Author Archive | Dan Hopkins

Waves are for Surfing

National elections sorted into two types, one with higher levels of turnout and a pronounced shift to one party. Yes, you might have thought that political science had left theories of “realigning elections” behind. But as the 2010 midterms approach, talk of “wave elections” seems to be all the rage, and the concept—though understandably fuzzy—has strong parallels with the electoral realignments of days past. So let’s put the idea to a few empirical tests.

One claim embedded in the “wave” thinking is that Congressional (or perhaps midterm) elections come in two types: those occasional elections that reflect dramatic shifts in partisan preferences, and the majority of elections that do not. If we look at the distribution of changes in average Congressional support for the Democratic candidate from election to election, the wave-based approach leads us to expect that there will be many small shifts punctuated by the occasional dramatic shift. Now, with only 19 observations on my computer, I’m not going to make any claims about the precise distribution of the changes in two-party support from election to election. But we can calculate the change from the previous election, and use a histogram to present the observations.


The key point here is a simple one: thinking of elections along a continuum seems to be more in keeping with the data than any binary concept of “waves.” The data themselves show no obvious cut-point between what constitutes a “wave” election and what does not. As an example, consider the 1982 election. Was 1982 a “wave,” with its shift of 4.5 percentage points in the average district vote toward the Democratic Party? If not, the reason is in part because the Democratic Party controlled the House of Representatives before the election, putting a ceiling on potential Republican loses. In 1982, the Republicans’ share of the two-party vote outside of the South was 43.9%, the lowest margin the GOP would win between 1974 and 2006.

Another claim is that “wave” elections are nationalized, whereas the typical election is fought more on local issues. To look at that possibility, I borrowed district-level partisanship measures from Matthew Levendusky, Jeremy Pope, and Simon Jackman from the 1960s until 2000. These measures are purged of factors like incumbency and national-level swings. I then looked at the correlation between district-level partisanship and the Democrats’ share of the two-party House vote in each year. As elections are more nationalized, we might expect higher correlations between baseline partisanship and Congressional support.


The understanding of 1994 as a nationalized election seems correct—that year was part of a dramatic upswing in the relationship between baseline partisanship and Congressional voting. But what is notable is that as the 1990s went on, that relationship remained stronger than in the prior decade. By this metric, 1996 was even more nationalized than was 1994. From these admittedly partial data, it seems that 1994 was not so much an outlier or a “wave” as it was the beginning of a newly nationalized era.

Continue Reading

The Demographics of 2010

Last year, political scientists Stephen Ansolabehere and Charles Stewart pointed out that most of Barack Obama’s increased vote total over John Kerry came from black and Hispanic voters. Those two ethnic/racial groups together accounted for an increase of 7 million votes for Obama, as compared to 3 million added votes from non-Hispanic white citizens. So in thinking about the upcoming elections for the House of Representatives, it makes sense to ask about how blacks and Latinos are represented in the most competitive districts. Consider the 42 seats currently held by Democrats that analyst Charlie Cook considers to be “toss ups.” As these races go, so goes the House in all likelihood. According to the Census Bureau, the median toss-up district’s residents in 2006-8 were 3.6% Latino and 4.8% black—as compared to shares of 15.1% and 12.3% nationally. Simply put, irrespective of turnout, the electorate that will prove decisive in which party controls the House has fewer voters of color than the electorate that proved decisive in electing Obama.

Continue Reading

Economic Perceptions in an Unequal Era

Like the economy itself, political science’s attention to economic voting is cyclical. In the late 1970s and the 1980s, the field had a vibrant set of debates about economic voting here in the U.S. One dominant argument held that economic voting was largely sociotropic—that is, responsive to economic conditions in the country as a whole. Voters were responding to their country, not their pocketbook. Before we apply the same thinking to today’s economy, though, we need to test whether its preconditions still hold. After all, sociotropic voting should only lead to clear shifts in candidate support to the extent that Americans agree about how the national economy is doing.

And there is good reason to think that we might not agree as much as we used to. In the decades since the major debates about sociotropic voting, the U.S. economy has seen important shifts, as an increasing share of income is concentrated among the wealthiest. In 1978, just prior to the peak of economic voting studies, Americans at the 95th income percentile made 6.7 times as much as their counterparts at the 20th percentile. By 2003, that ratio had grown by 27%, to 8.5. (Since much of the growth has been at the very top of the income distribution (gated, ungated), if anything, these figures understate the increase in inequality.) Given this rising inequality, it seems plausible that Americans’ perceptions of the national economy are diverging along with our actual economic fortunes. If so, the political impact of sociotropic voting might well change, since national economic performance would mean different things for different groups of voters.


The left side of the graph shows average annual incomes for Americans at different income levels, and tells the familiar story of rising income inequality. The right-hand side is less familiar. It shows the perceptions of national economic performance from the Michigan Survey of Consumer Attitudes for three income groups: the top 5 percent (black line), the top 25 percent (blue line), and the bottom 25 percent (green line). Higher numbers indicate a stronger economy in the past year. Can’t distinguish between the three lines? That’s the point: the trends are virtually indistinguishable most of the time, although those with higher incomes were quicker to perceive the economic recovery of the mid-1980s. So while Americans’ economic fortunes have diverged since the late 1970s, we still agree to a striking extent about how the national economy is faring. Inequality exists in our actual economic fortunes, but not in our perceptions of the national economy. And as the corresponding paper of mine illustrates, this consensus does not seem to be because Americans are disproportionately weighting the economic performance of those making the most money. Instead, when viewed in the aggregate, Americans’ perceptions of the national economy appear closely linked to the economic conditions of those in the middle of the income distribution. A key precondition for sociotropic voting based on actual economic conditions is alive and well.

Continue Reading

Spin and the Economy

In a blog post here yesterday, Andrew pointed out that for all of our collective insistence that the economy shapes voter decision-making, political scientists have more to learn about where economic perceptions come from. In a nation of more than 300 million residents, where economic conditions vary by region, it seems plausible that economic perceptions are shaped by what is reported in the news media. But does that mean that the economy is susceptible to media spin?

My ongoing work suggests that it is not. Or perhaps I should say “my computer’s ongoing work,” since this research makes use of computational tools to analyze more newspaper articles than I could ever read. After extracting all of the stories in the New York Times or the Washington Post that mentioned economic keywords from around 1980 until around 2006—that’s about 150,000 articles in all—I created an index of economic tone. The index adds up the monthly share of nine word stems that tend to track economic performance, word stems like “inflat” and “unemploy.” Separately, I used the Michigan Survey of Consumer Attitudes to measure Americans’ perceptions of the national economy over the last year. In both cases, the measures are of economic concern, so we should see higher values when journalists’ articles are negative in tone or when Americans are especially worried about the economy.


American public opinion typically looks stable and rational when viewed over time, and these results are no exception. Follow the blue line: Americans’ concern about the economy reliably grows during recessions and declines during expansions. But the key finding is in the relationship between the various trends. The tone of coverage in the Washington Post and the New York Times does not appear to systematically lead Americans’ economic perceptions, a point that formal statistical tests reinforce. If anything, the relationship is the reverse, with Americans’ economic attitudes shifting before we see similar shifts in print. These are only two newspapers, to be sure, but they are two prominent national newspapers. And they are commonly perceived (or derided) as influential. When viewed over the long-term, Americans appear to be responding to actual economic conditions—and not to the tone of these national newspapers.

Continue Reading

What Can We Learn From Enron’s Internal Emails?

The U.S. has “a campaign finance system that is nothing less than an elaborate influence-peddling scheme in which both parties conspire to stay in office by selling the country to the highest bidder.” That’s Senator John McCain in 1999. But it could have been many people from either side of the aisle. The view that interest groups trade campaign contributions for political favors is the prevailing wisdom.

But in an APSA panel later today, Lee Drutman and I will be challenging one part of that view using an analysis of the internal emails of the Enron Corporation. After Enron went bankrupt in late 2001, more than 200,000 internal emails from its upper-level management became public. These emails serve as a “currency of attention,” letting us see what the firm’s employees were doing on a day-to-day basis. Enron was a very politically connected firm, and its energy businesses put it into constant contact with public officials at all levels. Enron is also a good test case, because it wasn’t, um, especially worried about violating ethical norms or laws. If businesses are routinely trading campaign contributions for favorable policies, Enron should have been among the more active traders.

Using tools from computer science, we identified more than 1,000 emails that had political content, and then coded them as reflecting one of five types of activities a firm might engage in. The activities range from monitoring politics to contacting legislators, formally participating through hearings or other venues, shaping opinion, and intervening in campaigns and elections. The results?


Each of the graphs above plots the share of emails in a separate category of political activities over time. As it turns out, Enron’s staffers were not very interested in campaigns, elections, or political fundraising. An underwhelming 1.4% of political emails were about those topics. By contrast, Enron employees spent the majority of their political emails—66%—simply following events, or what we call “monitoring.” Certainly, they were also devoting considerable attention to their meetings with public officials, an activity which covers 15% of political emails. And to perhaps a greater extent than expected (9%), Enron was also involved in formal governmental processes, like rule-making or hearings. To be sure, Enron was closely intertwined with political figures—but it was only occasionally concerned with the elections that determined who those political figures would be.

Continue Reading

Hurricane Katrina and Political Science

In the five years since the Hurricane, what has political science taught us about Katrina and its aftermath?

  • Contrary to images of anarchy, there was substantial cooperation among the evacuees in the days after the storm. Sam Whitt and Rick Wilson’s research in Houston’s post-Katrina shelters undermines the popular image that the storm led to a breakdown in social norms. Evacuees played a “public goods” game, where they could divide $10 between themselves and the group. Donations to the group were then collected, doubled, and distributed equally. There is a clear incentive to free ride—to contribute nothing and reap the rewards of others’ generosity. But despite everything they had experienced, the evacuees showed considerable cooperative behavior, putting almost 40% of their money into the pool.
  • Communities that hosted Katrina evacuees responded with a remarkable mobilization of volunteers, much of it coordinated through churches. Some of my own research on Katrina details that the call to help the evacuees led to a midnight traffic jam in Houston, as thousands of volunteers rushed to assist. In fact, 36% of Houston residents reported volunteering to assist the evacuees. In Baton Rouge, the figure was even higher, at 52%. Arkansas opted to mobilize church-based networks; there, an astounding 87% of weekly church attendees made Katrina-related donations.
  • Networks could also work against the evacuees, keeping them out of communities. Daniel Aldrich and Kevin Crook (gated) find that communities with higher pre-storm levels of volunteerism were less likely to be sites for FEMA trailers.
  • Public opinion in the host communities were shaped by the unflattering public image the evacuees brought with them. That’s from more of my research. Looking at a survey conducted in the months after Katrina, I found that Houston residents’ political attitudes looked very similar to people with the same demographics living elsewhere in the South. The one important exception: Houston residents were notably more worried about crime. In Baton Rouge, residents stuck out for being more opposed to public spending on the poor.
  • The blame for Katrina’s aftermath didn’t fall entirely along partisan lines. Separate articles by Neil Malhotra and Alexander Kuo (gated) and Brad Gomez and J. Matthew Wilson (gated) consider how blame for Katrina was partitioned given that multiple levels of government were involved. As voters’ information improves, so too does their willingness to blame the poor governmental response on actors beyond the President.
Continue Reading

A Colorful Race

When political scientists arrive in Washington, DC over the next few days, they’ll be stepping into a close closely contested Mayor’s race between the incumbent mayor Adrian Fenty and the City Council Chairman, Vincent Gray. Color is playing a role in the race, and not just because the Democratic primary features candidates named “Gray,” “Orange,” and “Brown.” Voter support has split along racial lines, with African American voters preferring Gray and whites preferring Fenty. (Both candidates are black.)

One criticism of the incumbent mayor Fenty is that he has favored predominantly white neighborhoods, like APSA’s host Woodley Park and its Northwest neighbors. Chairman Gray has echoed those concerns, telling a local radio station that “[b]e it real or be it perceived, there is a view that people in certain parts of this city, especially predominantly African American [parts], have not been well served by this administration.”

And that criticism—which is by no means unique to DC—points to a gaping hole in a wave of recent studies of local distributive politics conducted by economists and political scientists. While several of our studies have looked at the relationship between local ethnic/racial diversity and public spending across cities (here, here, or here), there is little systematic work on how public goods are distributed within cities. Our recent work has focused on the share of the city’s money devoted to parks or libraries, paying little attention to where within the city those resources are going. Claims like Gray’s are common in big-city politics. But political scientists haven’t been testing them systematically.

That’s what makes work now underway by Georgetown graduate student Lindsay Pettingill intriguing. Pettingill collected all of the calls made to D.C.’s 311 hotline—over 1.5 million calls in all—from 2000-2009. These calls are service requests, and the District tracks its response times to each request. Pettingill shows below how these response times vary both by ward and by year.


First, we see substantial improvement in average response times. In 2000, it was taking the District upwards of 40 days to respond. By 2009, that figure was down to 11 on average. Most of the decline took place during the tenure of Anthony Williams, Fenty’s predecessor. The other key fact: response times across neighborhoods have converged over the years, with just two days separating the neighborhood with the longest response time from the neighborhood with the shortest response time in 2009. Calls from the heavily black neighborhoods like Berry Farm and Kenilworth don’t seem to go unanswered. In fact, it was the predominantly white neighborhoods in Northwest that initially saw the slowest response times, although those gaps have closed. Of course, this is not the only metric of bias in District services—and capital projects could tell a very different story. But if you see a broken meter outside the Woodley Park Marriott, don’t expect special service because of that green Fenty sign on the nearby lawn.

Continue Reading

Glenn Beck’s Rally and the Threat Gap

“We’ll be the checks and balances on this out-of-control, criminal government.”

Comments like that one were par for the course this past Saturday here in Washington, D.C. In what must have been a trial run for hosting the annual meeting of the American Political Science Association later this week, TV personality Glenn Beck led a “Restoring Honor” rally at the Lincoln Memorial. I want to focus not on the rally’s content but on the motives of the many thousands who attended.


(Photo credit: Lee Drutman)

Even before it took place, the rally was being discussed as another example of the “enthusiasm gap” between Democrats and Republicans leading into the 2010 midterm elections. I think a better name might be the “threat gap.” That’s because the Tea Party appears to be the latest example of a mobilization in response to a perceived policy threat—in this case, a Democratic policy agenda being pursued by at both ends of Pennsylvania Avenue. Nobel Prize-winning social science tells us that people are generally loss averse (gated), meaning that we’re likely to be more worried about shifts away from our preferred policies than excited about shifts toward them. That’s exactly what political scientists Richard Lau (gated), John Patty (ungated), and Joanne Miller and Jon Krosnick (ungated, gated) have separately argued: in motivating political behavior, threats are more energizing than opportunities. And that asymmetry hints at why there was no “Taxed Enough Already” movement driving the actual enactment of the Bush tax cuts in 2001.

Are such mobilizations in response to perceived policy threats rare in American politics? Not in the slightest: in October of 2002, January of 2003, and again in September 2005, many thousands of Americans of a very different political stripe came here to D.C. to protest against the War in Iraq. In fact, the quotation that opened this blog post, so seemingly at home in a discussion of the current right-wing mobilization, actually came from anti-war activist Cindy Sheehan at a 2005 demonstration. Yet strangely, discussions of the Tea Party almost never mention its recent left-wing analog.

At least in part, the Anti-War movement in the early part of the decade and the Tea Party movement at the end were both responses to shifts in public policy that many Americans—especially those not identifying with the party in power—saw as at odds with their values and deeply threatening. For one bit of evidence on this, take a look at national survey data from the Social Capital Community Benchmark Survey.


In both the fall of 2000 and the spring of 2006, the survey asked Americans about their political ideology, and also asked if they had attended a political meeting or rally in the last year. It’s not that ideology shifted much, at least on the left. In both years, only about 7.5% of Americans called themselves “very liberal.” But the participation of that 7.5% differed substantially. In the fall of 2000, with a Democrat in the White House and no major policy changes in the works, the very liberal respondents were only slightly more likely than other Americans to be manning the proverbial barricades: 25% had been to a rally or political meeting in the last year. In the spring of 2006, with the ongoing Iraq War and unified Republican control, that number had jumped to 39%. The participation of other ideological categories seemed unchanged. Not long ago, the threat gap cut the other way.

Continue Reading