Twigs Bent Left or Right

Understanding how liberals and conservatives differ, from conception on

green and yellow image

Illustration by Stuart Bradford

How did Franklin Delano Roosevelt ’04, born in 1882 to a privileged, aristocratic life in New York’s Hudson River Valley, become a liberal reformer? Historians have proposed several possibilities. It may have been the example of his father, who stood alone as one of the only Democrats in the Roosevelt family at the time. Perhaps it was the influence of his headmaster at Groton, who preached the gospel of social responsibility. Some say it was his struggle with polio, which gave him knowledge of suffering. Arthur M. Schlesinger Jr. ’38, Jf ’43, believes that one of the most powerful forces was FDR’s admiration for his larger-than-life fifth cousin, Theodore Roosevelt, A.B. 1880. “FDR was a Democrat, whereas Theodore Roosevelt was a Republican, but TR was a progressive Republican,” Schlesinger says. “He believed in government, he believed in public action to open up careers and to expand opportunities for the not-so-well-off. I think FDR imbibed that from TR.”

But the forces at work were unpredictable. “TR’s own sons, for example, did not assume the progressive Republicanism of their father,” Schlesinger says. What caused them to adopt different politics? After a pause, Schlesinger proposes an answer: “A mysterious chemistry, if you will.”

At Harvard and elsewhere, researchers in political science, sociology, psychology, and even genetics are attempting to assay this mysterious chemistry. These are particularly important questions right now. “When Goldwater ran in 1964, he said there wasn’t ‘a dime’s worth of difference between the two big parties,’” says Radcliffe Institute fellow and associate professor of government J. Russell Muirhead, who is working on a forthcoming book, Left and Right: A Defense of Party Spirit. “It used to be that most major legislation was passed by whopping bipartisan majorities, and lawmaking was characterized, in general, by a certain measure of bipartisanship. But beginning in 1980, and intensifying over the past five or 10 years, the parties as represented in Congress stand for very different things. The consequences of electing one party instead of the other are much greater than they used to be. Most people would agree there’s now at least a dime’s worth of difference between the two parties.”


Blue Genes or Red?

This polarization of the parties raises the stakes for the electorate, since it decreases the likelihood of moderate or centrist governance. Classically, the liberal-conservative division concerned the distribution of resources and power: conservatives sought to “conserve” or maintain the status quo, while liberals aimed to reform and redistribute. But other intersecting issues now complicate those assumptions. Religious conservatives, for example, highlight moral issues like abortion, euthanasia, and gay marriage, while having relatively little to say about broader economic and foreign-policy questions.

As the two parties have shifted, the electorate, too, is changing. New voters, including immigrants from Mexico, Central and South America, and Southeast Asia have created nonwhite majorities in several states, including Texas and California. Early in the twentieth century, immigrants—as economically disadvantaged groups—typically leaned leftward. But the political views of today’s arrivals are no longer so predictable. Many new Americans come from traditional societies and identify with conservative family-values positions. Eastern European escapees from communism harbor a deep suspicion of government power, as do Cubans in south Florida, and that can crystallize into a more general conservatism.

Recent immigrants are not the only voters defying conventional wisdom. “For a long time the parties were stable coalitions, and you knew someone’s politics if you knew something about them,” says Pforzheimer University Professor Sidney Verba, an expert on political ideologies and political participation. “The idea of a working-class Catholic from Boston being a Republican was crazy. Now, of course, Catholics in the United States no longer are overwhelmingly connected with the Democratic Party. There’s been a lot of shaking up of traditional party coalitions.” A Roman Catholic who opposes abortion, for example, might move rightward because of this single issue.

Individual political convictions aren’t as clear as we like to think, either. We may call ourselves “conservative” or “liberal,” but these standard categories leave a lot of room for individual ambivalence. We are not, like states on a television network’s electoral map, simply red or blue. “The reality is that people are much more purple than anything else,” says D. Sunshine Hillygus, assistant professor of government and a fellow at the Shorenstein Center for Press, Politics and Public Policy at the Kennedy School of Government (KSG). “There has almost always been about a third of Republicans who are pro-choice, and a third of Democrats who are pro-life. There’s also about a third of Democrats who are opposed to affirmative action, and more than 40 percent of Republicans who support gun control. When you look at individual issues, you don’t find nearly the polarization suggested by the media, or, frankly, by scholars.”

Still, given the structure of our electoral system, we can punch only one chad at a time on Election Day. And that chad may matter more than ever. Partisan redistricting of Congressional seats, changes in campaign finance, and other trends that put leverage in the hands of incumbents suggest that even those elected by razor-thin margins can stay in power for years, exercising profound, long-lasting effects on governance.

Mom and Pop Politics

At least one element of the “mysterious chemistry” is certain. Most scholars accept that our politics are inherited from our families, absorbed during the dinner-table conversations of childhood. Political affiliation does not often change dramatically from one generation to the next. “Children listen to their parents’ interpretation of news events. They listen to who their parents’ presidential heroes are,” says professor of health policy and management Robert Blendon, director of the Harvard Program on Public Opinion and Health and Social Policy at the Harvard School of Public Health and professor at the Kennedy School, who conducts frequent polls on domestic issues such as healthcare reform. “People who loved President Reagan sat at home and said, ‘He is the most wonderful person in the world!’ But up here in the blue states, I don’t think many children went to bed thanking President Reagan for all he had done.”

Indeed, the stereotype of rebellious teens who reject their parents’ politics is mostly false. One study, coauthored by Richard Niemi of the University of Rochester, tracked the politics of children and their parents over time. The parent-child pairs were first interviewed in 1965—when the kids were seniors in high school—and asked about their opinions on issues like the racial integration of schools and the use of prayer in schools. The data revealed that 60 to 70 percent of the teenagers agreed with their parents on these issues and nearly 70 percent affiliated themselves with the same political party. Only about 10 percent belonged to the opposing political party. During the next 20 years, as the offspring matured, they drifted away from their parents’ views on a number of issues, and the rate of agreement on a political party dropped from 70 percent in 1965 to 61 in 1973 and 1982.

Another study suggests that the rate of agreement on party affiliation dropped even further in the 1990s, to 57 percent. That may be due, in part, to the changing nature of dinnertime conversations about politics. David King, associate director of the KSG’s Institute of Politics (IOP) and lecturer in public policy, blames the Watergate scandal for causing political identification and interest in government to wane. “The Baby Boomers haven’t voted in nearly the numbers that the World War II generation did,” he says. “So their children didn’t witness their parents voting on a regular basis.” In the same post-Watergate period, many schools struck civics courses from their curricula. “You have to think about who was teaching social studies in the mid 1970s,” King says. “They were, for the most part, liberals who did not serve in Vietnam, who were unhappy and mistrustful because of Watergate, and who thought that in teaching about American politics they would be engaging in propaganda. They didn’t want their students exposed to this quote-unquote ‘propaganda.’” Civics courses have stayed off the curricula because of the pressures of statewide proficiency testing. Teachers are spending more class time preparing students for these tests, King says, “and we don’t have civics tests in most jurisdictions.” This means young people receive less political education from schools and parents, leaving their peers to play an important role.

Still, Niemi says that his study illustrates an important point, that we never really shake off our parents’ convictions. He offers the example of children of strong Democrats who, when asked about their opinions on specific issues, revealed themselves to be quite conservative. Yet on average, those people were unlikely to think of themselves as Republicans. Instead, Niemi says, they were unable to dodge the gravitational force of their parents’ politics and self-identified as independents who merely leaned to the right. “The real point we were making in this paper is that parental partisanship has some pull on offspring, even years after the offspring have flown the coop.”


The Crash, Selma, and 9/11

If our family traditions are the primary influence on our politics, then certain significant world events play a close second. Consider, for example, the Great Depression, which swung the nation left, and the Civil Rights and Voting Rights Acts of 1964 and 1965, which drastically altered life in the South. As the Democratic Party embraced the cause of civil rights, many white Democrats switched parties, Robert Blendon says. “As the civil-rights laws integrated the South, and pushed busing, affirmative action, and voting rights, whites moved to the suburbs, put their kids in church schools, and voted Republican,” he explains. “It was a transformation. We only have to look at a range of politicians. Trent Lott was a Democrat, Phil Gramm was a Democrat, Strom Thurmond was a Democrat. They became leading conservative figures in the Republican Party.” In time, 14 southern states changed from Democratic bastions to Republican strongholds.

Thinking about such major shifts, one wonders if more recent events, such as 9/11, or the war in Iraq, may prompt similar sea changes. Some scholars say it’s too soon to tell, while others, including Sidney Verba, doubt that these contemporary events have the power to alter our politics. He theorizes that frenetic media coverage prevents us from dwelling on an event long enough for it to make an impact. “What happens now is that you get a more rapid fluctuation in the attention span of the public,” he says. “No one remembers Rwanda anymore, or Somalia, or Bosnia, or Kosovo, though at one point they were on the front page.”

For an event to have a long-term impact on the party system, it must almost permanently alter the shape and views of the electorate, says J. Russell Muirhead. “If an event impoverishes large numbers of the electorate, like the Great Depression, or slaughters large numbers of the electorate, like the Civil War, or introduces lots of new people into the electorate by enfranchising people who had previously been disenfranchised, like the Voting Rights Act, then the event has a lasting impact,” he explains. “Events that affect our moods or our passions may have consequences for policies, but those consequences last only as long as the mood. September 11 affected our mood, but it didn’t change the demographic character of the country. The policies passed on account of 9/11—like invading Iraq—will have long-term impacts on our politics, to be sure,” Muirhead continues. “But I don’t think 9/11 will generate a realignment of the parties, because it hasn’t altered the basic opinions or demographic makeup of the electorate.”


The Magic of Demographics

Decades of research have generated many theories about how life circumstances influence political ideology. There are the issues of income (the higher your income, the more conservative you’re likely to be); education (an advanced degree makes a person at least a little liberal); region (residents of Utah are generally more conservative than residents of Massachusetts); race (minorities generally are more likely to lean left); and religion (regular churchgoers are more likely to lean right). Yet the variables interact in complex ways, and scholars do not agree about which demographic factors matter most. It is difficult to predict with certainty how a person with a Ph.D. who earns $100,000 a year and lives in Pennsylvania will vote on Election Day.

But two anomalous cases are instructive. The first case is Jewish voters, who make up just 3 percent of the voting population. “The Jewish voter tends to be among the most liberal of white groups,” says professor of education and social structure emeritus Nathan Glazer, a sociologist who writes about race, ethnicity, and the Jewish community in the United States. “Jewish voters take a liberal position, which is seen as less protective of income and property, despite the fact that they are a higher-income group.” Why do they vote against their apparent economic best interests? “The self-perception and historical experience as a threatened minority keep them on the liberal side, particularly because liberalism is so much oriented now to the defense of minorities,” Glazer says. “There is a suspicion that the American conservative is just not going to be sympathetic to minority concerns.” And so for Jews, he says, a desire for protection as a group trumps economic interest.

There’s also the case of Kansas, described by political journalist and Kansas native Thomas Frank in his 2004 bestseller, What’s the Matter with Kansas? Frank examines why poor Kansans living in decaying rural towns identify so strongly with conservative views and vote overwhelmingly for Republicans whose party platform promises to dismantle social programs to aid the poor. Frank maintains that they vote not according to their economic best interests, but according to their positions on cultural ideals and values issues like abortion, gun rights, and same-sex unions. Glazer agrees. “I think it’s the value issues, the notion of defending a traditional society and traditional values, which explain this kind of orientation,” he says. Although we might expect economic needs to drive a person’s voting behavior, certain dearly held values—protection for minorities or the preservation of traditional families—appear to win out instead.

Indeed, one body of political-science research suggests that the state of someone’s pocketbook actually matters little in voting. “Researchers have found that even though a lot of people vote on the basis of the economy, they’re generally thinking of the economy of the country, rather than their own economic situation,” says Hillygus of the government department. “They might think, ‘If the economy of the country improves, then that’s going to have a trickle-down effect and improve my own economic situation.’ But to say, ‘I earn $200,000 dollars, and if I elect the Republicans, I might get a tax cut’?—that doesn’t appear to be part of the decision-making process.”

That’s not to say that income level doesn’t influence voting. Exit polls from the 2004 presidential election showed that voters with incomes of $100,000 or more chose George W. Bush over John Kerry, 60 percent to 40 percent. (Compare this with voters with incomes below $30,000, who chose Kerry over Bush by the same margin.) But professor of government Eric Schickler, coauthor of Partisan Hearts and Minds, agrees that voters aren’t necessarily on a self-interested quest for beneficial tax policies. They may be motivated more by social conformity and their membership in an identity group. “Traditionally, the Republicans were the party of business,” Schickler says. “So if you identify yourself as, say, a small-business person, you’re more likely to see that identity as consistent with your own. ”

“Altared” Allegiances

Of all the demographic factors in play, religion appears to have made the biggest difference in recent decades. That wasn’t always the case. “In 1954, if you were a regular churchgoer, holding everything else constant, you were equally likely to be Democrat or Republican,” says David King of the Kennedy School. “Now, if you’re a regular churchgoer, you’re overwhelmingly more likely to be a Republican.” A paper coauthored by KSG lecturer Elaine Kamarck, which analyzed statistics from the University of Michigan’s American National Election Survey, shows just how wide the gap between religious and secular voters now yawns. In 1988, Democratic candidate Michael Dukakis “fared only two points worse among regular churchgoers than among those who attended church infrequently or not at all,” a difference typical of presidential elections since 1952, the paper states. But since the presidential election of 1992, the difference between these two groups widened dramatically, to nearly 12 points on average.

Ronald Thiemann, professor of theology at the Divinity School, says people of faith may be drawn to conservative politics because Republicans have been better at articulating religious issues. That wasn’t the case during the 1960s, he says, when liberal Jews and Christians stood united for civil rights and against the Vietnam War. Yet in the 1980s, when Republicans—and particularly the religious right—mobilized around the issue of abortion, religious liberals lost their voice. “I don’t think that people on the more [religiously] liberal side quite knew how to think about these new emerging realities,” Thiemann says. As the Bible Belt evolved into a Republican stronghold after the 1960s, the Republican Party came to be seen as the party of religious voters, especially evangelical Protestants.

But among college students at least, religious belief does not predispose one to conservative politics. Twice a year, David King and his colleague John Della Volpe, an opinion researcher and senior adviser to the IOP, conduct a survey of college students’ civic and political engagement. In the spring of 2005, they presented respondents with 11 statements—like “The best way to increase economic growth and create jobs is to cut taxes,” “Basic health insurance is a right for all people,” and “Homosexual relationships between consenting adults are morally wrong”—and asked them the extent to which they agreed or disagreed. Researchers analyzed the data to find groups of people who held the same constellations of beliefs. Two of the groups were familiar. Traditional liberals (43 percent of the student sample) were against the war in Iraq, believed health insurance is a right, and supported gay marriage. Traditional conservatives (14 percent) supported the Iraq war, opposed universal health insurance, and strongly opposed gay marriage.

Smack in the middle was a group of moderate students (21 percent) for whom faith was important. More than a third were African American or Hispanic, Della Volpe says, representing the largest cluster of minorities in the survey. Many came from immigrant families. “They’re very concerned about the moral direction of the country, ” he says. “They believe that religion should play a more important role in government, and they believe in school choice. But they also believe that healthcare is a right if it can’t be afforded, and they’re very, very strong advocates for the environment.” King theorizes that these opinions reflect the concern among today’s college students—who are volunteering in record numbers—for the health and well-being of the greater community. “Young people,” he quips, “have gone from the ‘me’ generation to the ‘we’ generation.” In the last presidential election, these religious centrists were equally divided between Kerry and Bush. They are not predominantly liberal or conservative, and their votes are up for grabs.

Psychological Inclinations

In addition to the influences of family, religion, and demographics, the mysterious chemistry partakes of the force of personality. The Harvard Study of Adult Development (originally known as the Grant Study) is a continuing project that began with 268 men who were Harvard sophomores between 1940 and 1942. The study was conceived in 1937 to identify factors leading to mental and physical health. It has also yielded interesting information about political preferences.

Few of the study subjects, for instance, were moderates; most were either solidly liberal (34 percent) or solidly conservative (37 percent). And their political ideologies were remarkably durable. “The interesting thing about these men is that over time, their politics didn’t change,” says professor of psychiatry George Vaillant, lead researcher of the study. “The Republicans at 25 were still Republicans at 85, and the same was true for the Democrats.”

In 1944, a psychiatrist evaluated the men and assigned them characteristics from a group of more than 25 possible traits. Those who identified themselves as Republicans “are more likely to be practical-organizing and pragmatic. They are ‘Show me, don’t tell me,’” Vaillant explains. “The Democrats are more likely to be cultural, verbalistic, shy, and to have a sensitive affect, or to be ‘thin-skinned.’” Aside from these traits, there’s little to distinguish the two groups. They were equally likely to have happy childhoods and to experience alcoholism, mental illness, and divorce. They were also equally likely to exhibit altruism, which the researchers defined as the ability to use personal difficulties to benefit others, as in the case of a childhood polio sufferer who went on to become a pediatrician to help disabled children.

They did differ in certain ways. The Democrats were more likely to have highly educated mothers. Republicans tended to make more money, and to be less open to new ideas. It’s also worth noting that the Republicans were more likely to be athletes, Vaillant says. He theorizes that the propensity for sports arose because the Republicans “were men of action, not reflection.” But in the end, these differences didn’t matter much; the conservatives and liberals aged equally well. These similarities surprised Vaillant, a lifelong Democrat who says he has never voted for a Republican presidential candidate. “I certainly had all kinds of prejudices, and doing the study got me to change them,” he says. “I thought that the Democrats would be a whole lot nicer and more altruistic, and that wasn’t the case at all.”

Some argue that what makes us conservative or liberal is a particular way of seeing the world. In his book Moral Politics: How Liberals and Conservatives Think, cognitive scientist George Lakoff of the University of California, Berkeley, argues that political choices come down to personal visions of the ideal family. Conservatives, he says, prefer the “strict father” model, while liberals prefer the “nurturant parent.”

To understand these two models, “Think of George ‘Bring ’Em On’ Bush and Bill ‘I Feel Your Pain’ Clinton,” says Muirhead. Bush, he says, embodies the “strict father” model with two qualities in particular: “an ability to pick out what’s evil, and a willingness to respond to it decisively, even ferociously. The strict father doesn’t waffle on what’s good and bad, and doesn’t question whether something he thought was bad might in fact be good.” In this paradigm, he explains, those drawn to the strict-father image prefer conservative leaders who believe that “we cannot be indulgent with criminals, or squeamish about using the military.”

The liberal, nurturant-parent model comes alive in Clinton’s famous ability to connect with people. “The nurturant parent is empathetic,” Muirhead says. “He identifies with you, sees things from your point of view.” This model corresponds to a vision of a world that might be improved. It values opportunities for cooperation, and seeks to build national and international institutions that “facilitate trade, diplomacy, and security,” says Muirhead.

It’s easier to trace the changes that altered America’s politics than to predict how events taking place now will unfold. Consider obviously significant happenings, like the influx of immigrants into Texas, California, and New Mexico that has created the first nonwhite majority states. “Changes like this must have some kind of impact,” Muirhead says. “But exactly what kind is simply impossible to say in advance. It will depend on how the economy treats recent immigrants, on the political salience of social issues, and on how immigration politics plays out. What we can say is that both parties will strain themselves to turn changes like these to their advantage.” Scholars also predict that the intense dialogue about faith, started during the last presidential campaign, will continue. Since the election, there has been much discussion about which party will be best able to attract religious voters, particularly those who don’t fit neatly into the other Republican or Democrat categories.

But no matter how political parties themselves evolve, they inevitably embody a form of groupthink that is removed from the lived experience of individuals. Muirhead rejects the idea that any person fits solely into a category of liberal or conservative. “Most people don’t go for ideological consistency,” he says. “They don’t walk around interpreting everything they see in terms of some neat ideological package.” Instead, we collect the ideas that make sense to us. “These packages of ‘liberal’ and ‘conservative’ each get something right about the world,” Muirhead explains. “Most people who live in the world understand that it’s very messy, and that good judgment is not based in ideological consistency. Good judgment is based in an ability to be conservative in some places, and liberal in others.”


Erin O’Donnell is a freelance writer in Arlington, Massachusetts.

Read more articles by: Erin O'Donnell

You might also like

Steven Pinker on Apple’s Vision Pro

Professor of psychology on the science and history behind the Vision Pro.

The State of Black America

Harvard African American scholars take stock of a difficult moment. 

Threats Foreign and Domestic

Joseph Nye discusses geopolitics and Harvard’s challenges.

Most popular

We Can Do Better

A human-rights expert on Central American child immigrants

Let the Sonnets Be Unbroken

Neil Rudenstine guides readers through Shakespeare’s lyric masterpiece.

Martha Tedeschi to Lead Harvard Art Museums

Hailing from the Art Institute of Chicago, Tedeschi will assume the role next July. 

More to explore

Photograph of Winthrop Bell 1910

Winthrop Bell

Brief life of a philosopher and spy: 1884-1965

Illustration of people talking to each other with colorful thought bubbles above their heads

Talking about Talking

Fostering healthy disagreement

Vacationing with a Purpose

New England “summer camps” for adults