The various turns of the Republican presidential primary have made it clear that there is a division in the GOP unlike anything since 1996.  There is, however, no consensus over what defines the sides of this split.

The media, largely left-of-center itself and with no deep understanding of conservatism, imagines a debate between "Romney" and a multiple-personality "Not Romney."  They also characterize the gulf as between a notional Establishment and the base (or Tea Party), or simply between those less or more "conservative."  (This last is the position promoted for obvious political reasons by Sen. Rick Santorum and Speaker Newt Gingrich.)

Yet general Tea Party and conservative support is distributed among candidates roughly according to the general Republican electorate in each state.  Jay Cost has pointed out the regional (South vs. North) disparities in support, and others like Sean Trende, while agreeing that "the split is about something other than ideology[,]" have highlighted the disproportionate role of evangelicals in voting for somebody other than Governor Mitt Romney.  These are closer to the more fundamental division, which reflects how conservatives perceive themselves and their beliefs: as individuals who hold in common a set of policies they believe to be right, or alternatively, as a group united in culture and ideology, engaged in political conflict with other opposing groups.

We are all simultaneously both individuals and members of various groups, so there is nothing "irrational" about either way of cognizing the political environment and one's own role in it.  Many people have both intuitions at different times.  However, dominant perspective taken -- group or individual -- will shape the way political information is processed and the emotional response to the candidate.  In particular, perceiving elections as primarily about the advancement of group interests will often lead to some form of identity politics.

Differences in cognitive style, as we have pointed out earlier, are a profound driver of political behavior.  The clearest predictor of not voting for Romney is not whether voters call themselves conservative, but whether their overriding concern -- beyond electability or experience or anything else -- is that the candidate be a "true conservative": implicitly, a person like them.  In Florida and Illinois, for example, this group of voters gave Romney only 11% of their votes.  Even "strong" supporters of the Tea Party, "very" conservative voters, and born-again evangelicals voted for Romney at roughly triple this rate, leading to narrow victories or strong second-place finishes in these demographics.

In other words, the marker most clearly associated with a preference for identity politics is also the strongest and surest predictor of the electoral division. An intuition of whether the candidate is a good "representative" of the group which the voter feels a part of undoubtedly plays a role in other voters' minds (especially those who think of themselves as "very" conservative or "strong" supporters), even if not always an overriding one.

But in the preference for '"true conservatism," the concern comes clearly to the fore and suggests that Romney has had little success as an identity politician -- likely because he has not sought to promote himself in this way.  His appeal is stronger to those who are, for want of a better term, conservative individualists: people with right-of-center views who do not usually think of themselves as part of group, movement, or cultural tradition.  The raging debate in the Republican primary over who is and is not a conservative is imprecise and beside the point -- there are simply different kinds of conservatives. 

The group vs. individualist tension is a second line of division in American political history, complicating the contest between conservative and progressive impulses.  The split is still apparent in the Democratic Party; although identity politics of the race, class, and gender variety has long had a dominant role on the progressive side, there remains a strain of liberal individualists.  Indeed, President Obama, cleverly playing against expectation, has sought at times to reach out to this individualist tradition, invoking Theodore Roosevelt, one of individualists' most important figures, as inspiration.  Whatever the validity of this invocation, it highlights the fact that in some sense there are four "natural" political parties which have been consolidated into two operating parties by our system, but not always in the same way.

For much of the twentieth century, Conservative Groups and Progressive Groups found an uneasy but lasting home in the Democratic Party, divided largely between South and North, but united at the presidential level.  Meanwhile, Conservative Individualists and Progressive Individualists, descendants of the 1912 split between Taft and Roosevelt, formed the "wings" of the Republican Party.  In the late 1960s Republicans included both John Lindsay as mayor of New York City and Barry Goldwater as senator from Arizona; meanwhile, Coleman Young ran Detroit as a Democrat, joined in political affiliation with George Wallace in Alabama.

The late-twentieth-century realignment of identity conservatives with Republicans and the corresponding realignment of progressive individualists with Democrats are well-known.  But the South, although its identity is no longer framed in racial terms (despite the strenuous claims to the contrary by many on the left), still has a larger proportion of people who go beyond policy preferences to think of themselves as "cultural conservatives."  Not surprisingly, the Southern states are a challenging area for Romney.  The real "outgroup" that most identity conservatives oppose is not ethnically or economically defined, but consists of a cultural elite that includes most "mainstream" news and opinion outlets, along with the arts establishment and academia.  So when the conservative voters of South Carolina saw Speaker Gingrich, "one of their own," attacked by their main adversary during the debates just prior to the primary election there, they came out in force to oppose this, boosting him to victory.

Ordinarily, successful presidential candidates unite at least two "natural" parties -- they are successful by the very fact that they are (or appear to be) in the middle of their electoral coalition.  So it is not surprising that politicians like Richard Nixon or Bill Clinton defy easy classification.  The factions are easier to see in unsuccessful, but serious, presidential contenders, four of which are listed in the figure below as exemplars of people who became clear leaders of their natural party, but who could not succeed in generating broad enough support from others to become president.  Only George McGovern acquired the nomination (although Robert Taft came very close), and he lost in a historic landslide.

In the early 1980s, Jesse Jackson clearly showed the expanding possibilities for an identity progressive, building into formidable electoral force his "Rainbow Coalition" of identity groups seeking to benefit from an expanded federal role.  But he was undone for the nomination by an inability to appeal to those outside those groups.  Pat Buchanan in the 1990s, meanwhile, had some success in the Republican primary by running on a platform of social conservatism (especially opposition to gay rights), immigration restriction, and a "culture war."  Yet his in-group vs. out-group appeal could never command majority support in the overall Republican Party, prompting Buchanan to later abandon it to run as a third-party candidate. 

McGovern's 1972 campaign was based on support for progressive "new rights" including abortion, drug legalization, the Equal Rights Amendment, and a guaranteed minimum income.  But he lost the formerly "solid South," and he failed to garner sufficient compensating support from the general population.  Taft, the mid-century's "Mr. Republican," was an indefatigable champion of free enterprise and traditional individual rights to property and freedom of contract.  His narrow defeat for the nomination in 1952 is commonly attributed to the establishment belief that his ideological consistency posed too much of a burden on his electability.

In other political systems, our natural parties could run under their own labels and form explicit rather than implicit coalitions.  For example, it would not be difficult to imagine that in a German context, someone like Santorum could emerge as the leader of the Christian Democrats -- a party guided in part by ideals of religious social thought and national traditions.  Meanwhile, someone like Romney would be a natural leader of its traditional coalition partner, the free-market Free Democrats.

In the United States, however, our parties are coalitions appearing as unitary entities.  The party that can field the broadest coalition by attracting more marginal voters to either of its wings will be victorious.  A conservative individualist must successfully reach out in a general election to group conservatives as partners, assuring them that the concerns of their groups will be addressed, while at the same time being able to attract moderate and pragmatic voters who do not identify with any group or movement.  This would appear to be the Romney strategy (as it was Reagan's), and it is plausible if well-funded and executed.

The open question is whether identity conservatism could truly engineer an electoral upset against an incumbent president who remains personally popular with centrist voters.  The hostility the media evinces toward identity conservatives is unfair -- why is identity politics acceptable on the left, but illegitimate on the right? -- but it is real and effective.  When Texas Governor Rick Perry was still seeking the nomination, he released this well-known ad, a classic appeal on multiple levels to conservative group affiliation, suggesting a "war on religion" whose first skirmishes are the suppression of religious expression in the schools and the open service of gays in the military.  The ad was widely ridiculed.

Our point is not to argue the rights and wrongs of the ad's substance, but to point to an observable fact: a massive counter-reaction is created by the rhetoric of identity conservatism, fed by the media but sustained because unaffiliated voters are easily made uncomfortable by language that suggests exclusion and division.

For example, although Gingrich asserts that he has a more moderate stance on illegal immigration than Romney, it was Romney who in fact overperformed among Latino voters in Florida, winning a clear majority.  So we have some evidence that Latinos (a crucial swing constituency in the general election) are comfortable with a conservative individualist holding a consistent rule-of-law policy position on illegal entry, but their willingness to support a more identity-oriented politician (even one making efforts to gain their support) is less certain.  Governor Romney's overwhelming victory (in excess of 80%) in Puerto Rico, following a statement by Santorum interpreted by Spanish-speakers as a demand that a shift to the English language would be required before statehood, simply strengthens this concern.

More broadly, any member of a group which has not historically been part of the conservative coalition (i.e., ethnic and religious minorities, gays and lesbians, and political independents of various kinds) will naturally be hesitant to support identity conservatives, in part from the aftereffects of past exclusion, but also from residual uncertainty about their current inclusion.  In a November election in which every vote counts, these are ones Republicans cannot afford to lose.

Charles N. W. Keckler, a deputy assistant secretary for policy, and Ryan L. Cole, a deputy counselor for policy and speechwriting, both served in the Administration for Children and Families in the administration of President George W. Bush.