The Ratings War: Does It Matter?

Posted on September 12, 2012 by

Nothing epitomizes the horse race nature of presidential politics quite like the conventions ratings war.  For example, Matt Drudge and Rush Limbaugh were quick to dismiss the DNC’s higher tv ratings by noting that Honey Boo Boo tied Bill Clinton’s much-anticipated speech.  Democrats, on the other hand, were quick to proclaim victory and trumpet their superior performance.  But do convention television ratings matter?  What can they tell us about the looming presidential election?

Nate Silver at fivethirtyeight.com examined television ratings as an indicator of the proverbial “convention bounce,” finding a strong historical relationship between the two factors.  And while I agree with Jonathan Bernstein that patience is required in assessing these sorts of things, we might reasonably conclude that television ratings have some predictive validity in this context (though there are certainly some causality issues involved).

What I wanted to know is whether convention ratings have any predictive power with respect to election outcomes.  So I calculated the difference in DNC and RNC ratings (X-axis).  Higher values indicate the DNC drew the higher rating.  The response (Y-axis) is the percentage of the vote received by the Democrat.  Though we see a positive relationship in the chart to the right, as we would expect, the effect is small in magnitude (b=.27) and statistically insignificant (p=.597).  Thus, it seems that victory in the ratings war is not a precursor of victory in the general election.  As an important side note, I suspect this conclusion is especially true today as surveys have shown that fewer voters make electoral decisions based on party conventions compared to thirty years ago.  The second chart (to the right) presents the percentage of ANES respondents who reported making their vote choice during the the DNC or RNC.  In 2008, less than 15% of respondents made their decision during a convention while in 2000 it was fewer than 10%.  So in this context I think convention television ratings are good for one thing: bragging rights.

I also examined whether convention ratings are predictive of voter turnout.  I used the aggregate ratings for both party’s conventions (X-axis) and turnout in presidential elections (Y-axis).  We see in the chart below a reasonably strong, positive relationship.  In fact, the relationship is statistically significant with only 13 observations (p=.01).  The magnitude of this effect is such that every 10 points in combined convention ratings adds about 3% to voter turnout.  A bivariate regression indicates that summed convention ratings explain about 46% of the variation.  If you look closely, you’ll see 2012 added to the chart based on the model’s predictions.  The estimated turnout for this election is 53%.  (disclaimer: yes, this is a very crude model of turnout).  So in this context, convention ratings might give us a window into an important element of presidential elections.  Notably, turnout matters in this election because higher turnout probably favors Obama.  See Nate Silver’s excellent post on this issue.

For those wondering, the chart below presents the raw DNC and RNC ratings from 1960 to 2012 (in the conventional red and blue lines). The data are available from Nielsen here.  We know that the closing night of the DNC drew 35.7 million viewers while the RNC drew 30.3 million viewers.  Both are down from 2008.  Interestingly, we can see that fewer Americans are watching party conventions today compared to thirty years ago.  One explanation for this may be greater electoral polarization.  This dovetails nicely, I think, with the decline in the number of voters who indicate making their electoral decision during the party’s conventions (see the ANES data above).  Another probable explanation is a general decline in network newscasts and a correspondent rise in electronic media.  Finally, this trend may be due to the fact that Americans are inundated with so much political information today that there is less desire for watching party conventions as a source of political knowledge.  (note: I’m not aware of any research in this area so I’m willing to concede these are simple conjectures.  If anyone knows  of a relevant study I’m happy to reference it.)

Posted in Elections, Polarization, Political Behavior, Political Parties, Voting Behavior | Leave a comment

Google Politics: “Register to Vote”

Posted on September 7, 2012 by

This from Google Politics today on the Twitters:

Here are some additional data from Google Trends for you to chew on.  Thoughts on the relevance or irrelevance of these data are welcome.

The chart below reports the search volume for “register to vote” over the past 12 months.

And here we see the search volume for “register to vote” over the past 30 days.  There is a clear spike during the RNC.  However, that spike appears to be dwarfed by the spike during the DNC.

Finally, the chart below reports the states from which “register to vote” was most searched over the past 30 days.

Posted in Elections, The Presidency, Voting Behavior | 3 Comments

Paul Ryan “Worked With Democrats as Often as Any Republican.”

Posted on August 27, 2012 by

The effort to define Paul Ryan is in full swing.  While most political observers are well aware of Ryan’s record, prior to his VP selection fully 43% of Americans reported having never heard of him.  So how should we define Paul Ryan?

Nate Silver of the New York Times reports that Ryan is the most ideologically extreme vice presidential candidate since at least the 1900s.  Silver uses Keith Poole’s NOMINATE scores to draw this conclusion.  Ryan has a NOMINATE score of 0.805 while the average Republican in the 111th Congress has a score of 0.62 (a standard deviation is 0.16, so Ryan is over a standard deviation more conservative than his colleagues).  Eric Schickler—a political scientist at UC Berkeley—is skeptical .  He notes, rightly in my view, that the scores used by Silver lack validity over long historical periods (specifically, the scaling procedure consists of different issues in the pre-New deal, pre-Civil Rights era).  However, Schickler goes on to note that Ryan is indeed a very conservative Republican.

Michael Tanner at the National Review and CATO offered a different take, claiming that Ryan is a pragmatist rather than a true ideologue…

Ryan has been the classic “half a loaf” type of conservative. Time and again, he has shown that he is willing to compromise and take far less than he had originally sought… You won’t find Ryan on the short end of any 434-to-1 votes.

and

According to the National Journal, Ryan works with Democrats about as often as any Republican does.

Unlike Silver and Schickler, Tanner’s conclusion rests on a limited set of anecdotes rather than a systematic examination of Ryan’s legislative record.

There are a couple ways to evaluate these claims on a more scientific footing.  On the one hand, we can simply look at Paul Ryan’s party unity score.  These scores are calculated as the percentage of votes on which a member sided with his or her party where 50% of Republicans voted against 50% of Democrats (the data are available here).  Over his
career in the House, Paul Ryan voted with Republicans 94.5% of the time on contentious votes.  The median for House Republicans 94.3%.  So in this case, Ryan is right at the median (unlike his NOMINATE score).  But this deserves an obvious qualification: voting with your party on contentious votes 94.5% of the time is hardly evidence of “pragmatism.”  I suppose it is true that Ryan “works with Democrats about as often as any Republican does.”  But that’s the same as saying Ryan hardly ever works with Democrats.  More on that in a moment.

The chart below reveals Ryan’s party unity score (blue line) per-Congress over his career in the House (the red line is the median Republican).  Once again, Ryan seems to mirror the general trend for Republicans in the House.  For example, though Ryan voted with Republicans on contentious issues 97.7% of the time, the median in the 111th Congress was 96.4%.  This is higher than the typical Republican, but not remarkably so.  Thus, if Ryan is a “pragmatist,” which seems like a major stretch, this label is probably more befitting his early career.

We get a similar picture of Ryan if we look at the bills he sponsored.  (This data are only available up to the 110th Congress, which is an important caveat).  What I did was calculate the percentage of Republicans who cosponsored Ryan’s bills (the data are available on James Fowler’s website here).  I used all bills as well as all bills where Ryan was in the majority.  I restricted the sample to only bills with at least 10 cosponsors.  The red line represents Ryan’s position.  We see, again, that Ryan falls somewhere toward the middle rather than the extremes.  For all bills Ryan attracted 72% Republican cosponsors while for bills where Ryan was in the majority the percentage of Republican cosponsors drops to 69%.  Again, yes, Ryan is more like a “typical” Republican.  But when only 30% of your cosponsors are from the other party this is hardly pragmatism.

Two important caveats about the co-sponsorship data.  First, the time span ends at the close of the 110th Congress.  I suspect that much of what Ryan has done that deserves the label “ideologue” occured in the most recent Congresses.  Indeed, the time series above suggests that Ryan became more ideologically extreme in the most recent years.  For example, his budget proposal, which has no cosponsors, passed with 0 Democratic votes and 4 Republicans no votes.  Moreover, much has been made of the fact that Ryan co-sponsored a bill with Todd Akin.  Second, I tried to further restrict the sample to only bills mentioned in CQ Almanac (as a measure of “importance”).  The data are available on the Policy Agenda’s Project webpage.  The cosponsor charts above almost certainly make Ryan appear more moderate than reality since some (perhaps many) of the bills are of little importance.  The key here is that—to my surprise—Ryan sponsored no important bills over his time in Congress (using this particular measure, and only up to the 2008).  Moreover, of his sponsored bills, only 2 were enacted into law (one modified the taxes paid on hunting arrows, Ryan is an avid bowhunter, and the other renamed a post office).

So in short, Ryan is a “typical” Republican in terms of his party unity score and cosponsorship network, and admittedly the figures above paint Ryan a less extreme light than his NOMINATE score does, but this is hardly evidence that he’s a “pragmatist” who “works with Democrats.”  He still votes with his party nearly 98% of the time and his sponsored bills attracted only 30% of Democratic cosponsors (and this number is surely inflated by non-contentious bills).  So this campaign of late to make Paul Ryan seem like a moderate, bipartisan guy who works with Democrats is just wrong.  The claim that Ryan “works with Democrats about as often as any Republican” is technically true, but very misleading.  But the important point is that the problem isn’t Paul Ryan, the problem is that the two parties (especially Republicans in the House) are so highly polarized that the average Democrat and Republican hardly ever works with his or her rivals.  Being “typical” in a polarized Congress does not make one a moderate.

Posted in Elections, Legislative Politics, Polarization, Political Parties, The Presidency | 1 Comment

The Veepstakes: What’s the Electoral Value of a Vice President?

Posted on August 11, 2012 by

A few hours ago, Mitt Romney announced Wisconsin representative Paul Ryan as his running mate.  It seems obvious that in picking Ryan, Romney is trying to solidify his position with fiscal conservatives and doubling down on the economy.  Personally, I think it’s a smart choice (see Bernstein for his thoughts).  But the important question is: Do vice presidential candidates have significant effects on presidential election outcomes?  This question is one that those in the media will presume to know, fueling endless speculation about the purported “value” of Romney’s decision.  Drink every time you hear phrases like “game changer,” “transformative,” “shakeup,” or “high-reward.”  We’ll see you back here next week…

But this issue is one political scientists have studied, and most works uncovers only minimal effects on a president’s vote-share.  Typically, researchers in this field have looked for evidence of a VP’s “home state advantage.”  Though this is just one way to examine the topic of a vice president’s electoral value, it seems that if the home state effect is minimal, there is little reason to believe a vice presidential candidate has significant electoral value nationwide.  The conventional wisdom—see here for a paper by Dudley and Rapoport— is that vice presidential candidates do  increase a presidential candidates vote share, but only in less populous states like Wyoming (Dick Cheney) and Delaware (Joe Biden) where the likelihood of swaying the Electoral College is minimal.  A more recent study by Devine and Kopko adds the qualification that this effect only exists when the state is less populous and the vice president has significant political experience in his or her state (which I’ll concede Paul Ryan has).  Still, further analysis has shown that the magnitude of this effect is quite small.  Research by Garand (1988) and Rosenstone (1983) shows that a president’s vote-share increased by between 2.5% and 3% in his VP’s home state (cited in Devine and Kopko).  In case you’re wondering, Obama won Wisconsin 56% to McCain’s 42%.

Thus, while the strategic selection of a vice president does seem tied to the electoral value of a candidate in the Electoral College (see a paper by Sigelman and Wahlbeck), political scientists have not found systematic evidence that a VP has significant effects on election outcomes.  Now admittedly there are other, perhaps more subtle but nonetheless important, ways a vice president can impact a presidential race.  Ryan may help Romney frame the political right’s economic message and keep the economy forefront in the minds of voters, for example.  But the purported value of Romney’s choice should be nuanced and tempered; this is not what we’re going to hear in the media over the next week or so.

Posted in Elections, Electoral Institutions, Political Behavior, The Presidency | Leave a comment

Some Thoughts on Americans Elect, Polarization and Gridlock

Posted on April 11, 2012 by

In January I was invited to speak at a roundtable hosted by Americans Elect–a nonpartisan presidential nominating organization.  As you may know, Americans Elect (AE) has garnered quite a bit of attention this election cycle from academics and pundits alike (see here for a Colbert Report bit and here for a forum at Vanderbilt).  Joining me on the panel was Americans Elect CEO Elliot Ackerman, veteran South Carolina political reporter Lee Bandy,  and my colleague at the College of Charleston David Gillespie.  I wanted to direct readers to David’s excellent article at the History News Network on Americans Elect and their prospects for changing American politics and in particular our party system.  I also wanted to put in a plug David’s most recent book–Challengers to Duopoly–which examines the important role played by third parties.

David’s comments on Americans Elect have been fairly optimistic in nature.  This optimism is revealed by the title of his article at HNN: Americans Elect Could Become a Viable Political Party.  My views on AE are more reserved and perhaps a tad skeptical (though I’m a supporter of their overall mission).  The difference in our views is attributable, I suspect, to the various effects Americans Elect might have and which of these possible effects the two of us have focused on.  Where David has discussed AE as a participant in the “market place of ideas” (where third parties, like the progressives and populists, have been rather successful), I’m more interested in the effects of AE on political polarization and legislative gridlock.  My interest in these factors stems from personal academic interest but also because AE is motivated, according to their own accounts, by these two factors (see their about section).  My overall message is relatively simple:  it’s not clear to me how AE’s non-partisan presidential nominating convention will alleviate these two problems.  Here are a few of my thoughts:

First, if AE delegates select a Democrat or Republican for their presidential ticket (which is very likely), that individual is required by AE bylaws to choose a vice-president from the other party.  The logic of this bylaw seems to be that a Republican president and a Democratic vice-president (or vice versa) somehow equals a moderate administration.  This, of course, ignores both the role of vice-presidents in our constitutional framework and the history of bipartisan presidential administrations.  Though the vice-president is first in line for presidential succession, he or she has few formal powers other than breaking tying votes in the Senate (a rare occurrence).  History is replete with VPs who lamented their bland existence (Roosevelt comes to mind).  In addition, there have been two examples of vice-presidents belonging to parties opposed to that of the president (1796, Adams and Jefferson; 1864, Lincoln and Johnson).  Neither worked very effectively toward “common goals.”  Moreover, party unity does not make a unified or “collaborative” administration.  Though they shared a party label, John C. Calhoun famously resigned as vice-president over Jackson’s support of federal tariffs.  In sum, I just don’t buy the logic that a president and vice-president of rival parties somehow equals a moderate or bipartisan White House.

Second, Americans Elect clearly states that part of the aim is to break gridlock, telling their members “you have the power to help break gridlock and change politics as usual.”  Now it is true that polarization is a cause of greater legislative gridlock (see Binder 2003 or McCarty, Poole and Rosenthal 2006).  But I don’t see how a non-partisan president can break gridlock when confronted with a partisan House and/or Senate.  Gridlock is hard enough to break even when the same party controls all three institutions.  If anything, a non-partisan president may have the exact opposite effect (exacerbating gridlock).  Single-party control is a solution to the constitutional barriers Madison and his colleagues erected, in other words.

Third, the rather simple fact is that polarization is the norm in American politics–not the exception.  Brendan Nyhan has made this point a few times on his blog (see the figure to the left from Poole’s VoteView).  As the norm, polarization is something very difficult to break or reverse.  So the question we should ask is: Why polarization?  (note: I highly recommend Sean Theriault’s excellent book “Party Polarization in Congress” on this topic).   Political scientists have some good answers to this question.  One is a process variously called geographic or partisan “sorting.”  Though there is some disagreement about whether redistricting has caused greater polarization, there is at least some consensus that at the state- and district-level, voters with similar demographic and socio-economic characteristics are grouping together resulting in more politically homogeneous locales (though see this post at TMC for a dissenting view).  Reforms in Congress are yet another source of greater polarization.  Simply put, the formal power of party leaders has expanded in the House and Senate since the 1970s while the power of committee chairmen–as a rival source of legislative power–has declined over the corresponding period.  Party leaders are able “stack the deck” in their favor leading to roll-call votes that more cleanly pit Democrats against Republicans.  Finally, it’s also generally known that the parties’ platforms have undergone significant change or “ideological realignment” over the past four decades.  The South has become homogeneously conservative and Republican dominated while the northeast has become homogeneously liberal and Democratically dominated (see here for an excellent video on this).  In sum, I’m just not sure how Americans Elect’s non-partisan presidential nominating process alleviates these things which political scientists know (or very strongly suspect) have caused the growth in polarization over the past thirty or so years.

Let me finish by saying there are some unclear strategic voting implications of Americans Elect’s success.  I suspect that–if anything–AE has a pro-incumbent bias.  I’m assuming that individuals who voted for the incumbent president–in this cycle, Obama, but I think effect flips from incumbent to incumbent–are probably less enthusiastic about AE’s nominating process than individuals who are seeking an alternative candidate.  This assumption is supported by AE’s own data, as all three of their “most supported candidates”–Buddy Romer, Ron Paul and Jon Huntsman–are on the political right.  Thus, if AE selects a member from the non-incumbent party, this will probably have the effect of disrupting a relatively simple two-candidate election and create incentives for strategic voting .  Overall, I would say this complicates our presidential elections (not that complexity is inherently bad, but I just don’t see what this improves about polarization and gridlock).

Posted in American Political Development, Elections, Electoral Institutions, Polarization | Tagged , , | 2 Comments

Reforming Polarization and Gridlock: Series on Congressional Reform

Posted on March 14, 2012 by

For someone who studies congressional development, the past couple years have been frustrating. Many people with noble intentions proposed reforms to remedy our dysfunctional Congress. However, these discussions have almost universally missed the causes of gridlock and polarization. They offer remedies rather than cures. So, in this series – that will continue until I run out of ideas – I’m going to offer reforms that, in my opinion, are more crucial and fundamental to congressional operation, polarization, and the like. But before I do that, I need to debunk the worst proposal of them all…

The filibuster. If I had a dime for every time somebody proposed filibuster reform… Now, before I lament why this reform is misguided, let me be clear: the filibuster needs reform. Its practice grossly misrepresents its intent (but not the Framers’ intent) and is generally a giant obstacle to majority government. But, that being said, it is not even close to the worst procedural device within the existing legislative process. Sure, it’s an easy target. I too often hear among colleagues, reporters, and friends alike, “if we could only reform the filibuster…” But more often than it is assumed filibuster reform is the silver-bullet for gridlock. We fix that, we fix gridlock and the problems it creates (i.e. the debt ceiling debate).

It’s not and here’s why: when polarization is the main problem, reforming Congress to make it more majoritarian will only exacerbate the existing problem. It will push the parties further apart, and effectively make the problem we have now much much worse over time. The already bad gap between the parties would become a chasm. Think about looking across the Grand Canyon. Now compare that to draining the Pacific Ocean to try and see Japan. That would be the size of the problem we would face and frankly, we can’t afford to make the already bad gap worse.

Here’s an interesting hypothetical. Let’s assume for a second that the filibuster, cloture, and the 60-vote Senate does not exist. Let’s assume, as Matt Glassman described so well, that we effectively have two Houses of Representatives. What happens in a situation like today, where the chambers have split control between the parties? Nothing. Absolutely nothing. But it would be worse than that: negative-nothing (is this a thing?). With the current ideological division between the parties and two majoritarian chambers, neither party has an incentive to compromise with the other because 1) party leaders control the legislative process and have even less incentive to compromise with members of the opposite party within their chambers 2) the two parties don’t agree on much to begin with, and 3) each chamber holds a veto over the other. Legislation created in each chamber would be more ideologically extreme (because each chamber is effectively controlled by each party’s leader), and there would be even less overlap on bills passed in each chamber. I’d be amazed if that hypothetical Congress is more efficient than the current one.

Adding to the difficulty is divided government. I’m not going to fully delve into divided government (president and Congress split control), but this much should be said. Pending one party had control of both chambers, presidents’ legislative power would diminish when competing against a truly majoritarian Congress (not necessarily a bad thing), and he/she would be forced to resort to more frequent vetoes (because legislation is passed more rapidly and because it is more ideologically extreme) to demand compromise… which probably wouldn’t happen to any significant extent. If the U.S. didn’t have an independent executive, this wouldn’t be as big a problem (like many parliamentary systems). But in the American system, it has just as much potential to deadlock the system as split chamber control.

Nothing about these scenarios suggest more “effective” government save one specific circumstance: unified government. Which, it turns out, resembles majority totalitarianism more than American democracy, something the Founders specifically sought to avoid.

So it’s not that filibuster reform is itself bad. It just leads Congress down a dangerous path: more extreme polarization in a system of shared power and checks and balances. That is, creating a majoritarian Senate is not inherently good when the results exacerbate an already bad problem, threaten just as much if not more inefficiency, and introduces the potential for tyranny of the majority (I mean this in a non-dramatic sense. I’m not suggesting 1984. But I am suggesting radical changes in policy from Congress to Congress). The Constitution is not a majoritarian document. In fact, it is decidedly non-majoritarian. It’s designed to prevent action/legislation, not spur it along. In a very real way removing the filibuster would strain the U.S.’s non-majoritarian Constitution… sort of like putting a jet engine in a pinto: lots of horse-power, not much control.

There are better reforms that could temper polarization and promote compromise, but the filibuster is not the first or even most effective change to acheive these ends. So, my main objective is to identify the institutional sources of polarization. My second, and much less definitive, objective is to offer some solutions that could help.

Next up: The Rules Committee.

Posted in Filibuster, Legislative Politics, Legislative Procedure, Polarization, Political Institutions, Senate, Separation of Powers | 1 Comment

Institutional Advancement: Do Dan Boren’s Claims Add Up?

Posted on March 8, 2012 by

Politico published an insightful article yesterday provocatively titled “Members of Congress:  This Job Sucks.”  Much of the article centers on and interview with Oklahoma Representative Dan Boren (D).  Boren is retiring at the end of this session, so he’s using his unique position to speak candidly about what he sees as the the sad state of the House.  Most of his comments revolve around gridlock and the rank-and-file’s inability to advance meaningful legislation.  On this matter, I have no qualms with Boren’s commentary.  One claim, however, does’t quite add up in my view…

In the old days, [Boren] would have been a sure bet to bide his time in Congress, win reelection by serving up earmarks to his constituents and, after a couple of decades, grab the prized gavel of the Armed Services Committee….”You want to get things done for your constituents. If you can’t ever become speaker or a committee chairman, why are you doing it?”

The implication seems to be that career advancement in the House–in particular advancing from lowly rank-and-file to committee chairman–is particularly arduous in the modern Congress.  Historically speaking, this claim doesn’t work very well.  Specifically, the aptly named seniority system–which served as the overarching rationale for promoting committee chairs from 1911 until the early 1970s–severely limited members’ advancement opportunities.  Under this rather simple system, the committee member of the majority party with the longest service in committee became chairman.  Thus, members with long careers were able to wield significant influence over policy outcomes, often at the expense of their party’s policy goals (particularly southern Democrats, who were over-represented in terms of occupying committee chairs).  Reforms in the 1970s, however, altered the workings of the seniority system, allowing both parties to deviate from the seniority norm (making it easier for other members to advance).  Moreover, in the 104th Congress, House and Senate Republicans implemented a rule limiting committee chairmen to six-year terms.  So if anything, there is greater access to top committee positions in the modern Congress compared to the 1960s.*

But what about more recent trends?  Has institutional advancement–in the form of assuming a committee chair–declined in recent years?  I feel a figure coming on…

What we see in the first figure is the median (blue) and mean (red) two-year terms served by committee chairmen and ranking members per Congress (from 1979 to today).  (note: the data are for standing committees only).  For both series there is a modest, positive slope.  Over the entire series the amount of change is about .3 terms (or just over half a year). In the past two decades, however, there has been a slight decrease in the terms served by chairmen and ranking members.  The second chart tells a similar story.  Here we are looking at the median (green) and mean (purple) number of terms served in committee by chairman and ranking members.  If anything, the average chairman or ranking member is serving fewer terms in committee before advancing to the top spot.  So in short, we just don’t see consequential changes in committee advancement of the sort that Boren laments, at least not in the past 20 or so years.  If anything it has become easier to assume a committee chair or become ranking member (though the trend is probably negligible).  Granted, there are probably more sophisticated ways of examining this issue (for example, to what extent party loyalty plays a role in advancement and examining at what point in one’s congressional career is a member most likely to chair a standing committee).  Nonetheless these simple statistics reveal some important trends.

As an aside, Boren is in his 4th term in the House.  In the 112th Congress, the data reveal that the median number of terms served by chairmen and ranking members is 10.

* One important counterpoint is that the rule limiting committee chairmen to six-year terms was undone by Democrats at the start of the 111th Congress.  Republicans, if I recall, have their own rule limiting members to six-years, though exceptions have been made in the past.

note: the data on committees and their members is available on Charles Stewart’s Congressional Data Page.  Many thanks to him for making this useful data publicly available.

Posted in American Political Development, Legislative Politics, Legislative Procedure, Political Parties | 3 Comments