A quick observation. To me, the most interesting subtext of the pundit/quant pissing match that erupted in the late days of the campaign is the similarity it bears to the hoary aggregator/reporter handwringing we’re all so used to.
Nate Silver is heavily dependent on publicly available polling data, much of which is funded by media organizations. Those organizations engage in polling both to allocate resources and — probably more importantly — to generate grist for stories. Attracting attention to stories written about the back-and-forth of individual poll results is an important part of how these polls are paid for. Silver’s work not only implicitly and explicitly tells us that such stories are a waste of time, but offers a substitute product that can be consumed more efficiently (and which, for the time being, has the benefit of novelty).
In both this and the familiar “oh no Buzzfeed/HuffPo/DCist/whatever” case, readers are being offered a distilled, intensified information product that has been created from more basic inputs. Those inputs are paid for by the value that their creators — media companies — derive from advertising. Aggregators capture some of that value for themselves. To the extent that all of this is zero-sum, the producers of the inputs feel that the aggregators are free-riding and endangering the entire ecosystem. I think that this explains a lot of the animus directed toward Silver.
In the case of content aggregation, we seem to have reached an uneasy truce. And good for us! But this is more reflective of the media’s admirable commitment to openness than it is of a sustainable equilibrium being figured out — though of course the increasing amounts of original reporting done by the aggregators is encouraging.
Perhaps poll aggregators will follow a similar route and begin collecting their own data, though this seems somewhat unlikely given the startup costs involved. Perhaps more plausibly, one could imagine a growing dependence on data gathered by universities. Everyone wants to be a research institution, after all — prestigious new polling operations seem like a viable destination for ballooning tuition fees, shifting the source of Silver’s received subsidy from media to the education sector. It also seems likely that polling will increasingly find ways to use the huge amounts of data that voters now accidentally generate on a continuous basis, and aggregators like Silver might be able to do some of this themselves thanks to the problem’s amenability to automation. (Notably, the price that this data’s owners will set still seems unclear.)
So I think we’ll be fine — to the extent that we need polls, I suspect that our need will be satisfied. (Though that extent is unclear — campaigns will continue to pay for private polls to guide strategy, of course, but I suspect that poorer information for election-watchers might be utility-enhancing, since uncertainty is entertaining.) But I do think it’s interesting to view the current, superficially inexplicable animus toward Silver through the more familiar lens of what’s been happening elsewhere.
UPDATE: Another perspective. I think a preference for simplistic accounts is made possible by structural factors like those discussed above, though.
ALSO: Seems like Henry Farrell wrote something very similar just after midnight this morning — I’ve only just read his post now (11PM) but it’s certainly worth a look. I swear I’m not just ripping him off! Simultaneous invention! Zeitgeist! Etc!
