Strengthening Constitutional Self-Government


The Permanent Democratic Congress

Norman J. Ornstein

 One of the most enduring, puzzling and contentious phenomena of modern American political life is the persistence of firm Democratic Party majorities in Congress even as the Republicans tighten their stranglehold on the White House. The enduring nature of the phenomenon can be seen easily with a few simple numbers. The Democrats have controlled the House of Representatives for 36 consecutive years, and 56 of the last sixty; they have run the Senate for fifty of the past sixty years. The Republican Party, meanwhile, coming out of the desert of two decades of presidential minority status in 1952, has successfully held the White House seven of the past ten terms, including three in a row and five of the past six. Since 1954, the United States has had 24 years with a Republican president and Democratic Congress (6 years with a Republican President and Senate and a Democratic House,) and only 12 years with a president and Congress of the same party.

The phenomenon has puzzled more than a few observers because it seems so irrational, even perverse. While coalition governments are common in parliamentary democracies, a stark division of powers not only between institutions but also parties was nearly unknown in the parliamentary world until the recent French experiment with “cohabitation.” Even seasoned American observers have a hard time understanding or explaining how voters can opt enthusiastically and overwhelmingly for a Republican president—who has campaigned by asking for a Congress he can work with—and then turn around collectively and choose by an equally overwhelming margin a Congress run by precisely the people he has asked them to reject. Indeed, the most common explanation has been that the electorate is not consciously making these choices, but is either being forced mechanistically or gulled by partisan congressional wiles to behave in a schizophrenic fashion.

That this phenomenon would bring contentiousness is not surprising; neither Democratic nor Republican partisans are happy at either the humiliation implicit in a long string of election defeats for major office, or in the loss of power and position in major national institutions. Democrats, the Denver Broncos of American presidential politics, are fearful that several generations of their talented youth will be shut out of any opportunity to occupy high executive branch office, leaving the party potentially without experienced elder statesmen into the twenty-first century. Meanwhile, Republicans, the Chicago Cubs of congressional politics, have not a single member of the House of Representatives who has ever been in the majority, with Republican leader Bob Michel possessing the dubious distinction of a record-breaking 32 consecutive years in the minority.

There are many reasons, in a volume about Congress, to explore the phenomenon of the persistent Democratic Congress through an era of Republican presidencies. There is the scholarly puzzle-sorting through why it exists. Beyond this interesting, if more academic issue, lie other, serious political, Constitutional and policy questions. Among other things, the persistent Democratic Congress has led to increasing frustration, and resulting hostility, among conservatives towards the First Branch of government.

For many, this phenomenon has called into question the very legitimacy of the Congress itself-since it cannot be natural, the unnatural causes and consequences must mean a chamber far removed from what the Framers intended, one so unrepresentative of the desires and wishes of the public that it should be denounced, delegitimized and defanged.

At the same time, the phenomenon has led to several new schools of constitutional thought. For decades the staunchest defenders of Congress were American conservatives, who generally agreed with James Burnham, one of the leading figures in conservative thought, whose bookCongress and the American Tradition was a forceful, well-documented and persuasive defense of congressional constitutional prerogatives against their encroachment by the presidency. In recent years, confronted by a presidency increasingly occupied by one friendly to conservatives and a Congress uniformly run by people and party hostile to them, many conservatives have reread the Constitutional scriptures and found new meaning in them, meaning that turns Burnham’s book on its head.

This would be no more than an interesting and amusing example of situational constitutionalism (the exact opposite hypocrisy, of course, is found among newly energized liberal defenders of Congress,) if it were not for the fact that serious efforts at major Constitutional reform are underway as a result. Some, like the line-item veto, are designed to enhance presidential authority and weaken congressional prerogatives. Others, such as a limitation in congressional terms, are aimed more directly at altering the long-standing balance of power in Congress.

Beyond these kinds of struggles, though, there is also no question that this persistent pattern of power has altered political dynamics and party psychology in Washington, shaping how we run our policy process, and what the outcomes might be. For all these reasons, a careful examination of the permanent Democratic Congress is in order.

Explaining The Permanent Democratic Congress

American history has tended to alternating eras of one party domination, rather than consistent close competition between the parties, with regular alternation in office. Most political historians divide the two centuries of American life into five distinct eras (or systems) of party dominance, each ended, replaced by a new age, with a party realignment. Of course, the dominance was never pure, and frequently involved different patterns for Congress than for the presidency, but it was still usually clear which party was the majority party in the country. The modern era of American politics is different. While we have had disagreements among scholars about whether in recent decades we have experienced a realignment or a dealignment, the divisions in power among the parties and institutions are very unusual, particularly for the Congress. Eighteen consecutive terms of one party majority in the House is twice as long as any other consecutive string in post-Civil War American history.

Why this unprecedented rule for the Democrats? To Republicans and many other observers of American politics, this Democratic success, especially through the 1980s when Republican party identification grew and Republican presidential control and popularity expanded, cannot be a natural phenomenon, or a reflection simply of Democratic political skill and/or Republican organizational failure. The key question for them is, why didn’t national Republican popularity translate into congressional strength?

The Power of Incumbency

One of the most widespread and popular theories is that the powers and perks of incumbency in Congress, which have grown exponentially in the past two decades, have created a situation where incumbents cannot be beaten except under the most extreme of circumstances. Since the Democratic Party has more incumbents and controls the system of power and perks, it has been able, by dramatically increasing the incumbent advantage, to institutionalize its majority, insulating it from normal political pressures and, in effect, mesmerizing voters to a point where they couldn’t hear an alternative message even if they wanted to.

This latter assertion is what gives particular bite and controversy to the incumbency explanation for contemporary Democratic dominance of the House. It suggests directly that the advantages of incumbents have grown so far beyond the “normal” or usual pattern that incumbents win even when voters really don’t want them to—in other words, that the strong Republican tides in presidential politics would have washed over Congress if it weren’t for the huge supply and size of incumbency sandbags erected in the last several years.

Proof for this contention comes from the astonishing success rates for incumbent House members, particularly in the past two elections (98.3% and 96%, respectively,) the growing number of House members winning elections with 60%, 65%, or 70% of the vote or more (in other words, the declining number of marginal members or seats,) and a catalogue of the impressive perquisites, from staff to mailing allowances and television studios, available to lawmakers. In addition, incumbents have had increasing success at raising the substantial sums of money needed to run modern congressional campaigns, especially money from organized sources like political actions committees (PACs), while challengers have had an increasingly rough time.

The success rates for incumbents, the phenomenon of the vanishing marginals, the perquisites of incumbency and the widening money gap, are all quantifiable and demonstrable. There can be no question that incumbency provides enormous advantages, and is at least a partial explanation for the phenomenon of Democratic one-party control. However, it is at best an incomplete explanation. To begin with, reelection rates for incumbents in the House have been high for more than a century, and have averaged more than 90% for several decades at least, with little real change coming with the expansion in incumbent perquisites or the changes in campaign financing. The mid-1950s, the late 1960s and the mid-1970s each saw a string of elections with reelection rates in the mid-90s, each followed by elections with more incumbent defeats. The mid-1980s cycle is at a marginally higher level of incumbency success than previous ones, it is true, and it is likely to persist through four elections while the earlier strings tended to be three, but it has also come with an unusually high and sustained level of voter satisfaction with the status quo.

Persistent high incumbency reelection rates tell us that incumbency is a powerful tool. But its power as a tool goes back many decades, before this unprecedented Democratic string. We have unquestionably seen incumbency solidify as a political phenomenon. But a sharper change than we have seen would be required before we could credit the modern revolution in congressional support for congressional Democratic hegemony.

Secondly, a major part in the revolution in incumbent advantage has come not from mechanical advantages but from the political savvy, energy level and drive of modern politicians—people who undeniably take advantage of the perks available to them, but who know how to reach their constituents with communications, insights, concerns and a work product that will succeed. We have some evidence to suggest that the infamous Class of 1974—the “Watergate Babies”—advanced and expanded the use of innovative political techniques to solidify their own sometimes precarious positions back home, with mobile district offices, twenty-four hour constituent hotlines and other modern marketing techniques. Whether they were pioneers or not, they and their contemporaries know how to communicate and campaign. Moreover, as the average age of members of Congress has declined (contrary to conventional wisdom, we have one of the least geriatric Congresses in modern times) the energy level has probably risen; combined with much more and more regular jet transportation out of National Airport, enabling lawmakers to spend full weeks in Washington and still have a highly visible and energetic presence in their districts, we have a formula for incumbent popularity back home.

In a recent issue of U.S. News and World Report, Michael Barone reported on visits he had made to the districts of two Democratic House incumbents who have been regular targets of Republican campaign committees, Robert Carr of Michigan and Tom Downey of New York. Barone portrayed two energetic, intelligent and attractive individuals who understood and responded to the local issues that mattered most to their constituents, and whose style, openness and drive overcame for many more conservative voters some of their non-conservative issue positions. No doubt, Carr and Downey would have had more trouble without the staff, franking privilege and other benefits available to every incumbent—but even without the perks, Barone’s reporting shows that they would not be easy to beat.

Incumbent advantage can be examined from many angles. As with Sherlock Holmes’ insight about the dog that didn’t bark, the most significant evidence that bears on the role of incumbency may come from the seats that have no incumbents—the open seats. If the most pointed theory of incumbency advantage is true—if Republican majorities have been thwarted by the shield of incumbency—then we should find major Republican successes where those advantages are absent. We don’t.

In seats where there are no incumbents running, there are no incumbent advantages; the sandbags disappear, and if national tides will shift local sands, it will happen in these seats. If the incumbency theory holds true, in other words, we should be able to find a large number of seats that had been held by Democratic incumbents going over to the Republicans when the incumbents leave those seats—and many fewer open seats held by Republican incumbents going over to Democrats when those seats open up.

The reality is the opposite. As Table I shows, since 1954, Republicans have won 77 open Democratic seats in the House of Representatives; Democrats during the same time period have won 101 open Republican seats—or 57% of the total party turnovers in the open seats. Republicans did do much better than their overall average in the tidal years of 1972 and 1980—but they did not distinguish themselves in the remainders of the decades of the ’70s and ’80s, including George Bush’s victory year of 1988.

If the incumbency theory is true, we should also see a pattern of GOP success in special elections, with seats opened up in the middle of the election cycle because of an incumbent’s death or resignation. Since 1954, there have been 159 special elections to the House; the net change in party control has been a GOP gain of one. Since Ronald Reagan’s election in 1980, there have been 36 special elections to the House; the GOP has made a net gain of one. In short, we see no evidence here for a GOP tide thwarted simply by incumbent advantages.


Table I. Open House Seats Changing Party, 1954-1990a

Year D->R R->D
1954 2 3
1956 2 4
1958 0 14
1960 6 6
1962 2 3
1964 5 8
1966 4 3
1968 2 4
1970 6 8
1972 9 5
1974 2 13
1976 3 7
1978 8 6
1980 10 1
1982 3 5
1984 5 1
1986 7 8
1988 1 2
1990 0 6


aSource: Vital Statistics on Congress, 1989-1990.True, if there were no incumbent advantages, the race for majority in the House would be considerably tighter; Republicans have more or less kept parity with the Democrats in open seats. But that is like saying that if there were no seven-foot tall players in the National Basketball Association, the Washington Bullets could win the championship. Incumbents could be outlawed from Congress, just as seven-footers could from the NBA, but otherwise, guile and discipline are required to overcome their natural advantages.

Why haven’t Republicans done better in open seats? One reason incumbency theory proponents will point to is gerrymandering—the process has been so rigged to create safe seats for each party that the opposition can’t win even when incumbents leave. It is true that the vast majority of open seats, like the vast majority of incumbents’ seats, do not change party in a given election (a phenomenon that precedes this era of Democratic control.) It is also true that more seats have become safe for both parties in recent decades. But many open seats still do change parties, and if the theory of an underlying Republican majority were true, more of them would change towards the Republicans than the Democrats.

There are other explanations—starting with the parties’ political will and skill. If latter day incumbents are nearly impossible to beat, then political parties have to devise other strategies to win seats in Congress—both targeting and challenging the group of incumbents who are vulnerable for reasons of age, scandal, inattention to constituent concerns or voting record, and laying the groundwork necessary to win a district when an invulnerable incumbent leaves office. This means having an experienced, well-known candidate in place at the right time, with an experienced campaign team that knows the district’s geography and politics.

The latter strategy means viewing congressional elections as investments; often it requires fielding and financing a candidate who cannot win to build experience and name recognition for a second try. Republicans have consistently failed to do so. Consider, for example, the number ofuncontested seats—seats held by one party where the opposition fails to put up even a token challenge. In most cases, obviously, seats go uncontested because incumbents are so strong that a challenge would have little if any hope of succeeding. But incumbents do not live forever; leaving their districts unchallenged means failing to create a congressional campaign organization or cultivate a viable candidate to run when the chance of winning increases.

From 1978 to 1988, on average, 14% of races for the House of Representatives went uncontested. Of these a total of 281 were held by Democrats—while only 84 were held by Republicans. In 1988 alone, 58 Democrats were left uncontested, compared to only 20 Republicans. Most of the uncontested Democrats were Southerners, usually from conservative districts; it is no wonder that Republicans have at best a mediocre record at winning Southern open seats. In 1990, continuing the pattern, Republicans left 26 Democratic House seats in the South without challenges, while Democrats let only 12 Southern Republicans similarly off the hook.

One of the races left uncontested in 1988 was House Speaker Jim Wright, despite the conservative nature of his Fort Worth district and the fact that the House investigation of allegations of ethics violations had begun well before the election filing deadline. When Wright resigned in 1989 and a special election was held to fill the seat, Republicans had neither an experienced candidate with name recognition nor a well-oiled staff in place. As a result, they lost narrowly (51.0% to 49.0%) an election that was a golden opportunity for a major GOP victory, and had much less chance for a victory against the new incumbent, Democrat Pete Geren, in 1990.

Of course, few critics of incumbency advantage are motivated simply by partisan or ideological concerns. To many, incumbents from both parties in Congress have engaged in symbiotic, if not conspiratorial, behavior to harden their silos against any outside challenge. This, congressional critics charge, has so limited turnover in Congress that even the Supreme Soviet has more openness and fresh faces by comparison.

To one who walked the corridors of the Capitol in the late 1960s, the era of the supreme seniority system when silver-haired graybeards were omnipresent, the idea that today’s Congress is geriatric and unchanging by comparison seems laughable. A glance at the House floor in the 101st Congress shows many more young faces than old—a casual impression of turnover that is borne out by the numbers. Turnover in Congress comes in many ways—retirements, runs for other office, death, defeats in primaries—and the turnover in the past two decades has been high.

As House Speaker Tom Foley has pointed out, 92% of the House is new since he arrived in 1965, 82% has changed since 1974, and 61% is new in the past decade. Since 1955, the average number of senior House members—those serving seven terms or more—:has been 139, or 32.0%; the current number is 168, or 38.6% of the House. The number of senior lawmakers was much lower in the mid and late 1970s, when we had high numbers of retirements and a series of elections where incumbent defeats were higher because many voters wanted change in the status quo. The last few elections, 1986, 1988 and 1990, we had the cyclical changes typical of American politics: lower retirements which occur when the retirement pool is diminished; fewer incumbent defeats, which occur when voters opt largely for the status quo. We do not require Constitutional change to expect more retirements and more incumbent defeats in the early or mid-1990s, as the retirement pool gets replenished and voters opt for change.

Certainly, as David Broder has pointed out, turnover does not mean competition, and competition has clearly declined in the House; fewer and fewer House seats are truly, competitively contested from one election to the next. This is a real and vexing problem, that stems more from the difficulties of securing good challengers than anything else. I discuss some of the recruitment problems below; it is worth emphasizing here that a necessary step to bring more competition to House elections is to change campaign finance laws to make it easier for challengers to raise the money necessary to run competitive races in the large and diverse districts we have. The main thrust of most reformers, including Common Cause and the Democratic Party is spending limits on elections. Well meaning or not, spending limits would almost certainly have the opposite effect, limiting the ability of challengers to get their messages across in districts that require sizable sums of money to do so.

Incumbent advantage could be dramatically reduced by forcing incumbents out through a Constitutional Amendment limiting terms of service. To do so would solve a problem, lack of turnover, that doesn’t exist; would cause a huge gap in the institutional memory of the Congress, encouraging historical policy mistakes to repeat themselves; and would give a tremendous increase in clout to congressional staff, lawyers, lobbyists, pundits, bureaucrats, public relations specialists, journalists, and other permanent denizens of the Capitol who are not limited in their service. It would probably bring into Congress a group of hyper-ambitious people viewing the job not, as Amendment proponents believe, as a way-station before returning to normal life in the district, but as a stepping stone to higher office or major money. They would have little stake in the House as an institution, and little incentive to think about the broader and longer-term implications of their actions. Perhaps current members also lack those incentives, but a Constitutional Amendment should have the clear promise of something more.


The evils of partisan manipulation of congressional districts, commonly called gerrymandering, have been a favorite target of political reformers for decades; increasingly they have been joined by Republican Party officials and conservative ideologues like the editors of the Wall Street Journal who view gerrymandering as a corrupt process that has wildly distorted the House away from a representative institution. Republicans and the Journal have been particularly exorcised over the fact that the Democrats’ percentage of seats in the House of Representatives has consistently been higher than their nationwide proportion of votes—evidence in their minds (not to mention their fundraising literature and editorials)—of the unfair bias created by partisan gerrymandering engineered by the Democrats.

There is no question that a clever and ruthless partisan with control over the levers of redistricting can wreak havoc on the other party. Democrats, led by Phil Burton, did so in California after the 1980 census; Republicans returned the favor in Indiana. But there are precious few additional examples that show clear cut and extreme partisan advantages. Indeed, the single largest effect of redistricting around the country tends to be the bipartisan agreements in most states that end up protecting incumbents of both parties from the kinds of colossal shakeups that would create larger turnover.

As for the celebrated gap between votes and seats, its notoriety, along with the broader attack on gerrymandering has stemmed largely from a misunderstanding of the inherent nature of the American political system, along with a lack of understanding of basic differences between Democratic and Republican voters.

The American political system, like the British system, is one of single-member, “first-past-the-post” districts. Each one has a winner, usually the one who captures the most votes (some have thresholds and run-off provisions.) All first-past-the-post systems have a votes/seats distortion built into them: the party that gains a majority or plurality of the votes cast nationwide will almost certainly win a greater proportion of seats.

Why is this the case? Political scientists and mathematicians have written many books quantifying the reasons and the relationships. To be simple and brief, a first-past-the-post system means a lot of “wasted” votes—but fewer for the most successful party. A candidate who wins with 51% of the votes “wastes” 1% of the votes; he or she only needed 50.1% to win. A candidate who loses with 49% of the votes “wastes” 49%; the result would have been the same if the candidate got zero. In a system like ours, each party ends up with a lot of candidates who get 45% of the votes and end up with nothing, and a lot of candidates who get 80% of the votes and win when 50.1% would do. The variations, however, do not exactly cancel each other out. All things being equal, if a party in a two party system captures 52 or 53% of the votes, that party is likely to end up with 55 or 56% of the seats, having been more efficient at translating its support into winning seats.

The skew can be even greater if there is a third party force in an election. Thus, Margaret Thatcher’s Conservatives won a staggering 58% of the seats in the British House of Commons-with barely more than 40% of the votes! (The Wall Street Journal, for some reason, has never questioned her government’s legitimacy on this basis.)

In the United States, through 22 elections since 1946, the average “bonus” for the majority party has been 6.4% more seats than votes. Since the “outrageous” gerrymandering after the 1980 census, the average bonus has been 6.5%—almost exactly the overall post-war average! The only time in this modern era in which the Republicans captured a majority of votes, they also took advantage of the built-in bonus. In 1946, Republicans captured 53.5% of the votes cast nationwide, and received 56.7% of the seats in the House, a bonus of 3.2%.

Why would the GOP bonus have been less? We have other evidence to suggest that Democrats generally will tend to do better than the Republicans in the votes/seats ratio—but not for malevolent reasons of gerrymandering. Rather, there are simple differences in the natural bases of party support. Democrats, not surprisingly, tend to attract votes more from poorer, less well educated and more minority-based constituencies than do Republicans. Their voters turn out less—and their districts, as a result, have lower turnout.

Assume you have two fairly-drawn districts of equal population, one a poor inner-city district that elects a Democrat, the other an affluent suburban district that elects a Republican. In the former district, 150,000 voters turn out; in the latter, 250,000. The Republicans have captured 60% of the votes cast in these two districts, but only 50% of the seats. Is that inappropriate? Not by most standards.

The following table shows the distribution of contested seats won by the Republican and Democratic parties in the 1988 congressional elections by levels of voter turnout. To summarize it, Democrats won 54.5% of their contested seats, or 110 of them, in districts where fewer than 200,000 people voted. Republicans won only 31.6% of their seats, or 49, in such districts. In contrast, Republicans won 35.5% of their seats, 55 in all, in districts where more than 225,000 people voted; Democrats won only 19% of their seats, or 38, in such high turnout places. The differences in party bases, and in resulting direct-based turnouts, alone account for a substantial amount of the differences between seats and votes, leaving the natural bias of the first-past-the-post system to account for less than the standard formulae might suggest. Indeed, given that the average bonus in the 1980s was less than the 44-year norm, one might even argue that the Republicans, not the Democrats, won the redistricting wars of the past decade.


Table II. Contested House Elections in 1988a

Turnout Democratic Winners Republican Winners
Less than 125,000 9 0
125,000-150,000 13 2
150,000-175,000 38 6
175,000-200,000 50 41
200,000-225,000 54 51
225,000-250,000 24 31
250,000-275,000 7 13
More than 275,000 7 11


aSource: Congressional Quarterly’s Weekly Report, May 6, 1989. There were seventy-eight uncontested elections.Bonus or no, the overall Republican problem in failing to win the House is best characterized by another set of numbers. Since 1952—the last time the GOP had a majority of seats in the House—it has averaged only 45.3% of the votes cast nation wide. Except for 1956, when the Republicans captured 48.7% of the votes cast, the party has not exceeded 48% in the past 36 years. The GOP, in other words, has not come close during its entire period of consecutive minority status to winning a majority of votes for the House. During this same era, the Republican Party has averaged 55% of the votes cast for president. A prime explanation for this disparity is the quality of congressional candidates.


In a broad sense, the Republican failure to win more House seats is directly related to a lack of success at candidate recruitment. Over the past several election cycles, the Republican Party has had great difficulty recruiting consistently top-flight candidates to run for the House of Representatives, either to challenge incumbent Democrats or to run in toss-up open seats. The reasons include:

*the vicious cycle. A party that is in the minority lacks the incentives to offer potential candidates that the majority party can bring, including future subcommittee or committee chairmanships, extra staff and other perks of majority status. The longer the GOP has been in the minority, the more remote look its chances of vaulting into the majority, and the more difficult it is to get ambitious people to be lured to run—and as a result, the longer it continues to occupy the minority.

*the partisan culture. While one must be careful about over generalizing, it is nonetheless clear that cultural definitions of career desirability and success vary between the parties. For the Republican Party’s best and brightest, the careers of choice are business, commerce or other professions. Politics, especially in Washington, would not be high on the list. For Democrats, a career in politics and/or government would be a much more acceptable career choice. This difference is partly a matter of:

*philosophy. To a considerable degree, the Republican philosophical and political message for the past several decades has been, to put it simply, “Let’s get government in Washington out of our lives.” To recruit for Congress based on that message means saying to a prospective candidate, “Spend your life in government in Washington.”

*political reforms. Well-meaning, necessary and important though they may be, there is no question that the web of conflict of interest and financial disclosure requirements put into place in the 1970s and 1980s has made it more difficult to recruit people for public service in Washington, especially for elective office. The more complicated one’s life and finances, the more sacrifices are now required to run for and serve in high office, and the more likely one is to be reluctant to bare all for open public scrutiny. While this phenomenon affects both parties, Republicans who tend to turn more to business executives and entrepreneurs, are affected more.

*the GOP Executive Branch. For many of the Republican Party’s best and brightest who have an interest in politics and public service, there is an alternative available to a campaign for Congress—namely, an appointive position in the Executive Branch. An appointment as an assistant secretary or deputy administrator of an agency may require financial sacrifice and Senate confirmation, but it involves more authority than minority membership in Congress, and does not require the fundraising, time commitment and politicking necessary to run a competitive race for Congress—which still may not win. Thus, the potential recruitment pool for Republicans for Congress is shrunk. Ambitious and politically attuned Democrats have no executive alternative; for them Congress is the only game in town.

To be sure, many of these problems for the Republican party are recognized by its top political figures, and steps have been taken to rectify them. Perhaps the most significant move was that of Ed Rollins into the top slot of the National Republican Congressional Committee (NRCC), where he moved aggressively to target vulnerable Democratic incumbents, including those who may require two election cycles to defeat, and to recruit and finance top-flight candidates with name recognition and political skill, pledging to them a commitment over more than one election, if necessary. This is a sharp departure from GOP practices of the past, and the change did not come easily; Rollins, from his first day on the job, was attacked vigorously by many Republicans more comfortable with the easy ways of the past.

While Rollins was initially pledged the time and resources to do the job, he found that the fund-raising for the Republican House campaign effort fell precipitously in 1990, particularly after President Bush abandoned his “no new taxes” pledge. Rollins’ efforts to separate House Republican candidates from the president on the tax issue met with fury in the White House. Although the GOP loss of seats in 1990, at nine, was far less than the norm for mid-term elections, Rollins announced his departure from the post in January 1991, leaving the committee, and its long-term strategy, in limbo.

At the same time, a number of Republicans like Jack Kemp, Vin Weber, Bill Gradison and Newt Gingrich have been trying to find a more positive message for the Republican Party that goes beyond simply getting Washington off our backs and government out of our lives. These efforts have the potential to bring more excitement to the message and more electricity to the rank and file—and more people eager to devote a major share of their lives to implementing it.

Moreover, at least some of the Republican support for the congressional pay raise in 1989 stemmed from the recognition that it would help recruit good candidates for Congress. And there is more openness in Republican ranks to significant changes in the campaign finance laws, even if some of them involve tax credits, i.e., indirect public funds, to foster more competition in congressional elections—and not just through heavy-handed and unrealistic mechanisms designed simply to punish Democrats or hobble incumbents.

These things should and will make a difference.

But there is another, ironic, roadblock for the Republican Party to consider.

Counter-cyclical Politics

In the 131 years of American political history since the Civil War, we have had 80 years with president and both houses of Congress of the same party; 22 years of president and one house of Congress of the same party, and 29 years of president of one party and both houses of Congress of the other (the latter, of course, mostly concentrated in contemporary times.) But though each of these political arrangements, across different eras of modern American history, one phenomenon has been almost universally true: the longer a party holds the White House, the worse it does in Congress.

Every presidential party has left office with fewer congressional party members than when it entered. The Republicans held the presidency from Lincoln in 1860 to Benjamin Harrison through 1884. When Lincoln was inaugurated, his party held 59% of the House of Representatives; after the last election before the GOP departed the White House, 1882, it held 36.3% of the House. The GOP held the White House again from 1896, with McKinley, through 1912, with Taft. It began this era of presidential control with 57.1% of the House; its last election in control left it with 41.3%. Woodrow Wilson, conversely, began his two terms as Democratic president with 66.9% of the House; after his sixth-year election, his party held 43.9%. (In every case, incidentally, the percentage dropped sharply more with the following presidential year election.) The table below shows the figures from Warren Harding to the present.

Of course, part of what is reflected in these numbers is the well-known phenomenon of the mid-term loss for the president’s party: in every off-year election since the Civil War, save one, the president’s party has lost seats in the House. But we also see another two phenomena at work: a president’s party almost never gains back as many seats in his re-election year as it lost in the previous off-year contest, and the losses in the second mid-term election of a two term president are nearly always greater than in the first—the fabled “six-year itch.”

There are many reasons for the mid-term loss and six-year itch cycles. A part of it has to do with differential turnouts and party coalitions in presidential and off-year elections. Turnout surges by an average of about 15 percent in a presidential election; the additional voters, who by definition are occasional and not habitual voters, are attracted by the excitement of the presidential campaign. Not deeply committed to politics or parties, they tend to go in substantial numbers with the trend of the campaign—with the winner. Less likely to pick and choose down the ticket, these voters also vote more often with his party further down the ballot.

In the mid-term election that follows, those voters are absent, and we are left with the regular voters, who tend to be more strongly rooted in their partisanship. The increment of votes for the president’s congressional party that comes with the surge of turnout by occasional voters in the presidential year is gone; the opposition party benefits.


Table III. Presidents and Congressional Seats, 1920-1988a

President Year Percent Number Percent Number
Harding (R) 1920 69.5 301/433 30.2 131
Coolidge (R) 1922* 51.7 225/435 47.1 205
Coolidge (R) 1924 56.9 247/434 42.2 183
Coolidge (R) 1926* 54.5 237/435 44.8 195
Hoover (R) 1928 61.4 267/435 38.4 167
Hoover (R) 1930* 49.2 214/435 50.6 220
Roosevelt (D) 1932 71.8 310/432 27.1 117
Roosevelt (D) 1934* 73.8 319/432 23.8 103
Roosevelt (D) 1936 76.4 331/433 20.5 89
Roosevelt (D) 1938* 60.8 261/429 38.2 164
Roosevelt (D) 1940 62.5 268/435 37.2 162
Roosevelt (D) 1942* 50.7 218/430 48.4 208
Roosevelt (D) 1944 55.8 242/434 43.8 190
Truman (D) 1946* 43.3 188/434 56.4 245
Truman (D) 1948 60.4 268/435 39.3 171
Truman (D) 1950* 53.9 234/434 45.8 199
Eisenhower (R) 1952 51.0 221/433 48.7 211
Eisenhower (R) 1954* 46.7 203/435 53.3 232
Eisenhower (R) 1956 46.2 200/433 53.8 233
Eisenhower (R) 1958* 35.1 153/436 64.9 283
Kennedy (D) 1960 60.2 263/437 39.8 174
Kennedy (D) 1962* 59.3 258/435 40.7 177
Johnson (D) 1964 67.8 295/435 32.2 140
Johnson (D) 1966* 56.9 247/434 43.1 187
Nixon (R) 1968 44.1 192/435 55.9 243
Nixon (R) 1970* 41.5 180/434 58.5 254
Nixon (R) 1972 44.4 192/432 55.3 239
Ford (R) 1974* 33.1 144/435 66.9 291
Carter (D) 1976 67.1 292/435 32.9 134
Carter (D) 1978* 63.7 276/433 36.2 157
Reagan (R) 1980 44.1 192/435 55.9 243
Reagan (R) 1982* 38.0 165/434 62.0 269
Reagan (R) 1984 41.9 182/434 58.1 252
Reagan (R) 1986* 40.7 177/435 59.3 258
Bush (R) 1988 40.2 175/435 59.8 260

(*) indicates off-year election.

aSource: Congressional Quarterly’s Guide to U.S. Elections, Second Edition.

The out-party benefits in other ways as well. Unhappy voters, those who quarrel with the status quo, are more likely to vote than those who are contented and complacent. Voters who have strong and uniform partisan feelings are more likely to turn out than those who feel cross-pressures. The unhappiness with the status quo will be more concentrated in those out of power, and usually grows in intensity in a president’s second term.

At the same time, divisions within the president’s party, often between purists and pragmatists, grow steadily throughout a president’s tenure; the argument in the first term that the president has to compromise his ideology in order to get re-elected does not apply to a second term. The more ideological members of the party, finding continuing pragmatism in the second term, are more likely to be disillusioned and cross-pressured—and thus more likely to sit on their hands—by the sixth year, leading to more partisan differential in turnout, and greater losses for the president’s party in Congress.

Cycles of recruitment of candidates matter here as well. A party out of power has a larger pool of potential recruits to run for office—more people eager to bring about a change, more people with no other outlets for their public service careers. A party that holds the presidency has thousands of executive branch positions to offer its best and brightest people, meaning fewer of them will be available to undertake arduous runs for office. The longer a party is out of the presidency, the more ardent become the potential recruits for the opposition party, and the more the talent pool of the president’s party gets drained by other jobs, ennui and disillusionment. The best GOP recruitment years for Congress in modern times have been 1946, 1952, 1966, and 1980—all times when the Democrats had held the White House for a long time. The best Democratic recruitment years, 1958, 1974, 1982 and 1986, generally fit the same pattern.

There is a simple and more general reason for this phenomenon. When things are going badly, especially in the economy, as they often are in a president’s second and sixth years, the public will blame the president’s party; but when things are going well, as they usually are when an incumbent president gets reelected, voters are inclined to give a pat on the back to all incumbent office seekers. In off years, most presidents and their parties find themselves on the defensive; in presidential re-election years, presidents run by proclaiming, “things are good, let’s keep them that way.”

This cycle has persisted for well over a century; yet, in no other period of partisan dominance of the White House did the opposition party grab a lengthy hammerlock on Congress. What makes the contemporary era unique? The roots lie in the 1930s and the 1950s, the Roosevelt and Eisenhower eras.

The drama of the Great Depression, coming after a decade of uniform Republican control, gave the Democrats a huge margin of power to work with in 1932. The Democrats had captured the House, after the 1929 Crash, in 1930; with the landslide of 1932, their edge increased overwhelmingly, to 72% of the membership. Both Woodrow Wilson and his successor, Warren Harding, had entered the White House with margins in the House of Representatives nearly as large—but they fell back dramatically in the off-year elections of 1914 and 1922, respectively.

1934 was different. For the only time in modern history, a president’s party actually gained seats in the House in the off-year. Democrats moved to 319 seats, or 74% of the total, and expanded to 331 seats, or more than 76% with FDR’s re-election in 1936. (Democrats also held nearly 80% of the Senate after the 1936 election.) From that point on, the typical patterns took over—but the erosion in congressional numbers in the president’s party came from such a huge base that the falloff was limited. Democrats lost 70 seats in Roosevelt’s “six year itch” election of 1938, and lost another 50 seats in his last midterm election, 1942—but they still managed to retain a majority!

When it all finally caught up with the Democratic Party in 1946, after a full 14 years in the White House, their loss of the House was narrow enough that they were able to win control back on the heels of Harry Truman’s comeback victory in 1948. When Eisenhower won the presidency in 1952, he, unlike Roosevelt, had no party majority already in place from the preceding midterm election, and his coattails were too short to build any significant cushion against the coming mid-term loss. Republicans lost the Congress back in 1954, fell back by a huge margin in the six year itch of 1958, and have been in a hole ever since. Every time they start to emerge from that hole, the counter-cycle of presidential politics digs it a little deeper.

Without the large advantages of incumbents, and modern redistricting, the outcomes of the last two decades in party control of Congress might still have been different. But they most certainly would have been different, and more advantageous to the Republicans, if the Democrats had won the presidency more often. (For obvious reasons, however, the Republican Party is unwilling to consider deliberately losing the White House in order to achieve the elusive goal of capturing the House of Representatives.)

Future Prospects

Does this then mean that the Republicans will not capture the Congress until the Democrats achieve a two-term presidency? Not necessarily. The GOP’s commitment to the arduous and difficult tasks of building a candidate base, investing in districts where victory might be years away, and capturing a new philosophical excitement may find a payoff in the near future.

There are other favorable signs. The reapportionment and redistricting of the early 1990s are unlikely to result in a large partisan edge for either party, but by creating a shakeup in district lines for nearly all incumbents in the House, they may create new opportunities for turnover. For many reasons, including generational change, the trauma of new district lines, and changes in campaign rules, we are likely to see more retirements from the House in 1992, leading to more open seat opportunities as well. If effective campaign finance reform passes Congress before then, a system that creates opportunities for challengers may enable more of them to actively contest for seats. If the underlying, rumbling discontent in the electorate about Washington politics-as-usual gets ignited by a catalytic event, such as the savings & loan scandal, the patterns of the past few decades may be broken, and large numbers of incumbents of both parties may find themselves unemployed.

These are all shaky, conditional statements, but they reflect a tangible possibility for a change in the House before the twenty-first century arrives. Would it be desirable? Beyond the obvious partisan considerations, many reformers, including such Democrats as James Sundquist and Lloyd Cutler, believe that divided government is harmful, even pernicious as a pattern. In terms of policy, however, political scientist David Mayhew has found no evidence of any real difference, in quantity or quality of outposts, between periods of unified and divided government in America.

As for the American public, surveys for decades have consistently shown two-thirds of voters indicating a preference for a president of one party and a Congress of the other over a president and Congress of the same party. They are not losing sleep over our partisan institutional arrangements. Americans like checks and balances; they see a split government as another. A Republican president and a Democratic Congress also meshes well with the mix of attitudes Americans have toward government. We like a tough American stance in foreign affairs, demanding respect for America, and a skeptical, green-eyeshade approach to government spending and waste of our tax dollars. A plurality of Americans associate these positions with the Republican party. At the same time, we want to rein in international adventurism, promote peace, take care of the disadvantaged and have a government sensitive to the needs and problems of average working Americans—all areas associated with the Democrats. A Republican president and Democratic Congress enables us, in other words, to have our cake and eat it too.

Nevertheless, the experience of the Senate, which shifted party control for six years in 1980, shows that a change of party can be healthy for both majority and minority—giving each side the perspective of the other, including the responsibility required to be in the majority and the frustration inherent in being in the minority. The Senate today, run by Democrats, operates better, in my judgment, than it would if the GOP had not won a majority a decade ago with Ronald Reagan’s landslide. The House could benefit form the same change in perspective.

At the same time, this lengthy period of divided majorities, unprecedented in American history, has created turmoil in the psyches of both parties, to damaging effect. Schizophrenic government has led to schizophrenic parties. Democrats and Republicans are at the same time swaggering and uncertain, secure and paranoid. Each side is confident in its own hegemonic domain, but thrown off stride by its abject failure to extend its popularity and control to the other’s turf.

Each party is fearful that it will make a mistake and lose its own empire—not just for one term, but for decades; each side is hopeful that it can force the other to make the fatal mistake, and finally capture its rightful, complete majority. The result is passive-aggressive politics, the politics of blame avoidance. Each side is so concerned about avoiding a mistake, and so intent on tarring the other, that taking risks to make better policy is increasingly less prevalent.

Conventional wisdom is that the safer a lawmaker, the greater his or her willingness and ability to take risks or cast unpopular votes. The reality, in the contemporary period, is the opposite. The safer the seats in the House, the more skittish the members. Changes that would bring more competition to House elections would help to break this pattern, even if they did not per se result in a different party majority. Most desirable would be campaign finance reform that would open up new sources of money to candidates and make it easier to run competitive campaigns, starting with generous tax credits and matching funds for individual in-state contributions, and some “seed money” mechanism to pay for start-up costs for fledgling challengers.

Campaign finance reform of this sort is desirable, as is more real competitiveness in House elections. But there is no case to be made for significant Constitutional reform. Like it or not, the Congress of today is a legitimate reflection of the electorate wishes, not an artificial and illegitimate construct of a sneaky Democratic Party playing by its own set of rules. To win the House and break the pattern, Republicans need to change their outlook and their campaign approaches—not alter the fundamental Constitutional rules we have utilized for two centuries.

Republicans tried the latter strategy once before; frustrated by Democratic dominance of the presidency under Roosevelt, the GOP pushed through the 22nd Amendment to the Constitution, limiting presidents to two terms. Ronald Reagan, among others, has forcefully underscored the self-defeating folly of that approach. It would be far better to mobilize to repeal the 22nd Amendment than to repeat the mistake with Congress just because political currents have changed.


About the Author

Norman J. Ornstein is a Resident Scholar at the American Enterprise Institute for Public Policy Research in Washington, D.C.. He is a graduate of the University of Minnesota and earned both his Master of Arts degree and Ph.D. from the University of Michigan. He has been a Fellow at the Center for Advanced Studies in the Behavioral Sciences at Stanford University, Phi Beta Kappa Visiting Scholar, Fortieth Anniversary Fulbright Distinguished Fellow and a professor of politics at the Catholic University of America. Dr. Ornstein has worked on Capitol Hill in a number of capacities. He has been American Political Science Association Congressional Fellow, a consultant to the Commission on the Operation of the U.S. Senate and a staff member/director of the Senate Committee to Study the Committee System, which reorganized the Senate. His books include: The People, Press & Politics; The American Elections of 1982: The New Congress; Interest Groups, Lobbying and Policymaking; and Vital Statistics on Congress. Also, he has written articles and reviews in numerous national and international newspapers and magazines. Dr. Ornstein served as political editor for the public television series The Lawmakers and as a political contributor to the MacNeil/Lehrer News Hour. He has appeared frequently on ABC’s Nightline, CBS’s Face the Nation, and all three networks’ news programs.