W. B. Allen

© W. B. Allen


Long meditation on the character of political parties in the United States persuades me at least that the two-party model is erroneous.  That observers have persisted for so long in a possibly erroneous view is not wonderful.  The mind’s eye focuses naturally upon the system of two-party competition in the United States precisely because the absence of anything more than two parties is so dramatically striking and seems unnatural.  Accordingly, any attempt to explain the system of two-party competition will be at as great, if not greater, pains to explain the absence of a plurality of parties.  That focus in turn will obscure the reality that what counts for characterizing the regime is far more the system of governing than the system of competition for office (although the two are not entirely unrelated).

Every credible observer who has commented on the system of governing in the United States has noted the stable pattern of party governance which, once installed, lasts for a prolonged period of time, for one or two generations or more, beginning with the Jeffersonian Democratic-Republican Party of 1800.  This pattern contrasts quite markedly with most constitutional polities we know, where oscillations between left-labor and conservative style governments occur with much greater frequency.

For the past forty years the political landscape in the United States presented a still more confusing prospect on account of the appearance of “divided government” as the dominant mode.  When the chief executive is of one party (Republican this past generation) while the legislature is controlled by the other party (Democrat in this case), we call this divided government.

That characterization perhaps oversimplifies the case, however, for it abstracts from the complex structure of political offices that run throughout this elaborate federal mechanism.  In that view, the system overall has been so thoroughly dominated by one party (Democrat) that it seems unreasonable to call it “divided government” solely on account of the presidency being held by Republicans.  That aberration clearly demands an explanation—given the importance of the office—but it does not alter the apparent reality of a system of governing dominated almost exclusively by one political party for sixty years.  Presenting this conception colloquially, if somewhat raunchily, I have observed that, looking at political power from the bottom up (as one ought to do), long before one gets to look upon the smiling face of a Republican president, one must smile upon countless Democrat bottoms.

A better model of analysis of party in America, therefore, may well be the model of the one-party state—allowing in this case for an intervenient but not truly competitive second party.  On this model a two-level dynamic would have to be explained.  First, of course, one would account for the emergence of the dominant party in some defining or aligning political event.  Further, one would expect it to sustain its dominance with political inertia, at a minimum, in the absence of any newly emerging defining political event.  The underlying premise here is that the public mind in this special circumstance resists shifting from the established dominant pattern and therefore must be “overcome” to produce a defining political moment.

In the second place the observer would have to account for the dynamic whereby a second and subordinate party persists despite its being fated never (for all practical purposes) to constitute the regulative principle of the regime.  The underlying premise here is that the interventions in an otherwise stable pattern permitted to the secondary party suffice for all the purposes for which that party has been organized.  That implies further that, although the two parties may well compete for office, they do not in fact compete over governing.  The one-party governance of the dominant party secures the interests of the secondary party no less than its own.

The foregoing speculations would ground a proper reappraisal of the character of party government in the United States.  That reappraisal in turn would enable clarity as to the future of political parties in the United States.  I shall forego the all important middle step, however, and proceed by means of a brief case analysis directly to the future of political parties.  I may do so, first, owing to the serendipity that my speculations, if read in conjunction with George Washington’s “Farewell Address,” will appear less innovative and already connected with a ready appraisal of the regime.  I may secondly proceed owing to the adventitious circumstance that the return of the decennial census at this historical juncture provides the case analysis that directly makes the point.

Washington observed that, in a government constituted so as truly to establish the people’s right to govern, “all combinations and Associations, under whatever plausible character, with the real design to direct, control, counteract, or awe the regular deliberations and action of the Constituted authorities are destructive of this fundamental principle and of fatal tendency.”

Consequently, only a party that begins by affirming the sufficiency of constituted authority—the party in power—can be accepted as a patriot party, so long as the principle of the power and right of the people obtains.  In these terms the secondary party is not a loyal opposition but rather a little brother longing to imitate his elder.  That explains, accordingly, why representatives of the secondary party may be elected from time to time and even to the highest office, without altering the character of the government itself.

The application of the 1990 census to our model calls for a more elaborate background statement before its implications will become plain.  The immediate issue raised by the census is of course that of legislative re-districting, which involves in the most intense and intricate manner the prospects of party competition and cooperation for the following decade.  (Observe that no decennial census in our history has ever served as the proximate cause of any political realignment!)

We must distinguish 1990 redistricting from all prior redistrictings, however, for it involves dimensions newly or at least for the first time fully developed.  A decade ago I anticipated redistricting with an essay on “proportional representation,”[1] for many speculated then on the possible advantages of substituting proportional representation for our traditional practices.  I discouraged the view that proportional representation was more democratic, pointing out that it marginalizes minorities by giving them a small voice once and for all.

This is the great flaw.

The proponents of proportional representation deny the idea of a whole community whose benefit the representative labors to assure.  Instead, they assert that democratic society must be a permanent warfare of implacable enemies—groups of persons whose interests are mutually exclusive.  They reject the idea that all citizens are represented by all representatives.

Proportional representation allows the voter to remain obstinately attached to “his” candidate no matter the political reality.  This can be seen clearly in the constant stalemate condition of many parliamentary systems.

The essence of an electoral system which produces only one winner is to induce the winner to be as careful about the interests of citizens who vote against him as of citizens who vote for him.

While more than a few observers in 1980 dreamed about a redistricting or Gerrymander designed to imitate certain kinds of proportional representation, they lacked both the fully developed manipulative techniques and a fully developed legal foundation to sustain their efforts.

In the decade since, both the law and the technical means have evolved so as to render a simulation of proportional representation a real possibility.  Additionally, the political climate has fostered an indulgent attitude concerning such efforts.  These recent developments have focused, not on political parties, but on racial and ethnic minorities.  Nevertheless, it is in the nature of things that instrumentalities invented for the purpose of permitting minority interests to drive redistricting must inevitably play into the hands of those entities which most essentially perform and live by redistricting—that is, political parties.  This observation clearly leaps from the page of Wilson’s textbook definition of the Gerrymander:

drawing a district boundary in some bizarre or unusual manner to make it easy for the candidate of one party to win election in that district.  In a state entitled to ten representatives where half the voters are Democrats and half are Republicans [he could as easily have said today “anglos and blacks”], district lines could be drawn so that eight districts would have a slight majority of citizens from one party and two districts would have lopsided majorities from the other.  Thus it can be made easy for one party to win eight of the ten seats.[2]

The process is not quite so easy as Wilson suggests, as I will show below.  Nevertheless, his point is made:  in the absence of any principle to constrain the process of redistricting, the process may be subjected to limitless contortions in the pursuit of partisan advantage.

Redistricting begins with the constitutional mandate to enumerate once every ten years all the persons in the United States who shall be divided within each state into units no smaller than 30,000 each, and to each of which units a representative is to be assigned, having been chosen by electors qualified for the purpose.

Apart from the fact that all persons are counted and represented, and that provision for the appointment of electors of representatives is deliberately separated from the redistricting provision, the Constitution does not expressly limit how redistricting must be done.  That the districts shall be as nearly equal as feasible (for there shall be one for a specified number of units of equal dimension) is almost the only additional principle that can be honestly derived by implication.  A further, indirectly derived principle builds upon the express restraint of containing representational units within states as opposed to following abstract district lines; one may thence infer a constitutional regard or preference for community integrity as an element of redistricting.

Upon this clear but nevertheless expansive canvas courts, politicians, and minority advocates have in the past decade painted a mosaic of standards and practices that permit almost any district configuration that will conform to a stated public policy interest, save that the rule of equal numbers yet prevails.  This particular evolution may be traced quite clearly, for the decade opened with a fairly resounding opinion from the United States Supreme Court:

There is no right to have minority candidates elected; but only to give them equal access to the electoral process.  Absent a clear demonstration of the proscribed discriminatory purpose in keeping minorities from participating in the electoral process, there is no Fifteenth Amendment violation.  These same arguments prohibit a finding of a denial of equal protection under the Fourteenth Amendment.  Disproportionate effects are not alone sufficient to demonstrate an improper motivation and a denial of equal protection.  There is no right to proportional representation.[3]

Just two years later (1982) the Congress “overturned” this Court opinion with its extension of the Voting Rights Act of 1965, in which an “effects test” was added to Section 2.  The case is further interesting because this legislation was passed by the dominant Democrat Party over the strenuous objections of the Republican President and most of the Republican members of Congress.

In the intervening eight years (to 1990) numerous court cases built upon the new legislation, culminating in a successful Department of Justice challenge to the 1980 redistricting of the Los Angeles County, California Board of Supervisors.  Thus, the Republican Reagan-Bush Administration employed the 1982 amendments on behalf of an explicitly hispanic supervisorial district.  In addition to the effect of the new law, new technologies were brought to bear enabling the Gerrymander to avert a far more serious pitfall.

In 1977 the Supreme Court resolved a case brought by a community of Hasidic Jews against a redistricting plan that divided their community in half, while concentrating black voters in New York so as to enhance black representation.[4]  The Court judged by plurality that the intent of the Voting Rights Act of 1965 was fulfilled in the plan, and thus the Jewish community could not obtain relief.

In the Los Angeles County case it were impossible to concentrate hispanic voters, in anything resembling a community, without splitting a black community.  How, then, was this paradox averted?

The Federal District Court and the many interveners took advantage of the high assimilation rate—and consequent wide dispersal—of hispanics, on the one hand, and greatly refined computer algorisms, on the other hand.  The result:  a “district” corrals  hispanic voters over far-flung lines even while not defining very much of a hispanic community.

Computer driven mapping via census tracts allows districts to be constructed purely in terms of defined demographic characteristics, with or without regard for community contiguity or defined spheres of interest.  This is the first time we have enjoyed this capability practically without limitation.

The 1990 census opens a new era, therefore, and the redistricting to follow will largely define the future shape of America’s parties—not so much in terms of their relative numbers (which are unlikely to change drastically) as in terms of the principles they will advance.  For the new capabilities both legally and technically tempt party technicians and party leaders in a way they have never been tempted before.

This applies especially to the Republican Party, our case analysis, and which is invited hereby to fight Democrats through race-based Gerrymandering precisely because, by doing so, it may at once appeal to certain Democrat constituencies and seek to limit the number of Democrat districts.  For every lop-sided black district guaranteeing a black representative in Congress, for example, Republicans will know that they reduce the effect of Democrat voters in other districts.  How will the parties respond to this temptation?  If past expressions can indicate the future, they will respond not at all well.  Let us consider the general case of the Republican Party as an example.

Republicans err greatly when discussing reapportionment, and their error creates a dilemma for those projecting the party’s future. In particular, the error suggests that the Republican Party may at length have become an entirely unsuitable vehicle for transmitting democratic principles.  The immediate reason is that the architects of Republican strategies are themselves fundamentally undemocratic.  To discover this fact one needs to pay close attention to their presentations on the matter of reapportionment, where one will discover undemocratic sentiments that in fact run through a host of Republican reactions, including the reaction to policies such as affirmative action.

Sticking with reapportionment for now, just reconstruct mentally the typical Republican demonstration.  It runs along the following lines:  First a number is cited, usually the total number of votes cast statewide or nationally in legislative races.  Next, it is noted that the voters who cast those votes are split almost evenly between Democrats and Republicans.  This would produce a bar that looks like the following:




The next step is to cite a new number, usually the distribution of legislative seats following the election in which the vote tally was made. That generally produces a bar that looks like the following:




The disparity between the proportion of Republican voters and the proportion of Republican legislators is obvious and the Republican politicos are quick to seize upon it as evidence of Democrat “gerrymandering” that disfranchises Republican voters.  The theory is that these voters are losing their representation to high-handed and disreputable practices.

I say this argument reveals a deep and perhaps reflexively undemocratic streak in Republican politicos.  The reason is straightforward and plain to see.  Start with a third figure, the number of constituents represented by each congressman.  One will find roughly five hundred thousand in each of 435 districts, and among that five hundred thousand constituents, one will discover a partisan background very close to that of the legislative graph, assuming one identifies not merely registered partisans but also family members and neighbors as indicative of a district’s partisan leaning.  The reason the bar for this number is much closer to the legislative breakdown is, quite simply, that there are more Democrats than Republicans in the country.

Why, then, do Republicans contend that Democrat “gerrymandering” costs them legislative seats?  They correctly point out that, whatever the demographic differences, the actual votes cast in elections are much more nearly equal between Democrats and Republicans.  They contend, further, that representatives ought to be apportioned in relation to the proportion of votes cast.  But this argument betrays them as uncomfortable with the rule of the Constitution, for the Constitution ordains that we count representatives in relation to the number, not of voters but, of persons.  In short, congressmen do not represent the voters who vote for them; they represent the people who live in their districts.  The voter himself is only a representative, an elector to whom the Constitution has delegated the task of selecting the person who will serve as the representative of all.

The political reality this produces is largely determined by the demographic differences between mainly Democrat and mainly Republican districts.  The fact is, it will take two to three times as many votes to elect a Republican representative, simply because the demographics will provide a much higher number of eligible voters and also a much higher voter turnout in Republican districts.  Democrat districts will tend to be younger (more under age citizens), have a higher proportion of ineligible immigrants, and also have a lower voter turnout rate.  Thus, two districts of nominally equal size could produce results like the following:

District Democrat—

          Candidate Democrat   75,000

          Candidate Republican 40,000


District Republican—

          Candidate Democrat    60,000

          Candidate Republican 150,000


Now add a third, evenly matched district, tending only slightly to the Democrat:

Even District—

          Candidate Democrat    90,000

          Candidate Republican  85,000


In this hypothetical election we will see 500,000 votes cast, with Republicans receiving 275,000 votes to just 225,00 for the Democrats—a 50,000 vote Republican margin that nevertheless leaves exactly two Democrats and only one Republican equally elected to office.

I maintain that to protest this outcome is effectively to seek to deny to the million persons represented by the Democrats, and especially those in the heavily Democrat district, their constitutionally guaranteed representation.  That is the undemocratic implication of Republican complaints.  They seek to substitute representation of voters for the representation of persons the founders thought necessary.  It will not take much analysis to show that, representing voters instead of persons would effectively build class and income distinctions into our politics that are inimical to the founding design.  Nevertheless that is exactly the end Republican politicos aim at when they hurl general charges of “gerrymandering” at a political system in which they actually fare quite well considering their general failure to challenge Democrat agendas among the persons most responsive to those agendas.[5]

Returning to the specific question of race-based Gerrymanders, it should be plain that Republicans can muster no principled resistance to such practices in light of their general characterization of the redistricting process.  Further, inasmuch as race-based Gerrymanders are only a specific application of the very process Republicans have aimed to install in general, it is most unlikely that such Gerrymanders will appear to party technicians so offensive as indeed they are.

If the Party’s future, in turn, will be built on easy recourse to race as an instrumentality in the competition for office, building on the very laws and policies successfully installed by the dominant party, it must follow that Republicans will by their actions ratify and enracinate those laws and policies.  To that extent, they will contribute to perpetuate the governance of the dominant party.  We may conclude, accordingly, that there will be for the Republican Party no future apart from a continued pattern of subordinate interventions in the otherwise stable Democrat governing of the United States.

[1]  Technically, proportionality does not depend on redistricting and operates rather by means of socially established party identifiers.  A group ordinarily qualifies to receive its proportionate share of total representation—no matter how the districts are drawn—by qualifying the group in the form of a political party for a place on the ballot.  Minimum thresholds of membership and prior political participation are often prescribed as the terms of qualification.  In the United States, of course, such an approach would require no insignificant change in the Constitution.  Accordingly, folk who seek a similar result, but must rely on ordinary legislation for the purpose, can at best simulate proportional representation, providing, for example, that 435 districts for the House of Representatives be drawn so as to elect in assigned proportions persons from different groups.


[2] James Q. Wilson, American Government: Institutions and Policies (Lexington, MA: D.C. Heath, 1980).

[3] City of Mobile v. Bolden, 100 S. Ct. 149 (1980).

[4] United Jewish Organizations v. Carey, 430 U.S. 144 (1977).

[5]  A similar demonstration is possible regarding the tendency to brand the concerns of the religious right as “social issues.”  The fact is, the issues germane to the religious right—such as excessive state control over religion or even abortion—are in reality individual rights issues and as such closely allied with the democratic ethos underlying the polity.  Yet Republican political operatives have defined these issues in such a way as to deny them mainstream status, just as they have effectively defined away the underlying democratic dynamic operating in legislative reapportionment.  These trends raise legitimate fears that the Republican Party may have become a party against the people.