Conceptual Map Connecting 2008 Presidential Candidates to Supporters’ Other Causes
One way to learn about an election campaign is to learn something about the people who support a particular candidate. One useful piece of information to gain is simply the number of people who actively support a particular candidate during any one period of time. This is the sort of horse race that mainstream media typically follow. But we can also delve a bit more into the sorts of people who support different candidates. In the past, for instance, we’ve looked at the candidates supported most often by residents of each of the fifty states. All of this information is gained from our analyses of bumper sticker, button and t-shirt sales at our Vote Democrat 2008 webshop.
We’ll look at these sorts of information again in the near future, I’m sure. Today, however, I decided it would be fun to look at the instances in which someone who ordered a sticker, shirt or button in support of a 2008 contender also bought another item — a bumper sticker, button, or t-shirt not directly related to the 2008 election at all. Alternatively, someone might have ordered a sticker, shirt, or button in support of a second 2008 contender as well. Each of these joint orders represents a conceptual tie, showing affinity between support for a particular candidate and support for some other idea or candidate. When we look at all the ties put together, they create a conceptual network (or, in graphical form, a conceptual map) that shows us the structure of ideas driving the 2008 presidential campaign.
Without further ado, I’ll share below just such a conceptual map. It is generated from Vote Democrat 2008 orders in April and May of 2006:
Candidates are represented by blue circles, and non-candidate causes by red circles. The thicker the line connecting circles, the more often people make orders for stickers, buttons or shirts related to both circles. When there is no line connecting two circles, there is no case in which someone buys items related to both circles. Only the more popular candidates are here with their associated causes, since unpopular candidates don’t produce a reliable enough picture of related causes (a “low n” problem).
At first pass, what’s interesting to me is the pattern of connections between candidates — the times when someone buys, for example, a bumper sticker in support of Hillary Clinton and then another bumper sticker in support of…. Oops! Bad example. You see, there is no case in which someone buys an item in support of Hillary Clinton, and then buys another item in support of another candidate. In other words, people who support Hillary Clinton seem to support her candidacy alone. John Kerry and Wesley Clark show the same pattern — people who support their presidential bids don’t support any other. On the other hand, people who have bought pro-Boxer, pro-Gore and pro-Obama stickers have also bought bumper stickers that support a candidacy of John Edwards. Another “cognitive star” in this network is Barack Obama — people who buy bumper stickers in support of Bill Richardson, Mark Warner, Joe Biden, Barbara Boxer and John Edwards also buy bumper stickers in support of Barack Obama. Is this some sort of way of hedging bets? Barack Obama is seen commonly as a “dream candidate” who won’t really run. Perhaps individuals buy a “dream” sticker and also purchase a “realistic” sticker to boot.
Turning to focus on non-campaign causes, some are especially well-connected to multiple campaigns. For each and every 2008 candidate, there is a case in which someone bought a sticker, shirt, or button supporting them and at the same time bought an item stating opposition to George W. Bush. Nearly as commonly tied to campaigns is the cause of peace (or opposition to war). Liberty, on the other hand, is an outlier, only connected to the candidacies of Russ Feingold, Wesley Clark and Al Gore. Health care is nearly an isolate, with an item in support of that cause only jointly purchased by a Kerry supporter. Finally, some causes, while associated with more than one campaign, are more strongly associated with one particular candidate than with others. This is most clearly the case with Al Gore, whose supporters also purchased huge numbers of items expressing support for environmentalist causes and interests. This is no surprise, given the impending release of Gore’s movie on global warming, An Inconvenient Truth.
What patterns do you see in these results? What do they tell you?