The Changing Political Landscape

Beating an incumbent member of Congress is incredibly difficult because they are prolific fundraisers, have the support of local and national party leaders, and successful campaign experience. If adept, they may also try to shepherd several perks of incumbency like their salary, travel, office, staff, and communication allowances, worth more than $1 million (Jacobson Reference Jacobson, Bond and Fleisher2000, 31); their ability to bring home pork to the district (Alvarez and Saving Reference Alvarez and Saving1997; Bickers and Stein Reference Bickers and Stein1996; Feldman and Jondrow Reference Feldman and Jondrow1984); and constituent casework (Fiorina and Rivers Reference Fiorina, Rivers, Fiorina and Rohde1989; Johannes and McAdams Reference Johannes and McAdams1981; Serra Reference Serra1994; Serra and Cover Reference Serra and Cover1992) to increase their chances of reelection. As a result, attempts to oust incumbent lawmakers from within their own party have been exceedingly rare in the past few decades,Footnote ¹ and successful attempts even rarer. Since 1980, incumbent House members pursuing reelection have lost to primary challengers 1% of the time (61 out of 5,969 reelection attempts).Footnote ² As a result, intra-party challenges have generally been reserved for incumbents tarred with scandal, when parties believe that the challenger may have a better chance of beating the opposing party in the general election, or launched by challengers with their own stores of personal wealth, independent from party prerogatives.Footnote ³

However, some gradual but profound developments are changing this calculus. The first is the growing prevalence and influence of ideological interest groups. As the purview of government programs has expanded and created constituencies with an incentive to organize and protect such programs through collective action, the number of interest groups has exploded over the past few decades (Walker Reference Walker1992).

Changes in technology and the rise of the Internet have dramatically and quickly affected grassroots organizing. Ideologically passionate activists, after all, are a tiny share of the population; finding others who share a similar zeal, mobilizing participation, and channeling resources were, in the past, activities reserved for the well moneyed and the socially connected. The now widespread use of the Internet has enabled those with even the most obscure political obsession to find individuals of like mind to chat, rant, and commiserate. For organizers, forming a large online community of individuals from across the country is cheap and easy; circulating news and messages to a vast, geographically dispersed memberships requires no paper, envelopes, or stamps, only the time it takes to type them; and recruiting new members is just a few clicks of a mouse away.Footnote ⁴

A second development is party polarization, a process noted by many congressional scholars (Binder and Smith Reference Binder and Smith1997; Fleisher and Bond Reference Fleisher, Bond, Bond and Fleisher2000; Jacobson Reference Jacobson, Nivola and Brady2006; Jamieson and Falk Reference Jamieson, Falk, Bond and Fleisher2000; McCarty, Poole, and Rosenthal Reference McCarty, Poole and Rosenthal2006a; Poole and Rosenthal Reference Poole and Rosenthal1997; Rohde Reference Rohde1991; Sinclair Reference Sinclair1997; Reference Sinclair, Bond and Fleisher2000; Stonecash, Brewer, and Mariani Reference Stonecash, Brewer and Mariani2002; and others) that began decades ago as conservative Democratic members from the South began being replaced by conservative Republicans. Different scholars place varying degrees of emphasis on different causal mechanisms: Southern economic development and integration (Key Reference Key1949), technological innovation and interregional migration (Polsby Reference Polsby2004), Democratic Party elites' decisions to embrace liberal racial policies in the 1960s (Carmines and Stimson Reference Carmines and Stimson1989), redistricting (Fiorina, Abrams, and Pope Reference Fiorina, Abrams and Pope2005),Footnote ⁵ and the changing stereotypes of party constituencies (Green, Palmquist, and Schickler Reference Green, Palmquist and Schickler2002). But all agree on a common result: white Southern conservatives, once a key component of the New Deal coalition, began to vote in greater numbers for Republicans in presidential elections, to elect Republicans to their congressional delegations, and eventually to identify as Republicans. Both among voters and in the halls of Congress, initial movement of Southern conservatives from the Democratic to Republican parties had two reinforcing effects: it made the Republican Party more homogenously conservative and the Democratic Party more uniformly liberal, encouraging further individual-level sorting as well as generational replacement that continues today.

The result is unprecedented disagreement between Democrats and Republicans on the major issues of the day. Polarization in Congress is at the highest level in 120 years, with virtually no ideological overlap between the two parties' memberships: practically every single Democrat is to the left of every single Republican (Poole Reference Poole2008).

This polarization has had a profound effect on the way Americans view the parties: they see more meaningful differences between Democrats and Republicans (Hetherington Reference Hetherington2001) and ideology is a more potent component of partisan stereotypes (Murakami Reference Murakami2008). Because ideologically atypical partisans, that is—liberal Republicans and conservative Democrats—are now so rare, they increasingly seem like the odd men out. To activists concerned with promoting an ideological agenda, they are at best antagonists who diminish the credibility of the party line, and at worst political traitors. Their growing exceptionalism makes their notable disunity all the more egregious, and also makes them more manageable, easily identifiable targets.

These developments have resulted in the increasing relevance of a particular kind of organization seeking to flex its political muscle: the party purity group. These organizations side ideologically with one the two major political parties, but are frustrated with moderates, who seem to betray core party values by siding with the opposition, even if only on a narrow range of issues. Changes in communication technology enable these groups to act on these growing frustrations stoked by the polarized political milieu. With the ability to garner large memberships, raise hundreds of millions of dollars, air attack ads, and recruit candidates, party purity groups are able to heavily influence party primary processes. But unlike the parties themselves, they are unhindered by the responsibilities of governing (and any accompanying necessary compromise) or maintaining a “big tent” essential to increase its share beyond only the most reliably ideologically friendly districts and states. Thus, in seeking to enhance the ideological purity of their home party, their efforts pit them in grueling, costly proxy battles with the party establishment itself.

Scalps on the Wall

The favorite electoral arena for party purity groups has been in open primaries, where less expensive airtime and a smaller pool of likely voters make them cheaper and easier opportunities to influence outcomes; where voters are much more ideologically driven than in general elections, making them especially receptive to messages advocating more conservative Republican candidates, or more liberal Democratic ones; and where the absence of incumbents makes running campaigns cheaper and less risky. Organizations like Emily's List (a group dedicated to funding and assisting pro-choice, women Democrats to Congress), MoveOn.org (a liberal, pro-Democratic group formed during the impeachment of President Clinton), and the Club for Growth (a conservative, anti-tax group) have become experts at identifying potentially competitive congressional primaries and promising political talent, recruiting large memberships to organize and contribute money, and assisting—sometimes independently—with messaging and campaign operations. Some of these organizations eventually found, however, that their influence in such open primaries was insufficient in producing the desired legislative results. As one reporter put it, they “began to see that candidates would say anything to get a pile of checks, only to ignore their promises … once they got elected” (Bai Reference Bai2003). They needed a new approach: targeting incumbents.

One of the first races of national prominence in which a party purity group launched a campaign to unseat a member of Congress took place in 2004, when the Club for Growth targeted veteran lawmaker Senator Arlen Specter of Pennsylvania primarily for his opposition to the full amount of President Bush's original tax cuts. The group's president, Stephen Moore, explains why the group decided to move beyond weighing in on open primaries, even though ousting an incumbent is much more difficult: “If we beat Specter, we won't have any trouble with wayward Republicans anymore. It serves notice to Chafee [R-RI], Snowe [R-ME], Voinovich [R-OH] and others who have been problem children that they will be next” (Dao Reference Dao2004a).

By taking down an incumbent, they can hit two, three, even four birds with one stone. As Specter himself noted, “he wants ‘my scalp on the wall’—so they can make other Republican Senators ‘behave’ “(Budoff Reference Budoff2004).

To garner the prize, the organization recruited Congressman Toomey—a conservative member of the House. Table 1, which displays the funds raised from all non-business, non-trade, and non-union PACs for this congressional primary (and other key races I will discuss shortly), shows that he was able leverage his House incumbency to raise some funds from politicians, especially former members of CongressFootnote ⁶ and his fellow congressmen.Footnote ⁷ In total, however, this amount was quite small: only 8% of his PAC funding from such Republican sources. This is by far the most a challenger recruited by a party purity group has raised from party insiders and was due in large part to his being a House incumbent and, in some respects, an insider himself. The lion's share of this type of campaign funding came from conservative organizations, including more than $440,000 from the Club for Growth and $20,000 from various right-to-life, pro-Israel, and anti-union interests.

Table 1.

The Funding of Incumbents and Their Primary Challengers

Source: The Federal Election Commission.

Note: Figures show the total amount of contributions from each category made to each candidate before the date of the relevant primary. (i) denotes incumbent running for reelection.

‡ Only donations made prior to the primary on August 8, 2006, are included, even though Senator Lieberman remained in the race to run (and ultimately prevail) as an Independent against Democratic nominee Ned Lamont in the general election, and include funds spent as independent expenditures on behalf of candidates. These figures do not include contributions from business or labor PACs.

The Republican establishment as a whole, in contrast, was not supportive of his bid. The National Republican Committee and the National Republican Senatorial Committee weighed in heavily in Specter's corner, citing the need for the moderate senator to maintain the Republican's narrow two-seat majority. Republican congressional strategists believed that Toomey would be more easily defeated in the general election and were depending on a Specter victory as part of their national calculus for retaining their two-seat majority in the Senate (Dao Reference Dao2004b). Articulating the party's viewpoint, Specter noted that “virtually everybody except the Club for Growth and my primary opponent agree that a big tent is necessary to be the prevailing national party” (Dao Reference Dao2004a). Republican strategists also believed that President Bush would be better served with moderate Specter downballot in this key battleground state. As a result, party stalwarts like Vice President Cheney, Karl Rove, and President Bush himself campaigned on behalf of Specter, who also received more than $634,000 from the Republican Party and leadership PACs of GOP office holders, including $550,000 from the National Republican Senatorial Committee. Moreover, the national party used the primary as an opportunity to test their get-out-the-vote operation in Pennsylvania.

Specter was also buoyed by a collection of organizations designed to support more moderate Republicans, which I dub big tent groups. Such groups like the Main-street Partnership, the Republican Majority for Choice, Republicans for Choice, and the Log Cabin Republicans were created to decrease party purity and added almost $40,000 to his primary coffers, and several pro-Israel groups gave close to $60,000 as well.

In the end, with all of the help of the Republican political machine, the Bush administration, and the perks of incumbency, as well as outspending Toomey by more than 4 to 1, Senator Specter squeaked by his challenger by a razor-thin 50.8 to 49.2%. He went on the win the general election by more than 10%.

Two years later, motivated anew by the close margin—rather than deflated by the loss, the Club for Growth tried again, this time to unseat Republican senator Lincoln Chafee of Rhode Island. This time they recruited Stephen Laffey, mayor of Cranston (the state's third-largest city). Like Toomey, his critical support came not from elements of the Republican Party organization or from Republican elected officials, but from a coalition of independent groups pushing for a more conservative Republican conference. The Club for Growth contributed more than $531,000 to his campaign,Footnote ⁸ but groups promoting a strongly pro-Israel stance as well as other conservative ideological groups chipped in for a grand total of almost $556,000. Unlike Congressman Toomey, however, Laffey was not able to earn a single dime from Republican leadership PACs. Senator Chafee, on the other hand, was able to gain over $1 million from the Republican Party and Republican politicians, including over $850,000 from the National Republican Senatorial Committee. He was also propped up significantly by liberal interest groups like the Human Rights Campaign, the Sierra Club, and NARAL Pro-Choice America, which contributed over $160,000; and a coalition of big tent groups similar to those who supported Senator Spector, which gave almost $43,000. Overall, over 85% of Specter's support came from party sources and big tent groups, while virtually all of Mayor Laffey's support came from conservative party purity groups. Following in Toomey's footsteps, Laffey lost the primary to Chafee, who went on to lose the general election by seven percentage points to Sheldon Whitehouse, the state's Democratic attorney general.

In the other fierce primary battle for a Senate seat of 2006, Democratic senator Lieberman was challenged from the left for his continued support for the War in Iraq. This race shows a similar pattern. Challenger Ned Lamont, a millionaire who largely self-funded his primary campaign, received PAC assistance only from liberal interests groups totaling about $268,000, of which more than $250,000 came from MoveOn.org. Senator Lieberman raised a smaller amount from liberal groups like the Human Rights Campaign and NARAL Pro-Choice America ($74,000), but that total was augmented by about $20,000 from conservative, pro-Israel groups, and over $94,000 from PACs run by fellow Democratic politicians. In the end, however, the support from Democrats and the benefits of incumbency were not enough: Lamont won the primary by 3.6 percentage points. Party purity groups would have had their first scalp, but Senator Lieberman petitioned to run in the general election as an Independent and beat Lamont with the broader electorate by 10 points.

In 2008, however, party purity groups of both the right and the left would score victories—both in House primaries, and both in Maryland. In the Republican primary for Maryland's first House district, we see the pattern of party establishment support for the incumbent pitted against ideological interests continue. Incumbent Wayne Gilcrest gained almost $224,000 from liberal group PACs—especially pro-environment organizations like the Humane Society, Council for a Livable World, Defenders of Wildlife Action, Friends of the Earth, and National Wildlife Action; as well as $72,000 from big tent groups like the Republican Mainstreet Partnership PAC, Republican Majority for Choice, Republicans for Environmental Protection, and the Tuesday Group PAC. These sums are overwhelmed, however, by the roughly $616,000 in support of challenger Andy Harris by conservative groups—again led by the Club for Growth, which spent in almost $598,000 on his behalf. On Election Day, Gilcrest lost to Harris by 10 percentage points.

In Maryland's fourth district, the Democratic primary pitted incumbent Albert Wynn, whose voting record revealed a social-conservative streak, against second-time challenger Donna Edwards, whose campaign depended heavily on the support of liberal groups like the League of Conservation Voters, the Sierra Club, Emily's List, and, again, MoveOn.org. She gained no PAC support from Democratic officeholders. Wynn, in contrast, while garnering only modest support from liberal groups and some funds from pro-Israel organizations, received many more contributions from his fellow Democratic incumbents—especially from members of the Congressional Black Caucus. All told, 71% of these PAC donations came from sitting Democrats. Despite support from his peers, he lost badly to Edwards by a margin of 22 points

Political Consequences of Warning Shots and Sinkings

A few details emerge from these data. First, there is a partisan asymmetry: virtually no PACs are devoted to supporting conservative Democrats, while several organizations reliably coordinate to support moderate Republican incumbents. However, the importance of this asymmetry's impact can be overstated: in each of the three Republican races, these big tent PACs were easily outspent by the Club for Growth and other conservative organizations by 11 to 1 against Senator Specter, by almost 13 to 1 against Senator Chafee, and by more than 17 to 1 against Representative Gilcrest. Incumbents from both parties have depended on lopsided-to-unanimous support from party organizations and fellow office holders to match the dollars spent by these party purity groups.

Second, some ideological groups do display a willingness to give modest levels of support for incumbents of both parties—environmental groups do come to the aid of a pro-environment Republican like Gilcrest. Most ideological or issue groups are not striving solely for party purity. But while the Club for Growth and MoveOn.org may be rare, they are disproportionately influential. Even when combined with big tent groups, in no case have other ideological groups been able to come even close to matching the funds they provide to promote party purity (and these figures do not include their bundling of individual contributions, which easily outstrip the PAC sums discussed here).

And third, elements of party organizations and partisans in government have been almost unanimously supportive of their ideologically atypical colleagues running for reelection (at least in the primary). Competitive primaries, and even competitive primaries borne of intra-party ideological conflict, after all, are not new. In 1938, for example, President Roosevelt targeted conservative Democrats, mostly from Southern states, in an attempt to “purge” both chambers of lawmakers who had proven unsympathetic to his some of his later New Deal legislation. What does seem noteworthy about these recent races (and many more not included in the table for space considerationsFootnote ⁹) is that they illustrate the tension, and diverging interests, between the party establishment and party purity groups. Parties and elected officials are concerned with maintaining a big ideological tent, tend to grant individual members significant leeway in responding to the ideological demands of their districts (Mayhew Reference Mayhew1974, 99–100), seem more likely to recognize the need to compromise in a highly polarized system of checks and balances, and also perhaps fear that they too might one day become the target of an ideological purge sometimes unpredictably and arbitrarily applied (for a good discussion, see Bai Reference Bai2003). Party purity groups, who believe that increasing party unity is strategically the best way to advancing their policy agenda, are unencumbered by such a panoply of concerns—especially when their home party is in the majority and they believe such a majority to be reasonably secure.Footnote ¹⁰ That a good portion of the partisan strife in Congress is occurring despite, rather than because of, the parties themselves is an underappreciated irony.