The Still-Vital Center: Moderates, Democrats, and the Renewal of American Politics

The Still-Vital Center: Moderates, Democrats, and the Renewal of American Politics

The Still-Vital Center: Moderates, Democrats, and the Renewal of American Politics

In 2006, 2008, and 2010, America experienced a series of wave elections that resulted in shifts of political control. At the same time, the country continues to grapple with serious, potentially existential threats: a weak economy, massive structural deficits and growing global competition. Both problems—political polarization and the failure of governance—share the same root cause: the failure to give moderates adequate voice in the nation’s electoral and policy processes. This report argues for a new moderate politics that can solve these twin challenges. For Democrats, a politics of the center presents a critical opportunity—to both rebuild a lasting governing coalition and to champion a new reform agenda that can heal a political process now fractured by polarization. For America, a politics of the center presents an opportunity to solve some of the biggest challenges it has ever faced.

The Challenge

A crisis of confidence in politics and in governance

In the aftermath of the November 2010 elections, the tone of American politics has shifted.

The lame-duck session of the 111th Congress was surprisingly productive, in part because Democrats and Republicans compromised to reach agreement, and in part because some Republicans abandoned their monolithic opposition to the Obama Administration’s initiatives. Since the Tucson tragedy, both parties have worked to reduce the incivility that has disfigured our politics, and President Obama delivered a State of the Union address widely regarded as an effort to take the edge off the often-harsh conflict that dominated his first two years in office. As a result, the political parties and the President have all seen their approval jump—evidence of the public’s yearning for a style of politics less geared towards point-scoring and more toward problem-solving.1 But despite the modest improvements of recent months, America enters the second decade of the 21st century with a public whose mood remains sour. Only a third of Americans think the country is on the right track, confidence in government stands near an all-time low, and partisan polarization has reached levels not seen since the 1890s, resulting in public disapproval of both political parties. And despite the generally positive reaction to Obama’s speech, only 34% think that it signals a major change in the types of policies he will pursue.2

It is not hard to find reasons for these sentiments: an economy mired in slow growth and high unemployment in the wake of the deepest recession since the Great Depression; two costly and seemingly interminable foreign wars; and record budget deficits as far as the eye can see. While America is stuck in neutral, nations such as China, India, and Brazil are surging ahead, generating widespread fears that our best days are behind us and that further decline is inevitable. The modest goodwill of the immediate post-election period is threatened by an impending series of high-decibel budget battles that will do little to address our unsustainable fiscal course.

This is more than a list of policy challenges; it represents a failure of governance—a system that heaps up problems without ever solving them. In response, the people have repeatedly resorted to the ballot-box, but without getting what they want. Remarkably, three consecutive elections have produced large shifts in the political balance without enhancing public satisfaction. In eight of the past ten years, the share of Americans saying that the country was headed in the right direction has declined (See Appendix #1).

We argue that this crisis of governance and the difficulty Democrats have had in sustaining a governing majority have the same root—namely, the failure to give appropriate weight to political moderates in our electoral and policy processes. But these problems also have the same cure—adopting the kinds of structural changes that will amplify moderates’ voices.

While the greater inclusion of moderates on both sides of the political divide would benefit the nation as a whole, we focus our case in this paper on why moderates have particular significance for Democrats. To flesh out this thesis, we present evidence and arguments for three basic propositions:

  • Moderates are an essential ingredient for building a lasting Democratic majority. Democrats cannot build sustainable state-wide or national majorities without winning a super-majority among voters who identify themselves as moderates. At the national level, self-identified liberals constitute barely one-fifth of the electorate; in most states, they are nowhere near a plurality—let alone a majority.
  • Moderates are not “liberals in disguise.” Moderates as a group have distinct viewpoints and policy preferences. They are neither liberals in disguise nor conservatives with a more conciliatory style. Nor are they Independents with a different label. Only 47% of moderates are Independents; only 43% of Independents are moderates.3 Their political orientations and attitudes differ significantly.
  • The current political process disfavors moderates. The basic structures of American politics—ranging from closed primaries to Congressional redistricting procedures—systematically tip the scales against moderates and reward candidates closer to the political extremes.

It is in the long-term interests of the Democratic Party to champion reforms that will level the playing field for the moderate voters and candidates that comprise such a critical segment of the Democratic coalition. But not only will Democrats benefit from giving moderates increased attention and weight, so will the political system as a whole. Polarization will diminish, and policy-making will be more likely to yield sustainable outcomes that large majorities of the electorate can endorse.

These reforms will address a central source of public dissatisfaction with modern politics: it will ensure that the system fairly and fully reflects the sentiments of moderate citizens whose voices are muted by current political arrangements.


Why moderates matter for a sustainable majority

While liberals are the ideological base of Democrats, they represent a small minority of the overall electorate. Democratic Congressional victories and President Obama’s election were not triumphs of liberal mobilization but the result of a basic axiom of contemporary political arithmetic: Democrats cannot win national majorities without winning a super-majority among Americans who regard themselves as neither liberal nor conservative. A “base strategy” may be barely viable for the Republican Party; it is an electoral and governing dead end for the Democrats.

As the dust settles on the rubble of the Democratic majority in the House, and eyes turn toward the 2012 general election, Democratic strategists are offering their party and the White House a familiar mix of competing advice. Some argue that without an agenda that re-energizes the base, both President Obama and the 23 Senate Democrats up for reelection will face tough sledding. Others point to the huge shift among Independents as the key problem.

The debate between a base strategy and a focus on Independents represents not only a false choice but also a flawed analysis. The Democratic base is defined—and should be understood—in ideological terms. Core Democrats are liberals, as that label has come to be viewed over the past four decades. But liberals constitute a small share of the electorate that has not increased in four decades.

Liberals—a small and static share of the electorate

Over the past three decades, the ideological composition of the electorate in presidential election years has remained remarkably stable, as shown in Table #14:

TABLE #1: The Composition of the Electorate in Presidential Election Years

Midterm elections, in which turnout is about one-third lower, are much more variable. For example, compare 2006 and 2010. In 2006, the “national” electorate (the total of all votes cast in House races) was 32% conservative, 47% moderate, and 20% liberal. In 2010, the liberal share remained the same while conservatives surged to 42% (10 percentage points over 2006) and moderates dropped by 9 percentage points, from 47% to 38%. Despite doing a bit better among liberals in 2010 than in 2006, Democrats suffered a huge reverse, largely because the composition of the electorate shifted so dramatically. This is not to say that the composition of the electorate is the only thing that matters in midterm elections. For example, if Democrats had done as well among moderates in 2010 as they did in 2006, their share of the total vote would have increased by about 2 percentage points, reducing the overall Republican margin by more than half. And if, in addition, they had done as well among conservatives in 2010 as in 2006, they would have held their majority, notwithstanding the electorate’s compositional shift.

Since 1980, no Democrat has been elected president without winning at least 60% of the moderate vote cast for the two major parties.

It is possible that the decades of stability in the presidential electorate may be coming to an end. Since Barack Obama’s election, when moderates and conservatives were about equally balanced in the adult population, conservatives have opened up a five-point edge (40% to 35%) over moderates in the Gallup tracking survey.5 (The liberal share has remained constant.) If so, the hill the president must climb in 2012 may be a little steeper.

Should this shift come to pass, it will be even more important for Obama to do well among moderates. In 2008, Obama carried 60% of the moderate vote, providing fully half his total vote. (By contrast, liberals constituted about 37%.) If the electorate in 2012 looks anything like the electorate in 2010, as it could if the Gallup numbers hold up, he would have to be even more dominant among moderates than he was during his first presidential campaign.

Moderates—the true presidential kingmakers

Moderates tend to vote Democratic. Since 1980, Democrats’ share of the moderates’ two-party vote has ranged between 46% in 1980 and 62% in 1996. 60% of moderates supported Obama in 2008, while the rest chose McCain. Obama did much better among moderates than either Gore or Kerry, but roughly the same as Bill Clinton did in 1992 and a bit worse than Clinton did in 19966.

TABLE #2 Democratic share of the major party moderate vote

President Obama’s 2008 victory is widely misunderstood as a triumph of liberal mobilization. The truth is more complicated. He received 52.9% of the popular vote, versus John Kerry’s 48.3% four years earlier—a gain of 4.6% of the electorate. Table #3 shows the ideological composition of that gain.7

TABLE #3 Ideological composition of the Democratic Presidential Vote

The largest increase in the Democratic vote share came from moderates, who were also the largest grouping in the electorate. Gains among moderates thus contributed more to Obama’s improvement over Kerry than did gains among liberals. In fact, gains among conservatives were as significant as those among liberals. Overall, liberal mobilization contributed only one-third of Democratic gains between 2004 and 2008—significant, to be sure, but hardly the whole story. Indeed, gains among liberals would not have sufficed to put Obama over the top. He prevailed by persuading substantial numbers of moderates and conservatives that he could be trusted to govern as president of all the people and not just as a liberal champion. If Obama had received only 54% of the moderate vote, as John Kerry did in 2004, his popular vote would have shrunk to 50.3% of the total, and he might well have lost in the Electoral College.

President Obama’s 2008 victory is widely misunderstood as a triumph of liberal mobilization.

“Base-ic” math—the necessity of a Democratic liberal-moderate coalition

There is no possibility of purely liberal or purely conservative governance that can sustain itself over time, for the simple reason that neither liberals nor conservatives constitute majorities of the electorate. To win, each party needs to form a coalition with moderate voters. But the structure of those coalitions is very different. Because conservatives command roughly twice as large a share of the electorate as do liberals, they can make do with less than a majority of moderates.

Table #4 illustrates the difference between the two parties’ winning coalitions. When Democrats win the presidency, their electoral coalition always consists of more moderates than liberals. When Republicans win the presidency, their electoral coalition typically consists of more conservatives than moderates—with the sole exception of the 1980 Reagan landslide where moderates were slightly more numerous than conservatives.8

TABLE #4 The Composition of Democratic versus Republican Governing Coalitions

(Plus sign (+) indicates more of a dependence on moderates than on the base of the party. Minus sign (-) indicates more of a dependence on the base than on moderates)

Democrats begin with a structural advantage among moderates, but they cannot afford to take it for granted. Conversely, while conservatives are less dependent on moderates to achieve electoral majorities, the Republican Party cannot afford to write them off either. In the long run, a hard-edged, uncompromising conservatism will alienate the substantial minority of moderates that Republicans need to govern effectively.


Moderate voters and what they believe9

Moderates as a group etch a distinctive profile. They are center-left on social issues, middle of the road on economics, and center-right on foreign policy. They are not “liberals in disguise,” nor are they a “mushy middle” of left-right compromise.

Because of the longstanding edge that Democrats have enjoyed among moderate voters, many analysts and party activists have concluded that moderates are basically liberals in disguise—that is, voters whose thoughts and sentiments correspond to those of the Democratic Party base but who prefer not to call themselves liberals because modern politics has given the term a bad name. Our research does not support this proposition.

n many measures, to be sure, moderates are somewhat closer to liberals than to conservatives. Still, the differences between moderates and liberals are very significant. Nor is it the case, as others believe, that moderates form a “mushy middle”­­­—an ad hoc series of compromises between two coherent ideologies. In fact, moderates as a group constitute a complex but distinct political orientation.

In the first place, moderates have mixed opinions about the overall stances of the two parties. When asked which party comes closer to their views on the three major baskets of issues—economic, social, and foreign policy/national security—moderates responded as follows:

Moderate Voters on Which Party Comes Closer to their Views on Issues

Moderates most resemble the Democratic base in their social attitudes, although even there (as we’ll see) there are some important differences. By contrast, voters who identify themselves as liberal or very liberal see themselves as closer to Democrats by overwhelming margins in all three areas.

Moderates versus liberals: Government, business and the economy

Compared to liberals, moderates are more skeptical of government, more supportive of business and much more likely to worry about deficits.

Moderates are more inclined than liberals to be skeptical of government, which makes them more favorable toward private sector-led growth strategies and deficit reduction than are liberals. Consider their responses to familiar survey questions:

Which statement comes closer to your view: (1) government is almost always wasteful and inefficient; or (2) government often does a better job than people give it credit for?



Very Liberal

Statement 1




Statement 2




If you had to choose, would you rather have a smaller government providing fewer services or a bigger government providing more services?



Very Liberal









50% of moderates, versus 29% of liberals, believe that the federal government has too much power. 52% of moderates, versus only 32% of liberals, say that “government is trying to do too many things that should be left to individuals and businesses.”

Given that moderates are more skeptical of government than are liberals, it is not surprising that they strike a different balance between the public and private sectors. For example, a Third Way survey conducted in mid-2010 asked:

Which of the following statements do you agree with more: (1) Large companies have too much power, hurt the middle class, and government needs to keep them in check; or (2) American companies are the backbone of the U.S. economy and we need to help them grow, whether they are large or small?



Statement 1



Statement 2



Only 31% of moderates, versus 47% of liberals, believe that there is too little government regulation of business. Indeed, fully half of all moderates believe that government regulation of business usually does more harm than good.The Third Way survey also asked:

Which approach to strengthening the economy do you prefer: Making new government investments to help jumpstart private sector job creation and economic growth; or Cutting taxes for businesses to help jumpstart private sector job creation and economic growth?



New public investments



Lower taxes



Not surprisingly, moderates’ lower regard for public investments goes along with a greater appetite for deficit reduction:

Which would you rather see Congress do to help the U.S. economy: Invest in U.S. research, innovation, and new technologies like clean energy; or Cut government spending and reduce the deficit?



Invest in innovation



Cut spending/reduce deficit



Which would you rather see Congress do to help the U.S. economy: Invest in new roads, bridges, and other infrastructure; or Cut government spending and reduce the deficit?



Invest in infrastructure



Cut spending/reduce deficit



Note that the contrast between moderates and liberals concerning public investment is not black and white: substantial percentages of moderates favor an investment strategy as well. The point is rather that while public investment is the default position for liberals, moderates need to be persuaded.

Given moderates’ skepticism about the role of government in the economy, it is not surprising that they have been less enthusiastic than liberals in their view of the economic policies President Obama pursued during his first two years in office.

Since taking office, have President Obama’s economic policies made economic conditions better, worse, or not had an effect so far?



Very Liberal









No Effect




Nor is it surprising that moderates—unlike voters to their left—are ambivalent about health care legislation, which stirred more controversy about the role of government than did any other issue during President Obama’s first two years.

Do you approve or disapprove of the health care legislation passed by Barack Obama and Congress in March?



Very Liberal









A post-election survey confirmed this finding: moderates favored the legislation by a narrow 7-point margin, compared to 48 points for liberals.10

Liberals, moderates, and conservatives agree that the cost of Social Security and Medicare will create major problems for the U.S. economy if no changes are made. But while liberals endorse raising taxes to finance these programs by a margin of 65% to 35%, moderates are split down the middle, 49% to 49%. (Solid majorities of liberals, moderates, and conservatives disapprove of cutting benefits in these programs.)

Other evidence confirms the gap between moderates and liberals on a wide range of economic and fiscal issues:

  • 61% of moderates think that labor unions have too much power, compared to 48% of liberals.
  • When presented with the proposition that protecting the environment should be given priority, even if that causes slower economic growth and some job losses, moderates split down the middle, 47%/49%. By contrast, about 70% of left-leaning groups agreed.
  • Only 46% of moderates endorse the proposition that “Government should help more needy people, even if it means going deeper in debt,” compared to 63% of those who consider themselves liberal or very liberal.

Moderates versus liberals: “guns, gays and God”

The one place where moderates come closest to being “liberals in disguise” is on social and cultural issues such as abortion, gay equality and guns. For example, 56% of moderates think that abortion should be legal in most or all cases. Only 42% think it is more important to protect the right to own guns than it is to control gun ownership. 72% agree with Obama’s decision to permit federal funding for most embryonic stem cell research. Only 24% believe that school boards should have the right to fire gay teachers. On the fraught question of whether gays and lesbians should be allowed to marry legally, a plurality of moderates (48%) answers in the affirmative. Like liberals, moderates do not believe that immigrants threaten traditional American customs and values. On the contrary, fully half believe that today’s immigrants strengthen our country because of their hard work and talents.

Despite these sentiments, moderates are significantly more socially conservative than are liberals. To illustrate: Moderates are more likely than are those to their left to think that poor people have become too dependent on the government, that too many people want to get ahead without working hard and making sacrifices, and that we have already made the changes needed to give African-Americans equal rights. Only 37% of moderates, but fully 59% of liberals, believe that the federal government is responsible for reducing income differences between rich and poor.

The differences between moderates and those to their left extends to religion and traditional values as well. They are more likely to believe in God and in a Day of Judgment, to say that prayer is important in their daily lives, to espouse “old-fashioned” values about marriage and family, and to endorse clear guidelines of good and evil that are binding on everyone.11

Positions on religion by ideology

Religious differences between moderates and liberals extend beyond belief to behavior. By a margin of 12 percentage points, moderates are more likely than liberals to report regular church attendance, a key indicator of religiosity.

Moderates versus liberals: Hawks and doves

When it comes to foreign and defense policy, moderates slightly prefer the Republican Party. It is no surprise, then, to find that they are less likely than are liberals to see diplomacy rather than military force as the key to peace and more likely to believe that we should fight for our country whether it is right or wrong. By a margin of 52% to 40%, moderates believe that the United States was right to use force in Afghanistan; by the same margin, those who consider themselves liberal or very liberal disagree. (When it comes to Iraq, however, moderates line up with the left against conservative and very conservative voters.) And unlike liberals, moderates embrace American exceptionalism: fully 54% endorse the proposition that “God has granted America a special role in human history,” compared to only 34% of liberals.

Moderates and liberals: Partial convergence

We are not suggesting that moderates disagree with liberals on every issue. Indeed, the areas of convergence are significant, and not just in the area of values-laden social policy. Moderates are strong environmentalists; they believe that the federal government is totally responsible for protecting consumers against unsafe products; they want government to provide jobs, a minimum standard of living, and health care for all. And they are far closer to liberals than to conservatives on the bitterly contested issue of immigration reform. Roughly half of both moderates and liberals endorse an immigration compromise that would combine better enforcement and border security with a path to citizenship for illegal immigrants who are willing to comply with the appropriate requirements. By contrast, half of all conservatives give priority to enforcement and border security.

To sum up: moderates as a group etch a distinctive profile. Although (unlike Independents), they tilt towards the Democratic Party, they are not liberals in disguise, and they are open to certain conservative appeals that few liberals would consider. A sustainable center-left majority will require a broad-based agenda that liberal activists are likely to regard as excessively compromised. Similarly, a viable center-right majority would require Republicans to embrace a big-tent approach to social issues, to put forward economic policies that reflect pragmatism and balance, and to acknowledge that military power is but one of the tools that a great power can deploy to advance its national interest.


The challenges of governance and the muting of moderate voices

Given the arguments made above, the following question arises: If moderates are a potent and coherent political force, why is modern American politics so polarized?

The answer lies in systemic structural problems that silence the voice of moderate voters. The most significant of these structural problems are: (1) the current system for primaries; (2) redistricting and (3) the leadership system in Congress.

Since Bill Clinton’s election in 1992, moderates’ share of the electorate has fallen by eight percentage points, from 43% to 35% of the electorate, while conservatives and liberals have each increased their share by four percent—conservatives from 36% to 40%, liberals from 17% to 21%.12 During the past decade, moreover, the bases of both parties have increased their share of their respective parties, in part because the ideological poles are stronger than they were, and in part because the parties have sorted themselves out ideologically. Many moderates have fled the Republican Party, while many conservatives have deserted the Democrats.

And as the ideological bases of both parties have become more dominant, they have become less willing to entertain the kinds of big-tent strategies and legislative compromises needed to forge stable majority coalitions. Voters outside the parties’ bases have become more frustrated, and their voting behavior more volatile.

But changes in the composition of the electorate and the parties take us only part of the way toward an explanation of both parties’ failure to seek stable coalitions with moderate voters. Why do Democratic Presidents and Democratic Congresses especially seem to lose their way (and subsequently their majorities) so quickly? The answer lies in some basic structural features of American politics ranging from Congressional redistricting procedures to closed primaries. These systematically tip the scales against moderates and reward candidates closer to the political extremes. These structural features have the same consequences for Republicans as they do for Democrats. But because the Republican conservative base is twice as big as the Democratic liberal base, the results for Republicans are not as consequential.

“Base-ic” instincts: Primaries and polarization

The primary electorate is different, and more extreme, than the general election electorate. This explains why, every election cycle, so many Americans are so unhappy with their choices. It also explains why it is so difficult for moderates to gain influence in American politics. All politicians, regardless of their expressed desire for moderate, bipartisan governance, have to cater to their primary constituency. As a result, the primary process exercises a tyranny over American politics that exacerbates polarization.

Unlike nearly all other democratic governments in the world, the American democracy is the only democracy that nominates candidates for major national offices using a primary system. In most of the world’s other, parliamentary, democracies, political parties assemble in annual or bi-annual conferences to conduct party business. Candidates for Parliament are slated in party conferences and assigned to districts. The most important criteria for selection to the ticket are loyalty to the party and the ability to contribute to government. Connection with the electoral district is not a very important consideration.

In contrast, the opportunity to stand for political office as a Democrat or a Republican is open to anyone in that district who wishes to so identify.

One of the reasons that the American primary system is so porous and therefore unpredictable is that only a very small portion of voters participates in primaries. Two factors are key. First, in half the states, voting in primaries is restricted to those who are registered to one of the major political parties (See Appendix #4). The eligible universe for primary voters omits larger portions of voters who consider themselves Independents or who simply don’t want to declare a party preference. Second, primary contests often attract little attention. By definition, primaries precede the high-energy biennial national elections and are therefore easy to miss. Dates for primaries vary from state to state, with the result that they do not attract the national media attention that general elections do. In addition, political parties often schedule primary elections at times—the summer, for instance—when the American public isn’t especially interested in politics. As Appendix #5 indicates, the 2010 primaries fell on 19 different days and were heavily concentrated in the summer months when people are on vacation and attention to politics is low. This contributes to the propensity for these contests to be very low turnout elections.

Table #5 illustrates the low levels of turnout in primaries for President­—including those years in which an incumbent President had little or no opposition. Overall average turnout in presidential primaries is exceptionally low because as the season wears on results from the early states winnow out the field of candidates. Because of its long history as the first in the nation primary (a position its citizens greatly value), New Hampshire regularly enjoys by far the highest turnout. In many years, however, the race is all but over after the New Hampshire primary, reducing incentives for citizens of states coming later in the cycle to participate.

The choice of presidential nominees is made by a substantially smaller number of voters than those who turn out in November. Not only is total turnout in presidential primaries low but turnout in each of the major political parties’ primaries is a fraction of those who participate the general election. As the political scientist Curtis Gans points out, with the exception of 2008, turnout in presidential primaries has been steadily falling since the early 1970s, and turnout in statewide primaries has been falling as well. (The 2008 presidential primaries were unusual in that the race between Hillary Clinton and Barack Obama was so close that it went on to the end, boosting turnout in states towards the end of the season.) Even in the record high turnout year of 2008, statewide primaries that were held on a different day from the presidential primary had miniscule turnout levels—slightly less than 15%!1314

TABLE #5 Turnout in Presidential Primaries, 1976-2008

Low turnout is an enduring characteristic of Congressional as well as presidential primaries. The following chart, taken from an article by David W. Brady, Hahrie Han and Jeremy C. Pope, compares turnout in contested primaries and general election turnout from 1956 to 1998 by Congressional district.15 As expected, turnout in the general election varies greatly from presidential election years to off years. Primary turnout, however, remains fairly stable and low, regardless of the year. And general election turnout is approximately three times as large as primary election turnout.16

FIGURE 2 A Comparison of Primary and General Election Turnout, 1956-1998

As Table #6 illustrates, turnout in Congressional primaries in the first decade of the 21st century has been no different—continuing a trend of consistently low turnout in primary elections. If general election turnout in Congressional elections is poor, turnout in primary elections is woeful. This is somewhat surprising for the 2010 election cycle, given the attention paid to those primaries because of the emergence of the Tea Party as a new and exciting political force. And yet turnout in those primaries is only about 20% of turnout in the general election—even lower than the averages Brady et. al. found for the previous decades.17

TABLE #6 Congressional Primary Turnout Compared to General Election Turnout

** Turnout for 2002 and 2006 is somewhat higher in Census reporting, however, the absence of Census data for 2010 caused us to use data from the United States Elections Project in order to maintain consistency.

Of course, the above figures are calculated by adding turnout in primaries for both political parties. In 2010 Democrats were unenthusiastic about their party while Republicans felt they had the wind at their backs. Therefore average turnout in the primaries should be greater for Republicans than for Democrats. While this is indeed the case, the differences in turnout between the two parties are not as large as the differences in publicity between the two would leave one to believe. As Table #7 illustrates, Republican turnout—calculated both as a mean and as a median—is greater than Democratic turnout.18

TABLE #7 Turnout in the 2010 primaries by Political Party

While energy and enthusiasm helped make Republican turnout higher than Democratic turnout in 2010, a bigger factor is whether or not the primary is open or closed. Closed primaries restrict participation in the primary to some form of stated affiliation with one political party or the other. Open primaries allow the voter to choose which primary to vote in. As expected there is a significant turnout difference between open and closed primaries as Table #8 illustrates.19

TABLE #8 Differences in turnout in contested primary elections 2010 –  Open versus Closed primary systems

So what does the fact that a very small portion of the population votes in primaries mean? Simply this: candidates who are most acceptable to the left of the Democratic Party and the right wing of the Republican Party tend to do better than moderates. Therefore, even candidates who are committed to working across the aisle find themselves having to constantly watch their left (or right) flank. Congressmen in safe districts need to be vigilant lest they face a contested primary and lose the race to a challenger who has managed to mobilize a large enough portion of the vote to prevail. The large uptick in the total number of contested primaries in 2010 and the surprise results in some of those primaries will reinforce the already powerful incentives that members of Congress have to watch their base carefully.

Although there is a wealth of data from exit polls illustrating the dominance of party bases in presidential primaries, exit polls for Congressional primaries don’t exist. Nonetheless, political scientists have figured out ways to test the ideological slant of primary voters. Using newspaper and journalistic accounts of primary losses from 1982 to 2000, Brady et. al. conclude that:

“As expected, Democrats who are conservative and Republicans who are liberal relative to their districts are significantly more likely to draw primary opposition… Incumbents who veer from their primary constituency are much more likely to draw challengers and thus are more vulnerable to primary defeat.”20

Thus Members of Congress learn to pay attention to their primary voters. Never mind that the victorious primary election faction is not ever large enough to carry a general election. The primary system creates a tyranny of the wings that haunts even the most determinedly bipartisan legislator.

This dynamic is even more intense in those states that do not use primaries to nominate their candidates. In those states the unanticipated presence of only a few hundred heretofore un-mobilized activists can deny seasoned politicians the nomination. This is very common in presidential primaries where the first-in-the-nation Iowa Caucus has long been known as a place where the Republican right and the Democratic left rule, and underdogs from Jimmy Carter to Pat Buchanan to Barack Obama have upset the party establishment. But it happens in Congressional races as well; witness the fate of Republican Senator Bob Bennett, who lost the nomination in the Utah state convention, where the outcome was controlled by 3,500 party activists.


The process of electoral gerrymandering (drawing Congressional district boundaries in such a way that they create a district which is almost impossible for one party to win—or lose) exacerbates the tendency towards the extremes in the primary system. Considered in tandem with low turnout primaries, gerrymandering further diminishes the impact of moderate voters in the American system.

At any given time in modern American history, no matter how upset the public may be at their politicians, the vast majority of members of Congress can expect to be re-elected. The recent 2010 midterms, where 63 seats changed party, is indeed historic. And yet 63 seats constitute only 14% of Congress!

The number of competitive House seats has declined dramatically over more than 100 years of American history—from 187 seats at the turn of the 20th century to an average of 58 seats at the turn of the 21st century. Table #9 shows that the number of very safe seats (defined as seats in which incumbents win with more than 60% of the vote) has increased in nearly every decade of the last half of the twentieth century. The most recent midterm election featured the second-highest number of House members defeated in a general election since the end of World War Two—54 in 2010, compared to 68 in 1948. But this is a rarity indeed. The median number of house seats lost in midterm elections between 1946 and 2010 the number is 22—5% of the House of Representatives.21

TABLE # 9 Average % of House Incumbents Winning  with At Least 60% of the Major Party Vote

Members of Congress in safe districts can still be defeated in a primary election. For many members of Congress, then, the election they fear most is the primary. Thus they pay particular attention to the policy desires of their primary electorates. And, as Brady have shown, they pay a price when they veer too far from the preferences of primary voters.

Gerrymandering of Congressional districts to create safe seats is the second structural factor that mutes moderates’ voices and increases political polarization. To be sure, other factors have also contributed to the creation of safe seats. In The Big Sort: Why the Clustering of Like-Minded America is Tearing Us Apart, Bill Bishop shows how Americans have been moving, voluntarily, into geographic communities where they find others who are like them—economically, culturally and politically.22 Given these trends, there is clearly a limit to the number of competitive Congressional districts that could be created if redistricting were removed from the self-interested political process in state legislatures and gerrymandering reduced. Similarly, Brookings’ Thomas Mann cautions that there is little data to support a link between independent redistricting and changes in Congress. Nonetheless, scholars like Mann agree that partisan redistricting has an effect and are involved in projects to increase the transparency of the redistricting process.

In the broader public, the same dissatisfaction with partisan gridlock that led California voters to do away with party primaries recently, has also been felt in states that have decided to put the redistricting process in the hands of non-partisan bodies. The number of independent redistricting commissions has grown over time, albeit slowly. The first four were established in Hawaii, Iowa, Washington and Montana in the 1980s. They were followed by Idaho, New Jersey and Arizona in the 1990s, and then by California’s passage of Proposition 11 in June 2008 establishing the California Redistricting Commission that will work on redistricting following the 2010 census.

While the experience with independent redistricting commissions is too small and too recent to enable us to draw any firm conclusions, there is some intriguing evidence from California, which has been at the center of redistricting battles for the past forty years. In 1971, and again in 1991, Republican governors blocked gerrymandering efforts on the part of the state’s Democrats. The issue ended up in the Courts and resulted in the creation of three-judge panels serving as “Special Masters” in charge of redistricting. Because of this, the Rose Institute in California was able to study a sort of controlled experiment. According to the authors,“The dark bars with white numbers represent the ‘Highly Competitive’ districts that elected a Democrat one year and a Republican in another year during the same decade (zero in the 2000s). The light bars with black numbers represent additional districts that were competitive in each plan, meaning either the average margin between the first and second place candidates in each election was under ten percent, or more than half of the elections in that district during the decade were decided by less than ten percent. The * appearing next to “1980s” denotes that this column’s figures include the few competitive districts of both the 1982 Democratic partisan gerrymander with those of the 1984 bipartisan incumbent protection gerrymander.”23 Sure enough, in those decades where judges drew the districts, there was an increase in the number of competitive districts. In those decades where the legislature drew the districts, the number of competitive districts remained small.24

GRAPH #1 Competitive Districts in the California Legislature

While scholars tend to study either Congressional primaries or Congressional redistricting in isolation, it is the combination of the two that heightens the impact of ideology. Over time we can see the effects of polarization in legislative behavior. Scholars who have studied Congressional voting behavior find that in the middle of the 20th century there was substantial overlap among the two political parties.25 In mid-century, American Congressional politics included Southern conservative Democrats and Northern, liberal Republicans. Each political party in this era was indeed a big tent, home to some number of House members and Senators whose ideological predispositions were closer to members of the opposite party than to their own. Reinforcing these broad coalitions was the fact that in this era all members of Congress shared a common understanding of the threats to America.

The collapse of the postwar consensus—on containing communism as the centerpiece of international policy, on government as the Keynesian manager of the economy, on culture as a sphere of contestation that should remain outside of politics—entailed the loss of shared assumptions. By the beginning of the millennium, Congressional voting returned to a pattern seen prior to the Great Depression and World War II. In our era, as in that earlier era, there are very few members of Congress who could constitute a moderate middle.26

Congressional Leadership

In the very first vote of every new Congress, those seeking leadership positions are required only to win votes from their caucus. They are not required to reach across the aisle. The result is that House and Senate leaders reflect a majority of the majority, which is almost always a minority of the entire legislative body. Rather than promoting comity and compromise, the method of selecting Congressional leadership reinforces the influence of each party’s base.

As we have seen, current Congresses are pulled apart by forces that are increasing partisanship by magnifying the clout of each party’s base. Opening up the primary process to more voters and reforming Congressional redistricting are two ways of increasing the clout of moderate voters in the governing process—a goal that is, as we have shown, important for the American political process but more consequential for Democrats than for Republicans. There is, in addition, a change in Congressional procedures that could move the system farther down that road.

Currently the Speaker of the House and the Majority Leader of the Senate are elected by a majority vote of the body. In practice this means that no matter how thin a party’s majority may be, that party will elect the leaders in the portion of Congress it controls. In turn, the House and Senate leadership is nominated in each party’s caucus. In those caucuses, the dominant faction with the most seniority tends to be made up not of representatives from competitive Congressional districts and states, but rather members who can count on winning general elections from districts and states where their party enjoys a supermajority. When the Democrats win, the leadership is determined by safe—that is, left-leaning—jurisdictions; when Republicans win, the leadership is determined by those that lean right.

Moving Forward

Three ideas for a reform agenda

The procedures we have discussed all have the effect of muting moderates’ voices, and denying them political influence commensurate with their share of the electorate. We offer three reform ideas that can help remedy this situation. These reforms will appeal to an electorate that has always been suspicious of political parties and supportive of more open electoral systems and, while it is in the long-term interests of all Americans to become champions of reforms that will level the playing field for the moderate voters and candidates, the consequences of these reforms are particularly important to the Democratic coalition.

IDEA #1: Open primaries

All Americans, but especially Democrats, have an interest in exploring ways to broaden participation in the primary process. There are two ways to do this. The first is to simply get rid of closed primaries. Given the greater interest and intensity of party activists in both parties, they would still dominate most primaries most of the time. But the possibility of expanding the primary electorate would cause candidates to look at primaries differently. It would open up the possibility that moderation and compromise might be rewarded rather than punished.

A more dramatic option would be to do what California did in its June 8, 2010 election. Proposition 14, called the “Top Two Primaries Act” passed with 53.8% of the vote. It overturned the closed primary process by creating, in essence, one political primary for both parties. Voters will be allowed to vote for any candidate, regardless of the voter’s party affiliation. The top two candidates will then face off in the general election. Prop 14 applies to all elections, with the exception of presidential primaries and elections for party office. Proponents of the Proposition argued that it might cause more moderates to be elected to the California state legislature. Passage of Prop 14 is evidence of the fact that many Californians felt that their government had become dysfunctional as a result of extreme partisanship stemming, at least in part, from the closed primary process.

It remains to be seen whether Prop 14 will result in more moderate governance. The United States has very little experience with so called “blanket” primaries. To the extent that they have been used, it has mostly been in one party states and for local elections. The exception is Washington State, where the blanket primary was approved in the courts in 2007. In the 2010 primaries, turnout over Washington’s nine Congressional districts averaged 155, 327 per district—nearly three and a half times greater than the average turnout in contested primaries. Some districts, such as the liberal 7th Congressional District, re-elected the incumbent Jim McDermott with 80% of the vote. But in general, the August 17th primary drew many candidates and resulted in high turnout. It makes sense that as, over time, the primary electorate expands to look more like the general election electorate, the impact of moderates will increase and members of Congress will be somewhat liberated from the tyranny of small primary electorates, potentially enabling them to engage in more productive governance without risking their political careers.

IDEA #2: Real redistricting reform

As we have seen, it is easy to overestimate the impact of gerrymandering on our politics. Polarization has been increasing at all levels of government—including those—such as states and counties—where lines are not periodically redrawn for partisan advantage. But for those who would like to decrease political polarization by enhancing the influence of moderate voters, reforming the process of drawing legislative districts by taking redistricting out of the hands of state legislatures and placing it in the hands of non-partisan commissions is an obvious place to start.

IDEA #3: Super-majority selection of Congressional leadership

Finally, what if it took 60% of each body to elect the Speaker of the House and the Majority Leader of the Senate? Unless one party commands a supermajority, the very first vote would then test the ability of aspiring leaders to construct the bipartisan coalitions that are so integral to effective governance. The 60% rule would require the leadership of the majority party to establish a relationship with at least some faction of the minority party. The consequences for legislation are obvious. In the Senate, the magic number needed to break a filibuster would be achieved first in the leadership elections; presumably this would make it easier to obtain 60 votes for legislation. In the House, the leadership election would establish a majority large enough to overcome the inevitable lapses in party discipline—assuming, that is, that the legislative agenda the Speaker supports is consistent with the thinking of the supermajority needed to attain that position of leadership.

This proposal goes somewhat against the grain of current thinking. Some have called for eliminating the filibuster rule in the Senate so that legislation could be passed with simple majorities. However, there are good reasons why we should want major legislation in the United States to be passed by large majorities. First super-majorities guarantee ownership by both political parties. In such circumstances, it is more likely that the legislative process can address difficult problems—for instance, our long-term structural budget deficit—without proposed solutions being manipulated by either party for political gain. Second, significant pieces of legislation require years of careful implementation. During this period, party control of government is likely to shift. Leaders of a political party that has no stake in a program are not likely to work hard to make sure that the program is well implemented. In fact, just the reverse might occur—with failure leading the former opposition party to say, in effect, “We told you so.”


We should heed the calls for greater comity that multiplied after the recent tragedies in Tucson. But the forces that push us apart are entrenched in dysfunctional institutions. Unless we adopt a reform agenda, civility will be at best cosmetic, hyper-polarization will persist, and the American political system will remain unable to address the problems that beset us. If that happens, we will begin the descent that our adversaries anticipate and our allies fear.

Although much of this paper focuses on the challenges the Democratic Party now faces, we have ended with proposals for broad institutional change. The reason is straightforward: the Democratic Party’s problems are embedded in the pathologies of the contemporary American political system. Trying to solve the former without facing up to the latter would be little more than a cosmetic exercise.

Our guiding premise is that today’s polarization is pathological. Not everyone agrees. And to be sure, the sorting-out of our party system has presented the American people with clearer, more intelligible choices that strengthen the tie between elections and governance. Citizens who vote for a Republican majority today can be confident that they’ll get an anti-tax, socially conservative agenda. As recently as the mid-1970s, they could have enjoyed no such confidence.

We believe, however, that the losses from ever-growing polarization greatly outweigh the gains. The analytical basis for our conclusion is well-established, and we need not repeat it here. Suffice it to say that polarization makes it far more difficult to conduct the essential tasks of governance, from staffing an administration to addressing complex problems whose solutions require bipartisan support. Polarization makes it easier for elected officials to cross the line separating robust democratic discourse from deliberate misrepresentation and personal calumny. It contributes to diminished trust and confidence in our public institutions. And it leads to a view of politics as warfare in which contested issues are never resolved but rather are endlessly relitigated. A nation that cannot adopt and then hew to a steady course is by definition incapable of solving problems that require sustained and widely supported purpose. The systemic reforms we have proposed will not by themselves solve the crisis of governance we face, but they will signal a willingness to get past the structures that serve to divide rather than unite.

The political logic is inescapable: Democrats cannot sustain a majority coalition without winning the lion’s share of the moderate vote; a Republican party that loses touch with moderates will have difficulty avoiding a hard-right turn that undermines the possibility of compromise. Without compromise, it will be hard to institute and maintain the policies we need to surmount our problems. Without compromise, the people will continue to regard both our government and our politics with disdain.

There are structural reasons why each party caters to its base. Party primaries are designed to be low turnout affairs catering mostly to activists. The drawing of Congressional district lines to create safe districts only increases the importance of each party’s base in primary elections. And the rules for selecting leaders in Congress do not include any test of a leader’s potential to work across party lines. The institutional reforms we suggest cannot by themselves overcome decades of polarization and create the more consensual politics that the majority of our citizens crave. But by making each party’s base more responsive to moderate voters, they would begin to move us away from the pathologies that disfigure our system.

We are not suggesting that either party should slavishly follow public opinion. The parties exist for a reason, their core commitments matter, and sound policy does not always emanate from the center. At the same time, it is good neither for the parties nor for the country that their core supporters are so disinclined to accommodate the views of moderates.

Too many Americans are losing confidence in the future. Restoring that confidence will take more than optimistic speeches and a gradual decline in unemployment. It will require, as well, the kinds of systemic changes that can begin to move us from the politics of point-scoring to a new politics of problem-solving.

This paper is intended as a contribution to that new politics. While we are Democrats, we do not present our proposals in a narrowly partisan spirit. Reforms that offer moderates political voice commensurate with their numbers—a situation that does not now exist would be good not only for our party, but also our democracy and our country.

Appendix #1

Americans’ views on the direction of the country

Average number of Americans saying the country was headed in the right direction by year.27

“All in all, do you think things in the nation are generally headed in the right direction or do you feel that things are off on a wrong track?”

























Appendix #2

Demographics of moderate voters

In the following paragraphs, the figures for moderates are followed by the corresponding figures for the electorate as a whole in parentheses.28

Gender: 48% of moderates are male (48%), while 52% are female (52%). Among groups to the right (conservatives and very conservative), the split is 51%/49%; to the left (liberal and very liberal), it’s 45%/55%.

Age: Moderates are somewhat younger than the total electorate. 26% are 18 to 29 (22%), while only 13% are 65 or older (16%). By contrast, 15% of conservatives/very conservatives are 18 to 29, while 20% are 65+. The age profile of moderates more closely resembles the cohorts to their left than to their right.

Education: In terms of educational attainment, moderates break down as follows: High school or less 32% (34%); some college 26% (24%); 4-year college graduate 18% (17%); post-graduate training 12% (11%). Compared with conservative and very conservative respondents, moderates are significantly more likely to have a B.A. or more, and significantly less likely to have gone no farther than high school. (Differences in educational attainment between moderates and those to their left are not statistically significant.)

Race/Ethnicity: The racial makeup of moderates is as follows: white 71% (73%); African-American 15% (14%); Asian-American 2% (2%); other 8% (8%). Only 10% of moderates are of Hispanic origin, versus 13% of the total.

Marital Status: As for marital status, moderates break down as follows: married 51% (51%); divorced or separated 14% (13%); never-married 25% (22%). In general, ideology and marital status are closely linked: the more conservative, the more likely to be married. More than 60% of conservatives/very conservatives are married, while only 42% of liberals/very liberals are married.

Religion: 47% of moderates are Protestant (49%), 24% are Catholic (24%), and 17% are unaffiliated. Only 37% of Protestant moderates describe themselves as born-again or evangelical, versus 48% for conservatives and 57% for the very conservative cohort. 36% of moderates attend religious services at least once a week (37%), while 28% seldom or never do so (28%). Although a smaller share of liberal/very liberals than moderates attend regularly, here again the sharper contrast is between moderates and conservatives (46%) and especially those who are very conservative (69%).

Income: There are no significant differences between moderates and the electorate as a whole. Indeed, there is virtually no correlation between income and ideology. For example, 13% of moderates report family income in excess of $100,000, as do liberals/very liberals and conservatives/very conservatives. In the middle of the income spectrum, 14% of moderates report family income between $50,000 and $75,000; for right-of-center groups, the corresponding figure is also 14%; for left-of-center groups, 12%. Even at the bottom, differences are modest: 29% of moderates report family incomes of $30,000 or less, but so do 30% of conservative/very conservative identifiers and 32% of liberals/very liberals.

Occupation/Economic status: Moderates are less likely to describe themselves as professionals or business people (27%) than are those to their left (32%). Only 19% of moderates say they are self-employed or operate small businesses, the lowest percentage of any ideological group, and half the share of the very conservative cohort (38%). Moderates are 10 percentage points more likely than the far right, and 8 points less likely than the far left, to see the United States as divided into haves and have-nots. Surprisingly, only 45% of moderates see themselves as “haves,” the lowest of any ideological group and 10 points less than liberals.

Appendix #3

How Moderates and Independents Differ29

As political scientists have long argued, the majority of Independents in fact lean toward one party or the other, and these “leaners” tend to think and act much like explicit partisans do. For example, the 2010 Pew survey found that of the total 37% pool of Independents, 17% leaned Republican and 13% Democratic, leaving only 7% as pure Independents.

This survey is hardly an outlier. The others we examined showed much the same thing. Overall, three quarters of Independents turn out to be leaners (40% of them toward the Republicans, 35% toward the Democrats) whose views and behavior resemble those of the party to which they lean (and from which many of them have recently come). 79% of former Republicans lean Republican; 76% of ex-Democrats lean Democratic. That is not to say that leaners are identical to the parties they have left. Former Republicans are more critical of George W. Bush and the Iraq war and less hard-edged on social issues such as gay marriage. For their part, former Democrats are more critical of Obama’s performance on health care reform than are Democrats. They trust government less and are more sympathetic to a smaller government offering fewer services than are their erstwhile colleagues. To a first approximation, nonetheless, partisans and leaners can be seen as a single group.

Since Obama’s election, there is evidence of a conservative shift among Independents. In 2008, 46% of Independents regarded themselves as moderate, 30% as conservative, and 19% liberal. By 2010, the moderate share had declined by five points, to 41%, while conservatives rose from 30% to 36% and liberals held steady.

Over the past decade, Republican leaners have become much more conservative and less moderate than they were a decade ago: as a share of Republican-leaning Independents, conservatives have increased from 44% to 57%, while moderates have declined from 43% to 34%. By contrast, the ideological composition of Democratic-leaning Independents has barely budged and stands at 52% moderate, 29% liberal, and 15% conservative. So the political center of gravity among Independent leaners has shifted noticeably to the right.

The ideological composition of non-leaning Independents, which constitute 25% of the Independent pool and roughly 10% of the electorate, hasn’t changed much in the past decade. But these Independents, too, are much more favorable to Republicans—and less to Democrats—than are moderates. 43% of non-leaners identify as conservatives, versus only 13% as liberal. In short, non-leaners are much closer to Republican leaners than to Democratic leaners, reinforcing the conservative tilt of Independents as a group.

The non-leaners are not a powerful political force, however. Compared to moderates (and to most other groups), they are very downscale in both income and education, are less likely to be registered or to vote, even if registered, pay less attention to politics, and tend to be disconnected from even the largest policy disputes. They are disaffected from both political parties and never thought of themselves as members of either. They trust neither government nor their fellow citizens. (Not surprisingly, they are very anti-immigrant as well.) And they are deeply pessimistic, not only about America’s ability to solve its problems collectively, but also about individuals’ abilities to solve problems on their own.

On closer inspection, then, Independents tend to dissolve into two groups: those who are hard to distinguish from partisans, and those who are so disconnected and disaffected as to have virtually no impact on political outcomes. As we will show, the same cannot be said of moderates, whose outlook and attitudes on a wide range of issues distinguish them from both liberals and conservatives.

Notwithstanding these core realities, many journalists and pundits persist in lumping moderates and Independents together. In this context, it is useful to point out some significant differences between them. Consider attitudes toward the two political parties: while moderates and Independents have statistically indistinguishable evaluations of the Republican Party (37% approval by moderates, 36% by Independents), moderates are far more favorable toward the Democratic Party (55%) than are Independents (36%). When asked to rank-order the parties, moderates systematically rate Democrats higher than Republicans, while Independents do the reverse, as revealed in a number of Gallup questions.

Table A Which political party do you think can do a better job of  handling the problem you think is most important

Table B Which political party do you think will do a better job  of keeping the country prosperous

Table C Which political party do you think will do a better job of protecting the country from international terrorism and military threats

Given these attitudes, it is hardly surprising that moderates approve of the way Barack Obama is handling his job as president by a margin of 54% to 42%, while Independents disapprove by a similar margin, 54% to 40%.

Moderates and Independents also differ in their stance toward government and its major institutions. While neither group is satisfied with the way the nation is being governed, moderates are less dissatisfied (34% satisfied, 65% not) than are Independents (24% satisfied, 75% not). Majorities of both groups lack trust in the federal government to do what is right, but moderates are more likely to express trust (26%) than are Independents (18%), a gap that spills over into specific policy areas and institutions. By a margin of 13 percentage points, moderates are more likely to express trust and confidence in the federal government’s handling of international problems; the margin is a smaller but still significant 7 points when it comes to domestic problems. Similar trust gaps exist for the executive branch, the legislative branch, and the mass media.

Given the attitudinal differences described above, it is not surprising to find that, in contrast to Independents, moderates tend to look and to behave more like Democrats, a fact that is often overlooked. For instance, only 17% of moderates identify as Republicans, 33% as Democrats, and 47% as Independents. When Independents who lean in one direction or another are factored in, 56% of moderates are Democrats or lean in that direction, versus 31% Republican. 65% of moderate Independents report than they have thought of themselves as Democrats during the past five years, while 46% report having thought of themselves as Republicans. (Remarkably, 11% of moderate Independents have thought of themselves as both at one time or another.)

Our focus on moderates is not meant to suggest that Independents are an insignificant electoral force. On the contrary, shifts among Independents have also shaped recent elections, presidential as well as midterm.30

Table C Which political party do you think will do a better job of protecting the country from international terrorism and military threats

Although Democrats’ share of moderate voters rose by twice their increase among Independents (6% points versus 3 points), the Independent contribution to Democratic gains between 2004 and 2008 is about equal to that of moderate voters.

Now let’s look at the states that Obama won and Kerry lost.31

Table C Which political party do you think will do a better job of protecting the country from international terrorism and military threats

In many of these states, Obama’s share of the moderate vote soared above Kerry’s—by 9 points in Colorado and Nevada, 10 points in Indiana, and a stunning 13 points in North Carolina. (Indiana and North Carolina may have been the most surprising Democratic victories in 2008, and Obama’s performance among moderates was the principal reason why.) In addition, two critical Democratic victories—in Florida and Ohio—occurred despite the fact that Obama’s share of the Independent vote actually declined relative to Kerry’s. In both these cases, Republicans’ share of the electorate declined significantly—from 41 to 34% in Florida and from 40 to 31% in Ohio.

It appears that many disaffected Republicans relabeled themselves as Independents but could not bring themselves to vote for Obama. Still, in both these states, Independents who supported Obama constituted a larger share of the electorate than had Independents for Kerry four years earlier.

The victory threshold for Democratic presidential candidates among Independents is significantly lower than for moderates. Clinton prevailed in 1992 and 1996, as did Obama in 2008, with 54% of the two-party Independent vote. Gore received only 49%, Kerry 50%.

As Table F illustrates, for the past 40 years, moderates have been more likely to vote for Democrats for president than have Independents, by an average of 8.2 percentage points. In fact, in the last 9 presidential elections Independents delivered more than half their votes to the Democrat in only one election—the 2008 landslide that elected Obama. In contrast, moderate voters have delivered more than half their votes to the Democratic candidate in six out of nine of the past presidential elections.32

Table F Percentage of Moderates and Independents who voted Democratic

Appendix #4

Presidential Primaries across the United States33

Presidential Primaries across the United States

Appendix #5

Dates for 2010 Primaries

Dates for 2010 Primaries

About the Authors

William A. Galston is Ezra Zilkha Chair and Senior Fellow in Governance Studies at the Brookings Institution. A former policy advisor to President Clinton and presidential candidates, Mr. Galston is an expert on domestic policy, political campaigns, and elections. His current research focuses on designing a new social contract and the implications of political polarization.

Elaine C. Kamarck is a Lecturer in Public Policy in the Kennedy School of Government at Harvard University. In the 1980s, she was one of the founders of the New Democrat movement that helped elect Bill Clinton president. Ms. Kamarck served in the White House from 1993 to 1997, where she created and managed the Clinton Administration’s National Performance Review, also known as reinventing government. She is also the author of Primary Politics: How Presidential Candidates have Shaped the Modern Nominating System (Washington, Brookings, 2009).

  • American Electorate224
  • Public Opinion201


  1. “GOP Image Net-Positive for First Time Since 2005, Democratic Party’s favorable Rating is Improved from last year,” Gallup, January 27, 2011, Available at: See also: “President Obama Job Approval,” Real Clear Politics, Available at:

  2. Jeffrey M. Jones, “Americans Divided on Obama, Republican Spending Proposals,” Gallup, January 28, 2011. Available at:

  3. “Independents Oppose Party in Power . . . Again,” Pew Research Center, September 23, 2010. Available at:

  4. CNN National Exit Polls, 1980 – 2008. Available at:

  5. Lydia Saad, “Conservatives Continue to Outnumber Moderates in 2010,” Gallup,  December 16, 2010. Available at:

  6. Authors’ calculation based on CNN National Exit Polls. Available at:

  7. Authors’ calculation based on CNN National Exit Polls. Available at:

  8. Authors’ calculation based on CNN National Exit Polls. Available at:

  9. Unless otherwise noted, all public opinion data cited in this paper are drawn from four major surveys:  (1) “Trends in Political Values and Core Attitudes: 1987-2009,” Pew Research Center, May 2009. Available at:;  (2) “Independents Oppose Party in Power . . . Again,” Pew Research Center, September 23, 2010. Available at:; (3) Gallup Poll Social Series: Governance, October 2010; (4) Robert P. Jones, Ph.D. and Daniel Cox, “Old Alignments, Emerging Fault Lines: Religion in the 2010 Election and Beyond,” Public Religion Research Institute, November 2010. Available at:  The authors are grateful to these organizations for their willingness to provide the detailed cross-tabulations on which our analysis is based.

  10. CNN/Opinion Research Corporation, December 17-19, 2010, Q2.

  11. Frank Newport, “Americans’ Church Attendance Inches Up in 2010,“ Gallup, June 25, 2010. Available at:

  12. Lydia Saad, “Conservatives Continue to Outnumber Moderates in 2010,” Gallup,  December 16, 2010. Available at:

  13. Curtis Gans, “The Primary Turnout Story: Presidential Races Miss Record High; Senate and Governor Contests Hit Record Low,” October 1, 2008. See summary charts:

  14. Curtis Gans. See also source for general election turnout: “National Voter Turnout in Federal Elections,” Available at:

  15. David W. Brady, Hahrie Han and Jeremy C. Pope, “Primary Elections and Candidate Ideology: Out of Step with the Primary Electorate” Legislative Studies Quarterly, XXXII, February 1. 2007, p. 80, Print.

  16. David W. Brady, Hahrie Han and Jeremy C. Pope.

  17. Many thanks to Professor Hahrie Han of Wellesley College for data from 2002 and 2006. Many thanks to Tess Stovall and Danielle Wilsey at Third Way for collecting and calculating the data for 2010.  General election data for 2002, 2006, and 2010 from Michael McDonald, “United States Elections Project,” accessed on February 7, 2011. Available at: According to the United States Elections Project, the VAP in 2002 was 215,825,547, the VAP in 2006 was 225,964,346, and the VAP in 2010 was 235,809,266. 

  18. Tess Stovall and Danielle Wilsey at Third Way collected and calculated data for 2010. 

  19. Tess Stovall and Danielle Wilsey at Third Way collected and calculated data for 2010. 

  20. Brady, p. 90.

  21. Norman J. Ornstein, Thomas E. Mann, and Michael J. Malbin, Vital Statistics on Congress 2008, Table 2-12 and William A. Galston tabulation from “Can A Polarized American Party System Be Healthy?” Issues in Governance #34, Brookings, April 2010.

  22. Bill Bishop and Robert Cushing, The Big Sort, Thomas Mann from Red State, Blue State.

  23. Douglas Johnson, Elise Lampe, Justin Levitt and Andrew Lee, “Restoring the Competitive Edge,” The Rose Institute of State and Local Government, Claremont-McKenna College, September 26, 2005, p. 10, Print.

  24. Douglas Johnson, Elise Lampe, Justin Levitt and Andrew Lee, “Restoring the Competitive Edge,” The Rose Institute of State and Local Government, Claremont-McKenna College, September 26, 2005, p. 10, Print.

  25. David W. Brady and Hahrie Han, “Polarization Then and Now: A Historical Perspective,” Red State and Blue Nation?, Volume I: Characteristics and Causes of America’s Polarized Politics, The Brookings Institution, p. 141, Print.

  26. For further analysis of these and related proposals see “Toward Depolarization” Pietro S. Nivola and William A. Galston in Red and Blue Nation?, Volume 2, The Brookings Insititution, 2008, Print.

  27. Source: Authors’ calculations of right track/wrong track polling for the past decade. Yearly numbers are presented in full at:

  28. Unless otherwise noted, all public opinion data cited in this paper are drawn from four major surveys:  (1) “Trends in Political Values and Core Attitudes: 1987-2009,” Pew Research Center, May 2009. Available at:;  (2) “Independents Oppose Party in Power . . . Again,” Pew Research Center, September 23, 2010. Available at:; (3) Gallup Poll Social Series: Governance, October 2010; (4) Robert P. Jones, Ph.D. and Daniel Cox, “Old Alignments, Emerging Fault Lines: Religion in the 2010 Election and Beyond,” Public Religion Research Institute, November 2010. Available at:  The authors are grateful to these organizations for their willingness to provide the detailed cross-tabulations on which our analysis is based.

  29. Unless otherwise noted, all public opinion data cited in this paper are drawn from four major surveys:  (1) “Trends in Political Values and Core Attitudes: 1987-2009,” Pew Research Center, May 2009. Available at:;  (2) “Independents Oppose Party in Power . . . Again,” Pew Research Center, September 23, 2010. Available at:; (3) Gallup Poll Social Series: Governance, October 2010; (4) Robert P. Jones, Ph.D. and Daniel Cox, “Old Alignments, Emerging Fault Lines: Religion in the 2010 Election and Beyond,” Public Religion Research Institute, November 2010. Available at:  The authors are grateful to these organizations for their willingness to provide the detailed cross-tabulations on which our analysis is based.

  30. CNN National Exit Polls, Available at:

  31. CNN National Exit Polls, Available at:

  32. CNN National Exit Polls, Available at:

  33. “Primary Election Systems,” National Conference of State LegislaturesJune 2010. Available at:


Get updates whenever new content is added. We’ll never share your email with anyone.