Category Archives: Elections

What Does Tuesday’s Special Florida Election Tell Us About the Upcoming Midterms?

Many analysts viewed last Tuesday’s special House election in Florida’s 13th district to replace longtime Republican Congressman Bill Young as a bellwether for the upcoming 2014 midterm congressional elections, particularly because they expected Obamacare to be the central issue in the Florida campaign and in November’s elections.  Thus, the author of the New York Times’ election preview, which was headlined: “Florida Race for House Sets Stage for 2014”, described it as “a contest in the first race of the 2014 battle for control of Congress, with both parties hoping for a victory and watching carefully how President Obama’s health care law may affect the outcome.” The prevailing media perspective did not change in the aftermath of Republican David Jolly’s narrow 48.5% to 46.6% victory over his Democratic opponent Alex Sink.  In a not uncharacteristic post-election interpretation, the Economist wrote: “On March 11th David Jolly, a Florida Republican, won a special election to the House of Representatives by relentlessly bashing Obamacare. His party hopes to use the same tactic to hold the House and capture the Senate in November. ”  Similarly, a Washington Post article opined that “Jolly’s win in a Gulf Coast district just west of Tampa illustrated the political toxicity of the law known as Obamacare. Jolly favored repealing and replacing the law, which was a central focus of the campaign, while his Democratic opponent did not.”

But did Jolly really win the special election by “relentlessly bashing” Obamacare – and will that tactic make a difference in who controls Congress come November?   The answers are not as clear cut as many analysts would have us believe.  It is true that Jolly ran on a platform that included repealing Obamacare.  But Sink’s stated position, which was to fix Obamacare, rather than repeal it, actually had strong support among likely voters, at least in the limited polling data* I was able to review.  For example, in a poll of likely voters conducted by St. Leo University, 40% of respondents wanted to keep Obamacare but make changes to fix some of the problems, compared to 20% who wanted to repeal it and replace it with a different health care plan, and only 20% who wanted repeal and a return to the pre-Obamacare status quo.

Of course, one needs to be cautious in using polling data to infer voters’ motivations. One difficulty is that those responding to surveys can only express an opinion on issues about which they are polled.  If voters are only asked about Obamacare, it will seem to loom large in their decision calculus. Moreover, responses differ according to what options are available. For example, this St. Petersburg poll of likely voters only asked respondents about their views on two issues: Obamacare and illegal immigration.  And on Obamacare, 35% of those surveyed said they supported repeal, 34% leaving it alone, and 27% advocated repeal once a “better plan” is in place.  Without specifying the details of that “better plan”, however, it is hard to interpret the results.  Does that mean a plan that is Obamacare without the blemishes, which was Sink’s stated position?  Similarly, in the St. Leo poll, it is hard to know what the 20% who wanted to replace Obamacare “with a different plan” had in mind, and how closely that option aligned with Sinks’.  (I did not see any crosstabs in either poll by which to judge the candidates’ respective supporters’ views on Obamacare.) Note that many of the ads attacking Sink for her support of Obamacare, such as this one by the Chamber of Commerce, did not necessarily advocate for its repeal.  Instead, they targeted its impact on Medicare payments.

So it’s not clear that Jolly’s message to repeal Obamacare carried the day.

In taking stock of Jolly’s victory, it is also helpful to remember that the smaller voter turnout characteristic of a special election is not always a reliable indicator of who will vote during a normal midterm election cycle. Note that while Sink appeared to have a slight lead based on early voting heading into Tuesday’s election, Jolly evidently benefited from a better Election Day turnout.  However, the election drew only about 184,045 voters, and Jolly won victory over Sink by less than 4,000 votes. Pre-election polls indicated that Jolly’s support was stronger among the older white voting population that is more likely to turn out in a special election. By comparison, in 2012 the long-time incumbent Young won this district by a much larger margin, 16% over his Democratic opponent, than did Jolly on Tuesday even though President Obama narrowly carried the district by about 50% to 49% in the presidential race. Of course turnout in that presidential year was almost 330,000 voters – much higher than it is likely to be in the coming mid-term election. However, in the 2010 midterm, turnout in Florida’s 13th district, (which encompassed slightly different territory) was almost 267,000 – much higher than what we saw in last Tuesday’s special election. (According the Florida Secretary of State’s office, Republican voters constituted about 38% of voters in the district, Democrats about 35%, and Independents about 23%, but I don’t know the partisan breakdown of Tuesday’s voting pool.)  So Jolly may have benefited by the smaller size and composition of Tuesday’s electorate.

Nor was this election only about Obamacare – voters were inundated with ads targeting Sink’s handling of a pension fund, Jolly’s views on entitlement reform including Social Security, and climate-change related issues including flood control, in addition to immigration. All of this makes me much more skeptical than are media pundits regarding what the results of Tuesday’s election tell us about the likely impact of Obamacare come November. Keep in mind that Republican control of the House is probably not at stake in 2012, regardless of how large Obamacare looms in voters’ decision calculus. And in the Senate races in which Democrats are vulnerable, Obamacare is only one of a myriad of issues, including jobs and the state of the economy, that are likely to influence outcomes. Finally, I expect most Democratic candidates to run on a platform of “mending” Obamacare, which may prove more appealing to many moderate voters than calls to fully repeal it.

So, does Tuesday’s special election result in Florida’s 13th congressional district provide an accurate preview of election events to come? Probably not nearly as much as some media pundits will have you believe.

*Thanks to Kate Hamilton for research on the polling data I referenced here.

The Media and the 2012 Election: Taking the “Fun” Out of the Fundamentals?

Although the next presidential election is more than two years away, pundits are already trying to parse the meaning of polling data even though trial heats polls this far in advance of the election are practically worthless. The same goes for the recently concluded  Conservative Political Action Conference straw poll won by Rand Paul for the second year in a row – it likely tells us little about who the Republican nominee will be. No matter – the fact that these polls don’t tell us much won’t stop pundits from endlessly hyping them – after all, they have to write something! Before we wander too far down the garden path of media misinformation, however, it might do well to revisit some of the more glaring media misconceptions regarding the 2012 presidential election.  In a piece I have coming out in the American Political Science Association journal Perspectives on Politics I take aim at what I see as five enduring myths promulgated by many (but not all!) journalists in the aftermath of that contest.

1.  Obama won despite a poor economy that normally would have doomed the incumbent president.  In May, 2012, for example, the New York Times’ columnist David Brooks, echoing sentiment expressed by many pundits, wondered, “Why is Obama even close? If you look at the fundamentals, the president should be getting crushed right now.”  In fact, however, most of the political science forecast models whose projections are based in part on the economic fundamentals had Obama a slight favorite to win the popular vote. In the aggregate, those models suggested Obama would garner a shade more than 50% of the two-party vote – he actually won bit more than 51%.

2. Obama’s victory owed much to the fact that voters found him more likable than Romney. A common conceit among journalists is that when it comes to attracting votes, candidate likability matters.  In a characteristic media assessment, the Washington Post’s Chris Cillizza and Aaron Blake argued, “Presidential elections are rarely won and lost on policy. Voters instead tend to choose the person they most want to be president based on who they like… Call it the empathy factor. And it matters. A lot.”  Actually, it is not clear likability matters much at all. Research by Morris Fiorina finds that in the 13 presidential elections from 1952 to 2000, the candidate with the net advantage in personal ratings won only seven times.

3. Obama’s early advertising blitz in key swing states effectively defined Romney as a heartless capitalist out of touch with the concerns of the middle class. One oft-cited example was the “brutally effective” series of ads run by the Priorities USA Super Pac in Ohio that depicted Romney “as a cold-hearted plutocrat, ruthless in his pursuit of profits and unaffected by the human toll of Bain’s brand of buying and selling companies”.  The best evidence, however, suggests these ads had limited staying power and likely didn’t change many votes in the long run.

4. Romney’s numerous campaign gaffes doomed his election chances. The most conspicuous, perhaps, was the infamous secret recording showing Romney condemning the 47% of voters who are “dependent on government,” but critics also cite his remarks favoring self-deportation by illegal immigrants, his failure to release individual tax returns, and his ill-timed criticism of Obama as sympathetic to anti-American interests in the Middle East in the aftermath of the attack on the Benghazi diplomatic compound, to name the most prominent. As one postmortem summarized, “Obama, weighed down by a poor economy, also needed help—and he often got it from Romney”. National and state level polling, however show very little variation in Obama’s lead over Romney throughout the campaign, with the exception of a short-lived post-convention bounce in Obama’s favor. In this vein, the “devastating” secret “47%” video recording appears to have had minimal effect on Romney’s polling support.

5. Obama’s vaunted ground game easily outclassed Romney’s poorly run get-out-the-vote efforts. This proved to be a staple of the numerous media post-mortems. However, it is not clear that the marginal effect of Obama’s allegedly superior ground game was big enough to change the outcome of the race. Indeed, overall turnout in 2012 was down from 2008, by 3.4%, as was Obama’s share of the vote, which dropped in that same period by 1.9%. Moreover, in five of the eight key battleground states where his get-out-the-vote efforts was concentrated, the drop off in Obama’s vote was greater than the decline in his overall national vote. In short, while it is possible that Obama’s vote share would have declined even more without his advanced voter outreach effort, it is hard to prove that he won because his ground game outperformed Romney’s.

So why did Obama win reelection in 2012, if not because of the factors frequently cited by many journalists and pundits?  For the same fundamental reasons that have largely dictated the outcomes of most presidential elections in the modern era.  Simply put, it is very difficult to defeat a sitting president in a time of economic growth – even tepid economic growth – particularly when that president’s party has been in control of the Oval Office for only one term prior to the election.  With the exception perhaps of media pundits, who depend on these campaign myths to attract and hold viewers, most of us should find this debunking reassuring. Although the horse-race perspective adopted by most journalists that focuses on candidate personality, campaign strategy and gaffes fits well with their need to attract and maintain a daily audience, it is not a narrative that is particularly flattering in its portrayal of the electorate. Fortunately, while entertaining, this media perspective is not very accurate. The overwhelming evidence is that voters, while perhaps not deeply informed regarding candidates and issues, do cast their vote based on their understanding of fundamentals, such as the state of the economy, as viewed through their own partisan predispositions. Viewed collectively, then, voters are quite rational. That is good news for the future of political science forecast models—and, more importantly, for the future of the US political system, and it something worth remembering as the media gears up for what promises to be another round of entertaining, if misleading, presidential campaign coverage.

Why Joe and Jane Sixpack – and James Madison – Are Likely Pleased With Tuesday’s Results

Bob Johnson, as is his wont, chastises me once again for implicitly suggesting in my previous post that Americans last Tuesday again voted for divided government.  Bob writes, “In fact I suspect that we could and in fact may be getting divided government despite the wish of every voter that the national government be unified — under their party’s leadership, of course.”  In one respect Bob is right, of course; as I should have said more directly in my post, most Americans are not voting for divided government.   However, that is not the same as saying most Americans who voted on Tuesday preferred to have unified government under Democratic control!  (Or Republican control, for that matter.)

It is true that, in the aggregate, Democrat candidates likely won slightly more popular votes in House races than did Republicans. House votes are still being counted, and we need to be careful about counting votes in districts where incumbents from the same party were pitted against each other.  But at this point preliminary numbers suggest that, in the aggregate, Democratic House candidates tallied about 50.3% of the vote to about 49.7% for Republicans, for a margin of about .6% in Democrats’ favor. And, of course, Obama won the popular vote; although the final numbers aren’t in, he’ll likely get close to 51% of the vote.  Not surprisingly, Obama supporters cite these numbers to argue that most Americans voted for unified government under Democratic control.  But this is probably not the case. Consider the National Election Study data regarding split ticket voting in the nine most recent elections to Congress and the Presidency, as summarized in this table:

As you can see, going back to 1976, not once did a majority of voters, at least based on the NES data, vote for unified government under Democratic or Republican control.  Bob is correct, of course, that a strong majority of voters would prefer unified government – if their preferred party was in control. However, there is always a small plurality of voters who, for whatever reason, split their ticket.   And that means a majority of voters typically oppose unified control under the opposition party.

Put another way, the 50.4% of voters who voted for Obama last Tuesday are almost surely not the same 50.3% of voters who voted Democratic in the Congressional race.  Indeed, on average across the last 9 national elections, about 13% of voters have supported the Democratic presidential candidate while voting Republican at the House level.   That percentage has dropped in recent years, as has split-ticket voting more generally, but even if we restrict out analysis to the last four elections, it is still about 8% of voters who split their ticket in this fashion.   Another 10% on average across the last four elections have opted for the Republican presidential candidate but supported a Democrat in the House race.   So, we see a bit more than 17% of voters splitting their tickets in the last four elections.  This is in part a testament, no doubt, to the power of incumbency in House races.

But it is also reminder, apropos Rob Mellen’s comment, that we do not have a parliamentary system, in which our president is selected based on the popular vote for the legislative branch. Instead, ours is a system of separated institutions, each with its own electoral base, sharing powers.  As Bob notes, it is that combination of staggered elections and separated  electoral constituencies that makes it easier for elections to produce divided government.

But is divided government really all that bad?  Bob will undoubtedly be happy to learn that David Mayhew has come out with still another book, Partisan Balance, extolling the virtues of our system of shared powers, despite – because of? – its propensity to return divided government. Mayhew’s essential point is that despite the intense partisan polarization that characterizes government at the national level, the system of staggered elections and different constituencies means the policy process never systematically tilts too much in favor of one party at the expense of the other.   This, of course, drives party purists at both ends of the ideological spectrum nuts – far better, they argue, that one party be allowed to control the government, pass their agenda, and be held accountable for the results, than to have to endure a policy process characterized by partisan bickering, fiscal cliffs and incremental change at best.

That might be true.  But polls indicate that, although there is variation across time, typically as many Americans support divided government as do prefer unified government, although opinion varies  by whether their preferred party controls the presidency or not.  Moreover, consistent with the NES data, there’s never been a majority of Americans surveyed who express a preference for unified government; most are either opposed or indifferent.

Yes, we are facing two more years of divided government.  That’s nothing new.   And it may not be a problem, at least from the perspective of Joe and Jane Sixpack.  And, for that reason, I’m guessing Tuesday’s results would please “Little Jemmy” Madison as well, even if modern-day party purists are frustrated once more.

Once More, With Feeling! Divided We Stand

As we begin assessing the results of the 2012 elections, perhaps the most important takeaway is that, once again, the country has voted for divided government.  When this newly-elected Congress finishes its term in January, 2015, we will have had divided government during 42 of the last 68 years, dating back to the first-post World War II Congress in 1947.  Much of the media focus, understandably, has been on Obama’s narrow victory, but we shouldn’t lose sight of the fact that, as of this writing, it appears the Republicans lost a net of about 6 House seats, and therefore they will retain their House majority with something close to 236 of the 435 House seats.  The Democrats, meanwhile, defied expectations (at least mine!) and actually gained two seats in the Senate, upping their advantage there (assuming the two independents Angus King and Bernie Sanders vote with them) to 55-45.

What explains Americans love affair with divided government?  In part, our explanations may vary depending on what type of divided government we see – right now the Congress is divided.  That is the more rare form of divided government, occurring in just 12 years since 1947. In most other years we see a unified Congress facing a president from the other party.  It is tempting to think, but harder to prove, that voters are engaged – consciously, or subconsciously – in some type of partisan balancing act.   If they are consciously dividing control, the question is why?  One explanation that was particularly popular when split control meant choosing a Republican president and Democratically-controlled Congress is that voters saw the parties having different strengths.  Republicans were stronger on defense, and hence better suited for the Presidency, while Democrats were more protective of the social safety net and therefore were given control in Congress.  However, the partisan roles have been more often reversed in divided government since 1994, with Republicans controlling at least the House and Democrats sitting in the Oval Office.  Note that, as Mo Fiorina reminds us, not all voters need to be engaged in this type of reasoning for split government to occur – divided government only requires a minority of voters to split their ticket.  Despite this, critics of the balancing argument suggest this still imposes a relatively high threshold of purposive voting on the part of voters.

A second argument explaining divided government focuses not on ideological or partisan balancing, but instead on structural factors.  Thus, one explanation for why Democrats retained control of the House for so long, even as the nation elected Republican presidents, is that Democrats, by virtue of more active political participation in state government, simply fielded a more experienced and hence stronger set of candidates at the national level.  That is, they benefited by, in effect, having a stronger minor league program. It wasn’t until 1994, after Newt Gingrich had been actively recruiting Republicans to run for office at the local level for a number of years that this structural imbalance was finally overcome.  In recent years scholars have cited a second structural factor – one that advantages Republicans.  They note that Republican voters are more efficiently distributed across the country, while Democrats are bunched up in large urban areas along the coasts.  This more efficient distribution – one accentuated by Republican-controlled gerrymandering in many states after the 2010 census – means Republicans “waste” fewer votes in congressional elections.

Does it matter that government is divided?  In particular, does it lead to legislative gridlock? On this question, scholars are also divided.  David Mayhew has shown that major legislation generally gets churned out whether government is unified or not.  Others argue, however, that under divided government it is more likely that the national government will fail to address pressing problems.   That is, our answer may depend on whether we are looking at pure policy productivity or at the saliency of the legislative Congress passes in relation to societal concerns more generally.

The answer may also depend on more than whether government is divided or not.   Larry Dodd and Scott Schraufnagel suggest that Congress’ legislative productivity depends also on the level of partisan polarization dividing the two parties.  When the two congressional parties are internally very unified, but the ideological distance between them is large, legislative productivity can be hampered even during periods of unified government.  This is because our bicameral legislative process, with its supermajoritarian hurdles, makes it easier for a unified opposition party to stymie action.  Under the current conditions of quasi-divided (split Congress) government and high polarization, the tendency toward gridlock is even greater.

The bottom line is that, for whatever reason, the voters last Tuesday essentially opted for the status quo.  Despite protestations to the contrary, Obama did not receive a mandate – if anything, his position is weaker now than it has ever been.  That means from his perspective,  the legislative window of opportunity starts small, and will likely close quickly.  To be sure,  I don’t expect that both parties, in their lameduck version,  will be willing to drive the country over the fiscal cliff come the start of the New Year. But the long-term outlook for inter-party cooperation on legislation in the incoming Congress is not promising.  Republicans, responding to their own constituent pressures, are likely to be as unified in the new Congress as they were in the last.  This does not mean major legislation won’t be passed.  Mayhew shows that it can still happen – but only when it addresses the political interests of both parties.  Those cases are rare indeed.

The Big Winner Last Night? Political Science!

For a political scientist, last night’s outcomes were very, very satisfying.  To begin, viewed in the aggregate, the structural-based forecast models issued by last September hit the two-party popular vote share almost exactly on the head, as of this moment.  (As long time readers know, because there are so many different structural models, I take their average and median forecasts as my best estimate of what is going to happen.)  To refresh your memory, those models indicated that Obama would win 50.3% of the two-party vote, on average, with a median forecast of 50.6%.  Right now, Obama’s share of the two party vote is about 51%.  Not bad.

Meanwhile, the state-based polling aggregators also performed as expected, with Sam Wang and Drew Linzer and Simon Jackman (I apologize to the others out there who also got it right) – pending the Florida outcome – also hitting their Electoral College projections exactly on the mark.   Yes, these models don’t tell us why the election turned out as it did, but they demonstrated once again that the best way to predict an election is to ask a sample of voters the day before how they are likely to vote.

So yesterday was a huge victory for political scientists.  But we can’t, as a profession, let down our guard.  There are pundits out there, still roaming the political landscape, spreading their punditry to the unsuspecting masses.  As I drove home last night, I heard on the NPR the first discussion of the “M-word” – (pssst – “mandate”).  Let’s be clear, no matter how much pundits say otherwise, Obama did not win a mandate last night, either prospectively or retrospectively.  What he won was a seat at the governing table for another four years – a seat from which he will find his reach growing gradually shorter as his term progresses.  All this seat provides is an opportunity to do what most presidents are allowed to do: suggest an agenda, and then draw on one’s formal powers and whatever residual influence one might have by virtue of public support and reputation to bargain with the opposing party to implement that agenda.  In this case, Republicans are going to point out that the election essentially was a vote for the status quo – not for change in Democrat’s direction.  Let the bargaining begin, starting with that fiscal cliff.

I’ll be on with a more extensive post-election analysis, but I leave you with a final warning:  now that the election is over, pundits can go back to pundicating without fear that results might prove them wrong.

Meanwhile, I leave you with this visual image (pardon my French):