On Tuesday, I wrote a post that raised questions about New York Times statistical ace Nate Silver’s recent analysis of campaign spending. In that post, Silver argued that Karl Rove’s new super PAC created to defend mainstream Republican candidates against Tea Party challengers was unlikely to make much difference—because, according to his examination of primary Senate contests over the last two cycles, mainstream GOP candidates already hold a significant financial advantage over their hardline opponents.
But Silver’s analysis omitted what I believed was essential information: the expenditures by outside groups such as super PACs and nonprofits. Instead, he considered only traditional campaign donations in his data. In addition, the story included two references to super PACs making “contributions” to candidates—a practice that is in fact illegal. I, and others, wondered if Silver had misunderstood the rules governing outside spending groups and believed that they contributed directly to candidates.
Since then, the language of Silver’s post has been tweaked to remove one of the references to a super PAC “contribution.” (A related clarification has been added to the bottom of the post.)
And on Wednesday we heard back from Silver, who offered an explanation of his thinking. Silver noted that he did accurately understand the legal restrictions on super PACs. And in response to the central question about why outside spending was omitted from his analysis, Silver replied that he did not believe including those figures would have made a substantial difference in his findings.
Here is what he wrote:
With respect to spending by independent groups like Club for Growth in past primaries: one issue is that the record-keeping is not all that great, especially as compared to FEC reports for public financing which are very good. But from the spot-checks I did, it also seemed as though their contributions were quite small in most cases, especially in the 2010 cycle, and that the “insurgent” candidates were still at a very substantial financial disadvantage in most cases.
Endorsements from these groups might have been valuable in securing favorable media coverage for an insurgent candidate, and might have been valuable to the campaigns in other ways. But part of what the piece is arguing is precisely that these “intangible” contributions can outweigh paid media campaigns on behalf of a candidate….
The editorial issue here is one of how much detail to include in a 538 article. As a rule, 538 posts are quite detailed as compared to most blog posts, and reasonably detailed as compared to many or most “traditional” news articles. But there are judgments involved in terms of when including too many qualifications or parentheticals will district from the main thrust of an article, particularly when (as I assert is the case here) the additional details do not much weaken the thesis of the article.
It’s clear that Silver thought out his decision carefully, but I still believe his approach is incomplete. Past expenditures by outside spending groups are essential to a data-driven argument about the likely future impact of an outside spending group, and outside groups in fact played a significant role in the latest Republican primaries.
Silver also sent us a link to the hardline Club for Growth’s spending records in the 2010 cycle, in which the group spent $5 million in total, almost all in general elections. But a recent CJR analysis based on data then available found that in the 2012 cycle, Club for Growth Action, the group’s main campaign affiliate, spent $16.6 million in total, with nearly half of that—$7.9 million—going to Republican primaries. Three-quarters of these dollars backed the winning candidate in these primaries.
And outside spending can meaningfully shift the balance of campaign spending power between candidates in particular races. Silver’s post shows the moderate Republican incumbent Richard Luger outspending his insurgent challenger Richard Mourdock by a margin of $5.5 million to $2.1 million. But according to a CJR analysis of data from the Sunlight Foundation, Club for Growth Action alone spent more than $1.5 million either to attack Lugar or support Mourdock—who prevailed in the primary before losing to a Democrat in the general election.
Silver’s post does include thoughtful points about the limited effectiveness of campaign spending beyond a certain threshold, and about the potential that Rove’s high-profile push could “backfire” by making insurgent candidates appear more sympathetic to right-wing media and hardline donors. Those points may hold up after a fuller analysis. And his note about the relative difficulty of working with data on outside spending is well-taken—for example, while reviewing Sunlight’s data this afternoon I found different values in different places for the amount Club for Growth had devoted to attacking David Dewhurst, the mainstream GOP candidate for a Texas Senate seat.