge spillover because there are over a thousand adherents spread out in the metropolitan area and beyond, as compared to Madison, which might have a hundred or so. However, when you compare the weekly and monthly evening dances in the Boston area to those in Madison, you'll find that the quality and diversity of the dance repertoire in Madison are regularly better than in Boston. It turns out that when you have a large dance community, there are more dance events, so the leaders (the knowledge sources) only go to a select few dance events. The other events struggle for energy, and often dances are requested that have no leader. Dance event coordinators from closer to the Boston area often envied those that lived further out because further out dances would regularly be well attended by a local contingent. That contingent would then develop a core repertoire familiar to most members, and (here I'm conjecturing) a larger percentage of members would take on leadership roles, including dance repertoire maintenance and expansion. This appears to be the opposite finding from Audretsch and Feldman's article, where more intense competition and spacial proximity created a richer response.
Perhaps the reason that Audretsch and Feldman's findings don't hold out here, besides the fact that I'm analyzing a radically different situation, is that I'm looking at it in the wrong scale. When you analyze at the size of a city full of dance groups, the competition is detractive to attendance. When you analyze at the level of a group; however, where competition occurs between dancers, and where proximity is measured in how often they dance together or have danced together, I think these factors do add up to a richer setting and more leadership. Friendship and shared dance skill sets can take the sting out of showing off and competitive behavior, and instead of dividing the group into leaders and followers, it will improve the overall atmosphere. This may be why the Madison group is so strong, and why a few groups that are defined by social unity more than geographical proximity in Boston seem to be more energetic. When we bring the inter-group competition for attendance back into consideration, we find that places with high dancer populations see more stratification in dance events based on skill sets than those with low dancer populations. Dancers will go to the dances their friends are at in Boston, and their friends are probably similarly skilled in dancing. This makes the atmosphere more attractive socially but can lead to problems when the the leader peer group does not attend. Madison does not see this trend because we all attend the same events, and therefore our peer group is based on geographical restrictions instead of skills sets. I would argue that that makes the Madison community stronger overall because knowledge is spilling over from those of high skill to those of low skill regularly. However, we do have a smaller total number of leaders in Madison than in Boston. It is hard to say if we have maintained our repertoire better than the Boston community however. It is possible.
Returning to innovation, I wonder if we could explain the successes of these industries with geographical focuses using peer groups. Could competition occurring within peer groups be more valuable? and geographical proximity allows for more specialized and more stratified peer groups to emerge? This would bring out a larger total number of more skilled people in one place, but could a weaker overall community result? I guess it seems unlikely, but I'm kind of curious what other factors are involved. The fact that people are employed in these industries instead of simply pursuing a hobby probably makes a big difference. Audretsch and Feldman's factor of access to innovation through institutions like universities also would make a difference, and I could probably write even more about that, since international folk dance groups are almost always in cities with big educational institutions, but I think this has become long enough. Thank you for reading.