Feb. 24, 2011, 11:40 a.m.
In the long march toward a more parliamentary and partisan Washington, National Journal’s 2010 congressional vote ratings mark a new peak of polarization.
For only the second time since 1982, when NJ began calculating the ratings in their current form, every Senate Democrat compiled a voting record more liberal than every Senate Republican — and every Senate Republican compiled a voting record more conservative than every Senate Democrat. Even Nebraska’s Ben Nelson, the most conservative Democrat in the rankings, produced an overall voting record slightly to the left of the most moderate Republicans last year: Ohio’s George Voinovich and Maine’s Susan Collins and Olympia Snowe. The Senate had been that divided only once before, in 1999.
But the overall level of congressional polarization last year was the highest the index has recorded, because the House was much more divided in 2010 than it was in 1999. Back then, more than half of the chamber’s members compiled voting records between the most liberal Republican and the most conservative Democrat. In 2010, however, the overlap between the parties in the House was less than in any previous index.
Just five House Republicans in 2010 generated vote ratings more liberal than the most conservative House Democrat, Gene Taylor of Mississippi. Just four Democrats produced ratings more conservative than the most liberal Republican, Joseph Cao of Louisiana. Every other House Republican produced a more conservative vote rating than every other House Democrat, even though a substantial number of those Democrats pursued a relatively moderate course overall. Of the nine members who were outliers last year, only one — Republican Walter Jones of North Carolina — is still in Congress. That makes him the only lawmaker in the House or Senate this year to have a 2010 vote rating out of sync with his party.
The results document another leap forward in the fusion of ideology and partisanship that has remade Congress over the past three decades, the period tracked by NJ‘s vote ratings. For most of American history, the two parties operated as ramshackle coalitions that harbored diverse and even antithetical views. Each party’s Senate caucus housed ideological antagonists, such as progressive Democratic titan Hubert Humphrey of Minnesota and segregationist stalwart Richard Russell of Georgia, or New Right Republican firebrand Jesse Helms of North Carolina and silk-stocking New York City liberal Jacob Javits. Such contrasts are not extinct. But since the early 1980s, they have vastly diminished as the differences within each party have narrowed and the distance between them has widened.
Over that period, “it’s just a straight, linear increase” in congressional polarization, says Gary Jacobson, a University of California (San Diego) political scientist who specializes in Congress. “There’s a little bit of bumping around in the numbers here and there, but the basic movement is toward the parties moving further and further apart. The 1970s are a high point of all the cross-party [coalitions]. The last three decades are ones of pulling apart.”
In 2010, the vote ratings show, the ideological consolidation was greater among Republicans than Democrats. Almost without exception House Republicans generated strongly conservative voting records, regardless of the demography or political leanings of their districts. By contrast, House Democrats from districts that voted for John McCain in 2008 or are dominated by working-class whites produced much less liberal records than their colleagues from districts that strongly supported Barack Obama or are more racially diverse and well educated. In the Senate, just eight Republicans notched a composite conservative score of less than 70, while 21 Democrats received a liberal ranking of less than 70.
The results capture the continued remaking of Congress into an institution defined by much greater partisan discipline and philosophical conformity. Occasionally, legislators can still build idiosyncratic coalitions across party lines, as occurred during some of the votes on the free-wheeling House debate over spending earlier this month. Likewise, a bipartisan group of senators is attempting to build a cross-party alliance to advance the recommendations of President Obama’s debt-reduction commission.
But increasingly, on the biggest issues, the parties line up in virtual lockstep against each other, as they did on many of the key measures in the 2010 rankings, such as the Senate votes on health care and financial-services reform. (Even on the House’s final vote last weekend on funding the government through September, every Democrat voted in opposition and all but three Republicans voted in support.) All of this is fundamentally changing the way Congress gets things done — when it gets things done at all. “If you are the whip in either party you are liking this, [because] it makes your job easier,” says Mississippi Republican Trent Lott, the former Senate majority leader (and before that the GOP Senate whip). “In terms of getting things done for the country, that’s not the case.”
For those who have come of age in today’s hyperpartisan Congress — with its near-parliamentary levels of party discipline on floor votes, jagged ideological confrontations, and dominant role for leadership — it’s easy to forget how different the institution looked as recently as the early 1980s, when NJ began measuring members’ votes on a liberal-to-conservative scale.
The first time NJ calculated congressional votes using the scale it employs now, in 1982, the results revealed a Congress that operated in a manner that would be unrecognizable today.
John Danforth, a moderate Republican senator from Missouri, was finishing his first term in 1982. He remembers that soon after he arrived, Russell Long of Louisiana, the venerable Democratic powerhouse who chaired the Senate Finance Committee, gave him a singular piece of advice. ” “˜Don’t ever hold grudges, because your strongest opponent today could be your ally tomorrow,’ “ Danforth, who retired in 1994, recalled in a recent interview.
That advice made sense in the Senate of those years, because both caucuses were much more diverse and unpredictable than they are today. In NJ‘s 1982 vote ratings, fully 36 Senate Democrats compiled records at least as conservative as the most liberal Republican, Lowell Weicker of Connecticut. From the other direction, 24 Senate Republicans compiled voting records at least as liberal as the most conservative Democrat, Edward Zorinsky of Nebraska. Zorinsky, in fact, received a rating exactly as conservative as Arizona Republican Barry Goldwater, whose 1964 presidential campaign ignited the modern conservative revival.
The senators with voting records that fell between the most liberal Republican and the most conservative Democrat represented a pool of idiosyncratic, unattached pieces that could be assembled and reassembled in constantly shifting coalitions to pass or block legislation. In such a fluid environment, it virtually defied conceptualization to define a typical Democrat or typical Republican senator.
The Democrats who generated less liberal records than Weicker included New South moderates such as David Boren of Oklahoma and Sam Nunn of Georgia and Old South conservatives such as ancient John Stennis of Mississippi and Harry Byrd of Virginia, as well as coastal neoliberals such as Bill Bradley of New Jersey. The Republicans more liberal than Zorinsky included a phalanx of brainy New England moderates, among them Weicker, William Cohen of Maine, Warren Rudman of New Hampshire, John Chafee of Rhode Island, and Robert Stafford of Vermont, a champion of the modern environmental movement. Issues frequently divided the parties along ideological and regional lines. When Helms pushed a constitutional amendment to allow school prayer, Weicker and Danforth helped lead the fight to stop him.
“The overarching point is that the Senate was comprised of 100 individuals who had a loose binding with the respective parties,” says Weicker, who left the GOP in 1990 to win the Connecticut governorship as an independent. “There were more conservative Democrats, more liberal Republicans. You had people who stood on their own two feet.”
In the three decades since, NJ‘s vote ratings have tracked the narrowing of that Senate center. By 1994, the second year of Bill Clinton’s presidency, 27 Democrats compiled more conservative NJ voting records than the most liberal Republican, James Jeffords of Vermont (who also later left the GOP to become an independent). Just nine Republicans compiled voting records more liberal than the most conservative Democrat that year — Richard Shelby of Alabama; Shelby, too, later switched parties, joining the GOP. In 1999, with Clinton’s impeachment looming over the chamber and the parties recoiling in the aftermath of the grassroots conservative backlash against the 1997 balanced-budget deal, NJ found no Senate crossover between the parties for the only other time.
By 2002, the second year of George W. Bush’s presidency, some overlap returned, but just two Democrats compiled a more conservative voting record than the most liberal Republican, Rhode Island’s Lincoln Chafee. (Continuing the pattern, Chafee was elected governor as an independent last November.) Just seven Republicans racked up voting records more liberal than the most conservative Democrat, Georgia’s Zell Miller (who never switched parties but did endorse Bush at the 2004 GOP convention).
In 2010, the second year of Obama’s term, this process of separation reached another apex, with no overlap between the ideological scores of senators from the two parties. Taking the long view, the trajectory from Ronald Reagan’s second year to Obama’s is stark: In 1982, 58 senators compiled voting records that fell between the most liberal Republican and the most conservative Democrat. By 1994, the number was down to 34. By 2002 (after touching zero in 1999), it stood at just seven. And now it has returned to zero. “Over the years, there is no question that the middle in the Senate has shrunk considerably,” says Lott, now a Washington lobbyist and a senior fellow at the Bipartisan Policy Center.
In the House, as noted earlier, some ideological overlap remains. But the basic story is the same — and in some ways is even more dramatic. In 1982, the days of conservative Democratic “Boll Weevils” and liberal Republican “Gypsy Moths,” fully 344 House members received NJ vote ratings between the most liberal Republican (Rhode Island’s Claudine Schneider) and the most conservative Democrat (Georgia’s Larry McDonald). Even as recently as 1999, 226 House lawmakers compiled ratings between the most liberal Republican and the most conservative Democrat. By 2005, the number between those two poles fell to 54. By 2010, the number of members between those two boundaries had shriveled to seven.
The separation between the parties might not always be as pronounced as in the 2010 ratings. Some Senate Republicans (Scott Brown of Massachusetts, say, or Mark Kirk of Illinois) might easily compile more-moderate voting records than Democrats Nelson of Nebraska or Joe Manchin of West Virginia, particularly if both tilt to the right in anticipation of tough 2012 reelection campaigns. As Michael Franc, vice president for government relations at the conservative Heritage Foundation, notes, it may have been easier for Republicans to achieve unanimity in opposition to Obama’s agenda than it will be for them to do so while trying to pass their own programs.
Yet, the underlying trend toward the parties pulling apart in Congress is unmistakable, and, in the eyes of many analysts, probably irreversible. “The two parties,” says Washington lobbyist Vic Fazio, the former chairman of the House Democratic Caucus, “increasingly are at polar opposites.”
Though the dominant trend is increasing convergence within the parties, and widening divergence between them, the 2010 vote ratings reveal enduring fault lines in each chamber, particularly among Democrats.
The ratings measured 427 House members and 94 senators; the missing House and Senate seats were held by a person (or persons) who did not cast enough votes last year to warrant a score.
The results reaffirm the link between senators’ voting records and the behavior of their states in presidential elections. Senators whose states reliably support candidates from the lawmakers’ party in White House races have consistently compiled more-ideological voting records than senators whose states often prefer the other party or swing between them. (See “Serving Behind Enemy Lines,” NJ, 4/24/10, p. 25.) That pattern was vivid again in 2010.
Of the 21 Democratic senators with the most-liberal overall voting records, according to the ratings, 18 were elected from “blue wall” states that have voted Democratic in at least the past five presidential elections. The only exceptions to the pattern are Senate Majority Leader Harry Reid of Nevada and Ohio’s Sherrod Brown, who tied with seven others for the most liberal Democratic score; and first-termer Tom Udall of New Mexico, who tied for the 15th-most-liberal score.
Among Republicans, the 22 senators with the most-conservative vote ratings were all elected in states that voted Republican in at least the past three presidential elections. That group includes the eight who tied for the most conservative score — among them Jim DeMint of South Carolina, John Cornyn of Texas, and Mike Crapo of Idaho. In a striking measure of his repositioning since 2008, Arizona’s McCain also tied for the most conservative score among Republicans; as recently as 2001, in the aftermath of his defeat by Bush in the 2000 GOP primaries, McCain had generated the 39th-most-conservative record in the Senate.
In both parties, dissent is more common among the senators elected, in effect, behind enemy lines. These are the lawmakers who are often most interested in exploring compromises that round off the sharp edges of partisan conflict. Overall, the 30 GOP senators, for instance, elected from states that voted Republican in each of the past three presidential elections compiled an average composite liberal score of 17, meaning that as a group they were more liberal than 17 percent of their Senate colleagues. But the three GOP senators elected from states that voted Democratic in each of the presidential contests since 2000 — Collins and Snowe of Maine and Brown of Massachusetts — generated an average liberal score more than twice that, 37 percent.
The same holds true for Democrats. The 30 Democratic senators elected from states that voted Democratic in the past three presidential elections compiled an average liberal score of nearly 76. By contrast, the eight Democrats elected from states that voted Democratic for president only once since 2000 compiled an average liberal score of 67, and the dozen from states that have not voted Democratic since at least 2000 amassed an average liberal score of only 60. Except for iconoclastic Connecticut independent Joe Lieberman (who is retiring after next year), all 14 of the Senate Democrats with the most-conservative voting records, relatively speaking, represent states that have not voted Democratic more than once since 2000 — a list that includes Nelson of Nebraska, Jon Tester and Max Baucus of Montana, Mark Warner and Jim Webb of Virginia, Kay Hagan of North Carolina, and Michael Bennet and Mark Udall of Colorado.
Those pairings underscore another striking trend in the Senate results: the convergence among the ratings of senators from the same party who represent the same state. Although the gap between senators from opposite parties who hail from the same state (say Democrat Tom Harkin and Republican Chuck Grassley of Iowa) remains large, partisan pairs increasingly follow the same course. For instance, Democrats Carl Levin and Debbie Stabenow of Michigan and Barbara Mikulski and Ben Cardin of Maryland all tied for the most liberal ranking (as did Democrat Patrick Leahy and independent Bernie Sanders of Vermont). The other end of the ideological scale finds the overlap between Tester and Baucus, Webb and Warner, Bennet and Udall. Similar patterns are evident among Republicans. In all, 22 states were represented by senators whose vote ratings were within 5 percentage points of each other. In only 12 states did senators have vote ratings more than 25 percentage points apart.
This convergence may illustrate the diminished ability of senators to sail a distinct course, independent of the dominant political currents in their state. The frequent pairings suggest that senators are aligning more closely with their state’s underlying political balance, or at least the consensus in their party within their state.
Those who break from that consensus face an increasing risk of primaries driven by activists of the Left or Right; three senators — two Republicans and one Democrat — were denied renomination in 2010, almost as many as in the previous 26 years combined. “There is more of a demand in each party for a degree of purity or inflexibility that was not there before,” says Danforth, now a lawyer in St. Louis. Lott notes that the growing threat of such primary challenges (at least three more Senate Republicans could face serious opponents in 2012) powerfully reinforces the trend toward partisan and ideological conformity evident in the ratings. “You really need to toe the line,” he says. “That affects people’s thinking — both Democrats and Republicans.”
In the House, as in the Senate, Republicans pursued a more unified course in 2010 than Democrats did. The contrast between the parties was arguably even greater in the lower chamber. What’s more, many House Republicans compiled conservative voting records regardless of the demographic or political bent of their districts, while Democrats differed substantially based on those factors.
Among Democrats, for instance, there was a clear relationship between their 2010 vote rating and the way their district voted for president in 2008. The 124 House Democrats representing districts where Obama won at least 60 percent of the vote compiled an average liberal score of nearly 81, well above the party average of 70. In stair-step fashion, the average liberal score dropped to 69 for the 48 House Democrats in districts where Obama won between 55 and 59 percent, and to 63 for the 35 in districts that he carried with less than 55 percent of the vote. Most strikingly, the average liberal score of the 47 House Democrats from districts that McCain carried in 2008 stood at just 50 — fully 30 percentage points below the number for those holding the safest seats. Of the 50 House Democrats with the most-conservative voting records, 35 were from districts that McCain carried.
Among Republicans, the variation was much smaller. The 52 House Republicans from districts where McCain won at least 60 percent of the vote produced an average conservative score of nearly 83. The 33 members from districts where he won between 55 and 59 percent generated a slightly more conservative ranking of 84, and the number fell only slightly, to 78, among the 54 lawmakers in districts that McCain won with less than 55 percent of the vote. Even the 34 House Republicans from districts that Obama carried compiled an average conservative score of 72 — only about 10 percentage points less than those from the safest seats.
The story is similar when looking at the House through a demographic lens. In 2009, National Journal divided the chamber into four quadrants based on whether the share of the white population with college degrees exceeded the 30.4 percent national average, and whether the district’s minority population exceeded 30 percent, the level that an earlier NJ analysis found to be a revealing dividing line in election results. (See “The Four Quadrants of Congress,” NJ, 2/6/10, p. 20.)
In 2010, as in earlier years, House Democrats from districts high in both diversity and education posted much more liberal scores than those from districts low on both measures: the predominantly blue-collar small-town and rural seats represented largely by members of the Democratic Blue Dog coalition. Democrats from the “high-high” districts posted an average liberal score of 79, compared with an average liberal score of 62 for those from the “low-low” districts. In stark contrast, the 84 House Republicans from low-low districts posted the exact same 79.6 average conservative rating as the 30 Republicans representing districts high in diversity and education.
Why did House Republicans display so much more ideological unanimity than Democrats? One reason is that the GOP’s sweeping losses in 2006 and 2008 reduced the party mostly to strong Republican seats in the past Congress, leaving few members with an ideological inclination or electoral incentive to cooperate with Obama. Rep. Tom Cole, R-Okla., a former chairman of the National Republican Congressional Committee, says that the ambitious Democratic agenda also helped Republicans coalesce in opposition — in the political equivalent of Newton’s principle that every action generates an equal and opposite reaction. “I remember when the stimulus package came up for a vote: There are 178 of us, we just got our clock cleaned, and the normal reaction was to make peace with the winner,” Cole says. “And, of course, every single Republican voted no. I remember at the time somebody telling me, because I’m a deputy whip, “˜You guys did such a great job whipping that.’ I said, “˜It really was not hard to whip.’ You can’t be a Republican and be for this. Our members didn’t feel agonized.”
But the results also reflect a longer-term dynamic: Although both parties are growing more ideologically homogenous, the trend is affecting Republicans more powerfully and more thoroughly. Democrats remain more of a coalition party than the GOP. The roots of that trend extend to the foundation of each party’s electoral base. The Republicans’ voting coalition is much more ideologically uniform than the Democrats’: About three-fourths of GOP voters identify as conservative, while only about two-fifths of Democrats consider themselves liberal, with the rest calling themselves moderate or conservative. That creates a more consistent set of expectations from the base for congressional Republicans than it does for Democrats, no matter what part of the country they represent.
Because about twice as many voters consider themselves conservatives as liberals, Republicans are typically less dependent on support from moderates to win elections, which further amplifies the conservative influence over the party’s elected officials. “The Democrats are always going to be fractious and divided if they want to aspire to majority status,” says Jacobson, the political scientist. “That’s just the nature of their coalition. The Republicans don’t have to be that broad; they can be much more unified. Republicans are now a conservative and very conservative coalition, and their share of moderates is minuscule. The pressure right now is not coming from their centrists; it’s coming from their extremists.”
The difference is already apparent in the early months of 2011. After last November’s landslide, House Republicans hold 61 districts that Obama carried in 2008, but those GOP members are not straying from the party agenda nearly as much as McCain-district Democrats did from their party’s priorities over the previous two years. Every House Republican has voted to repeal the Obama health care plan; all but two voted on February 18 to block the Environmental Protection Agency from regulating carbon emissions; and all but three (just one of them from a district that Obama won) supported last week’s continuing resolution on funding that imposed the largest domestic discretionary spending cuts in modern times. “The polarization around here, the formation of two homogenous parties at the poles, is really asymmetrical,” says Rep. David Price, D-N.C., a former political scientist. “The ideological cohesion and voting discipline represents a homogenization of the Republican Party that just hasn’t taken place to the same extent on the Democratic side.”
At the broadest level, the trends in NJ‘s vote ratings over the past three decades track the decline of individualism in Congress. Throughout congressional history, the most respected legislators — from Henry Clay and Stephen Douglas to Lyndon Johnson, Bob Dole, and Edward Kennedy — have been those who through force of personality or intellect have been able to assemble coalitions and forge compromises that would not have coalesced without them. Such personalized acts of consensus-building still occur but much less frequently, and those who try face much steeper walls of resistance to compromise. “You don’t have so much individualism [anymore],” says Weicker, now president of the Trust for America’s Health.
Primarily, legislators in both chambers (especially the House) are asked to simply be foot soldiers — to support policy choices that their leadership forges, almost always in close consultation with the constituency groups central to the party’s coalition. Rather than being heralded as iconoclasts, those legislators who deviate too often from that centrally directed consensus now face pressure from their colleagues; a cold shoulder from leadership; blistering criticism from the overtly partisan media aligned with each side; and, with growing frequency, primary challenges bankrolled by powerful party interest groups. “A lot of these institutions have become [ideologically] monolithic in their own right, and that just reinforces the political divide,” the Heritage Foundation’s Franc says. “If you are a charismatic senator or House member who wants to change an issue, you are going to be swimming not just against your own caucus but all of these outside interests and the blogosphere.”
In all of these ways, Congress functions as a more top-down, parliamentary-style institution — the results of which are evident in the relentless separation between the parties and the decline of mavericks in NJ‘s vote ratings. Meanwhile, the results from the 2006, ‘08, and ‘10 elections suggest that congressional campaigns as well are operating in a parliamentary manner, in which assessments of individual candidates matter less than broad judgments about the two parties. Put another way, increasingly in congressional campaigns (especially for the House), it appears that the color on the front of the jersey matters more than the name on the back. That means members have less ability to separate themselves from attitudes about their party by voting against key elements of its agenda.
That trend screams from the latest House vote rankings. House Democrats who broke the most often from the party’s liberal consensus — the agenda that contributed to last November’s voter backlash — suffered by far the greatest losses in that election. Among the 81 most liberal House Democrats, just one who sought reelection was defeated (the Democrats lost the seat of one other who retired). By contrast, among the 98 Democrats with the most-moderate scores, 45 who sought reelection were defeated and the party lost the seats of 10 others who retired. Those legislators didn’t lose because they compiled more-conservative voting records, but neither was their distancing sufficient to save them from the tide that crested against their party in all but the safest Democratic districts.
“What you’re seeing now is, it’s harder to survive [a wave],” Price says. “The survival techniques that people adopted in these swing districts are less workable now. The elections become nationalized, and it’s a harder environment for members to deploy their usual survival tactics, like constituent service, and being nice people, and all the things members have counted on to protect them.”
The 2012 election will test how heavily these patterns extend into races for the Senate, where candidates typically have established more independence from general attitudes about their party, largely because they become better known than House members. Many of the Senate Democrats facing tough races next year compiled some of their party’s most conservative voting records — among them Ben Nelson, Bill Nelson of Florida, Claire McCaskill of Missouri, and Tester. But the past three elections suggest that that may be thin insulation unless attitudes about the overall Democratic agenda and Obama improve in their right-leaning states. “One lesson “¦ is you cannot localize elections like this,” says Matthew Bennett, vice president of Third Way, a centrist Democratic group. “You are a Democrat and an Obama Democrat “¦ [and] you had better find a way of explaining what you did and putting in context what he did, because that is going to define you.”
One final question raised by the long-term trends in NJ‘s vote ratings is whether a Congress attuned to these quasi-parliamentary legislative and electoral rhythms is more or less reflective of public opinion than the more fluid and unstructured institution of earlier generations. Bennett says that the polarization evident in the ratings has produced a Congress more divided than the country. “There’s been a lot of sorting out that has gone on the electorate, and there’s no question that districts and states are brighter hues of red and blue than they used to be,” he says. “But in aggregate, the plurality of the electorate is still moderate, and they are the most underrepresented category of voters at the moment.”
Cole, the Oklahoma Republican, disagrees. He believes that the hardening lines between the parties in Washington reflect a widening disagreement in the country over “fundamental first principles” revolving around the role of government. “Most of the Republicans I talk to, and my constituents, really believe that what’s at stake is, we are going to be a fundamentally different America on the other side of these [Obama] policies, and they feel it really strongly,” he says. “It’s not created by the system, not created by Washington politicians, but is a really profound debate that is beginning to emerge about what kind of country we are going to be.”
With all signs indicating that that debate will roar through Washington for the rest of Obama’s term, no one should expect the systematic separation of the parties that defines NJ‘s latest vote ratings to reverse anytime soon. Pulling apart has settled in as a defining characteristic of political life in modern Washington.
Contributions by Scott Bland
