Who's Watching the Conventions?

The following is a guest post from Jonathan Ladd, an associate professor at Georgetown University and the author of Why Americans Hate the Media and How it Matters (Princeton University Press, 2012).

This week (baring a major disruption from Hurricane Isaac), Republicans will hold their presidential nominating convention in Tampa Bay, FL. It will be followed one week later by the Democratic nominating convention in Denver, CO. Historically, conventions have produced “bumps” in the trial heat polls. On his blog, Tom Holbrook of the University of Wisconsin, Milwaukee, produces a chart showing the bounces generated by major party conventions since 1964.

Some of these are quite large, including a 14.1 percentage point bounce for Nixon in 1968, a 13.6 point bounce for Clinton in 1992, and a 12.9 point bounce for Goldwater in 1964. Holbrook uses a regression model to predict that this year Romney will receive a 3.6 point bounce and Obama a 1.1 point bounce.

What are the important considerations that we should keep in mind when thinking about the importance (or lack thereof) of the conventions for the presidential race? 

In general, political scientists find that the biggest effect of major campaign events is to activate partisan predispositions. The campaign as a whole has this result, but the effect may be even more dominant for conventions and debates that require one to self-select to be a viewer. This activation effect is driven by that fact that those most interested in politics are also the most likely to have strong existing views (see here and here). A very large portion of those who tune in are political junkies/activists who made up their minds long ago or those who are at least partisan enough that watching these events reminds them what they like about their party and dislike about the other one.

Of course, this does not preclude large post-convention bounces. But the biggest bounces tend occur when a candidate has a divisive nomination fight or for some other reason has failed to previously consolidate his own coalition behind him. This partisan reinforcement (or activation) can produce a surge in the polls without converting many swing voters. Voters who are truly up for grabs have the least interest in, and knowledge about, politics. They are simply unlikely to pay attention to conventions.

This was all true when the parties were less polarized and there were far fewer media choices. Yet it is possible that polarization and the proliferation of media options have made conventions even more primarily about rallying the base. Ben Lauderdale pointed out on twitter yesterday that convention bumps appear to be getting smaller over time. This makes sense if the bumps seen in past decades largely resulted from the partisan activation of candidates’ own coalitions. In today’s polarized environment, almost everyone exposed to convention messages comes in with their partisanship already activated.

How will people watch the conventions? To get a sense, I assembled available data from the Pew Research Center’s Project for Excellence in Journalism, as compiled in their annual “State of the Media” reports. All figures below are from that source.

Total Households (in millions) Watching Conventions on Cable or Network TV

Source: “State of the Media, 2009”

The first thing to mention is that it is hard to predict how many total people will tune in to the conventions this year. While the networks have shrunk the amount of prime-time coverage over the years, many people still do watch the network coverage and it is relatively easy to switch over to complete coverage on cable. The total number of households watching the conventions actually shows little clear trend over time. There was a decline after 1992, but that was reversed with a big increase in viewership four years ago. It will be interesting to see if we return to the post-1992 norm or see viewership more like 2008.

Percentage of Households with TVs Turned to the 3 Network Evening New Programs Based on Nielson Data

Source: “State of the Media, 2012”

The three major broadcast networks audience has been steadily shrinking for decades. The most popular source for television news in the U.S. remains local news broadcasts, but these contain very little national political information. For this, you need to use network or cable news or some other source. On a normal evening, even the diminished network news audience is much larger than the cable prime time audience, but that is not necessarily the case for special events like party conventions, where people are more likely to seek out cable news channels.

Political Convention Viewership in 2008

Source: “State of the Media, 2009”

In 2008, the number of people watching the conventions on cable was not that much less than the number watching on the networks. This was especially true for the Republican Convention, likely driven Republicans’ affection for Fox News and distrust of the networks. People were just about equally likely to watch the Republican Convention on cable as on a network.

We can see a bit more about what this means by looking at the breakdown across cable channels. The Project for Excellence in Journalism doesn’t appear to have this data available for 2008, but I could find it for the 2004 conventions.

Democratic Convention Viewership on Cable News (in thousands)

Source: “State of the Media, 2005”

Republican Convention Viewership on Cable News (thousands)

Source: “State of the Media, 2005”

While cable viewers of the Democratic Convention were roughly evenly divided between the three major cable new networks, a majority of cable viewing of the Republican Convention was done through Fox News. The biggest difference between 2004 and 2012 will likely be MSNBC being used by a higher percentage of Democratic Convention viewers and a smaller percentage of Republican Convention viewers.

Does this tell us anything more about the effects of the conventions on public opinion? To the extent that Republicans are likely to watch their party’s convention on Fox, the network’s style will likely enhance the partisan activation effect. And to the extent that Democrats watch their convention on MSNBC, it will likely have a similar effect. To reiterate, this is the main effect that conventions have always had, only now more so.

[Cross-posted at the Princeton University Press blog]