You are using an outdated browser.
Please upgrade your browser
and improve your visit to our site.
Skip Navigation

What Is Representative Government?

In a recent radio interview, New York Governor Nelson Rockefeller was being pressed to explain why he refused to call a special session of the legislature to consider revision of the state’s inequitable system of apportionment. As the relentless questioners poked pins into the various defenses of Rockefeller, the Governor finally turned on his assailants. “But what would be your basis for apportionment?” he asked. “Have you got a formula?”

The answer to the Governor’s question is presently being sought in about half the states of the union where courts, legislatures, constitutional conventions and the electorate are discovering that it is one thing to agree on the principle of more genuinely representative government and another to set workable standards for reapportioning our legislative assemblies.

In the brief three months since the Supreme Court (in Baker v Carr) announced its readiness to invade the “political thicket” of reapportionment, there has been court action in 22 states. Further suits are being readied. Before the process is finished, virtually every state in the union will be affected.

The way in which the search for a “formula” proceeds varies from state to state. In the “parent” Tennessee case, the Supreme Court held that legislative apportionment was based on “invidious discrimination,” in violation of the Fourteenth Amendment and its guarantee of “equal protection of the law.” The case was remanded to the US District Court in Tennessee for corrective action.

The federal court decided to give the Tennessee legislature a chance to straighten itself out. The legislature came in with a new apportionment. The court found it unsatisfactory, but gave the legislature a second chance to correct the inequities by June, 1963, while threatening to reopen the case on failure to act.

In the Tennessee case, the court was not immediately confronted with the problem of a proper formula for apportionment. The state constitution provides that both houses be based on population. Court intervention was required to force the legislature to obey its own state constitution, one that it had whimsically ignored for half a century.

The knotty problem of an apportionment formula was first handled in the Georgia case, where the Supreme Court decision has so far had its most immediate and revolutionary impact. In this state, the “county unit” system gave a rural minority control not only over the state legislature but also over the nomination of candidates for state-wide office and even for the House of Representatives. On April 28, a three-judge federal court struck down the use of the existing system in the primaries. On May 25, another three-man court (with two of the same judges presiding) ordered a reapportionment of the Georgia House. Both these decisions sought to establish “formulas.”

The first of the two decisions allowed the county unit to stand PROVIDED — provided there was “no discrimination in excess of the disparity that exists against any state in the most recent Electoral College allocation, or under the equal proportions formula for representation of the several states in Congress.”

One part of the formula is perfectly clear: the section using Congress as a measure. The “equal proportions” formula is a mathematical equation that divvies up the seats in the House of Representatives among the states as exactly in accordance with population as can be hoped without dividing a Congressman into fractions.

The other part of the formula, modeled after the Electoral College, is somewhat more ambiguous. It could mean that the voters of one county shall have five times as much voting strength as voters in another; or it could mean that no county shall have a per capita voting strength more than 15 percent below the average. Both interpretations require explanation.

Under the most recent allocation of electoral votes, based on the i960 census Alaska has three electoral votes for a population of about 225,000: one elector for each 75,000 people. California has 40 electoral votes for a population of about 15,700,000: one elector for every 393,000 people. Each Alaskan has an advantage of five to one over each Californian.

The average number of people per elector, nationally, is some 333,000. In California, as noted, there is one elector for every 393,000. Every Californian, consequently, will have about 85 percent as much influence on the outcome of the next Presidential election as the average person nationally.

In either event, the populous states (or counties) have a compensating advantage; namely, the fact that the biggest votes are cast en bloc. The fear of the bloc vote from the populous counties may have been responsible for persuading the Georgia State Democratic Committee on June 28 to hold a primary by popular vote. The Committee was also warned by Governor Vandiver that any other system might invite further court intervention. As a result, Georgia will hold a Democratic Party primary this year on a popular basis for the first time in this century.

In regard to the state legislature, the federal three-man court found that its composition was “irrational” and “arbitrary.” Senators representing only 21.4 percent of the population are a majority. In line with the general trend to let the legislature itself have the first crack at repairing its own damage, the court gave the legislature until January 1963, to “meet constitutional standards.” Failing proper reapportionment, the court threatened such total but indefinite action “as is necessary and feasible to accord plaintiffs their rights.”

The court did not feel that it had to order both houses to be constituted according to population. This question was left open, with an order to reapportion at least one house, while noting that it is “not necessary … at this time to make a final determination” on what happens to the second house.

The Michigan Case

While the Georgia formula is based on federal analogy (Congress and Electoral College), the state of Michigan decided to improvise an original recipe by mixing four parts of people with one part of land and water to cook up the Senate seats. This formula comes from the need for a balanced political diet in the Michigan GOP.

The Michigan Senate was not a product of whimsy. The districts were apportioned in 1952, on a fixed basis, through a measure submitted to and approved in referendum by the voters of Michigan. Despite this firm legal basis, Gus Scholle, President of the state AFLCIO, challenged the apportionment in the state Supreme Court, where the case was thrown out because the state judiciary refused to enter the old “political thicket.” Scholle took the case to the federal Supreme Court, which remanded the case on April 23 to the state court for action.

Just what action the state Supreme Court is to take, no one quite knows. The Supreme Court “studiously avoided” guide lines for the state, as Justice Harlan pointed out in his lone dissenting vote. The Michigan court, argued Harlan, could do only one of two things: either reaffirm its original decision to do nothing; or invalidate the Senate apportionment which would mean that a state could not constitute a Senate with “numerical voting inequalities, even though that course reflects the express will of the people.”

The real effect of the Supreme Court decision was its psychological impact on the Michigan Constitutional Convention then in session. The GOP, out of regard for George Romney, its Presidential hopeful, had to keep face with the urban liberals of the state while keeping peace with the rural conservatives in his party. The convention itself reflected the conservative bent of the malapportioned legislature. The Democratic opposition, heavily unionist, had made reapportionment a major issue over several years. Above all else, the convention realized that it was under the scrutiny of the courts and that any plan it proposed would not be sacrosanct just because it was enacted by a convention. The resultant plan claims that it would represent people and area at the rate of 80 to 20 percent.

In practice, the new Michigan system is not an 80-20 system at all, despite appearances. Keewenau county, with 2,417 people and 27,500 acres of inland water and 348,200 acres of land, would have a representation equivalent to 22,141 people. In other words, as a critic of the plan, James K. Pollack, argues, “Keewenau county’s representation is based 89.1 percent on land and water areas and 10.9 percent on its human habitation.” The combined areas of national and state forests and inland waterways (not counting people living there) would have a voting worth equal to almost half a million Michiganders.

Whether this system of apportionment will stand up in the Supreme Court is still undetermined. But undoubtedly the Court will view this system as it must evolve through the years when the mounting population of Michigan will steadily enhance the voting power of soil over souls. At present, 19.05 acres equal one human being but 20 years from now 9 or 10 acres of mud may equal one man.

The Michigan case may in itself become a “parent” case, similar to the Tennessee case. The latter applies to states where legislatures violate their own constitution; the former applies to states where legislatures, or some other apportioning agency, may be obeying their own constitution. In this respect, Scholle v Hare (the Michigan decision) may be controlling in a present New York suit, brought by Station WMCA. There are many striking parallels between the Michigan and New York cases—both legally and politically. In New York, as in Michigan, both houses are apportioned according to constitutional formulas. In both states, the legislatures, to use Al Smith’s descriptive phrase for New York, are “constitutionally Republican,” with Democrats repeatedly winning a majority of the popular vote and the Republicans winning a majority of the seats. In both states, prominent Republican aspirants for the Presidency are on the spot. In both states, the impact on the federal House of Representatives could be monumental, because GOP state legislatures have designed Congressional districts in New York and Michigan to add almost a dozen unwarranted votes to Republican strength “on the Hill.” Loss of GOP power in these states would mean loss of seats in Washington.

Governor Rockefeller has refused to call a special session of the state legislature to consider reapportionment on the ground that to do so this year, when Governor and legislature are up for election, would make the issue too “political.” As a consequence, his very refusal is becoming a key issue in the state campaign, with the Democrats “for” and the Republicans “against” revision.


The WMCA case has become a further embarrassment, legally and constitutionally, to New York Republicans. On June 11, the US Supreme Court ordered a three-man federal court to reconsider this case (dismissed in January for the usual reasons) in the light of Baker v Carr. The same day. Rockefeller released a letter to the Citizens Union in which the Governor leaned on the original finding of the three-man tribunal. The unhappy coincidence of dates caused Mayor Wagner to comment that Rockefeller’s comments were both “ill-fated and ill-timed.”

Still undeterred. Rockefeller took to the air on June 17 to point up the complexities of reapportionment. “I have not seen anybody,” he said, “a lot of criticism—but nobody has come forward with a formula.” In reply. Station WMCA, through its president, R. Peter Strauss, has been circulating the summary of a report prepared by Dr. Ruth Silva of Pennsylvania State in her role as consultant to the Peck Commission on constitutional revision, in which are spelled out five possible formulas for reapportioning the Senate and four more for the Assembly. WMCA also charges that this report was “lost” precisely because it was so clear in its attack on the present system and in its proposals for a future plan. Although many of Dr. Silva’s recommendations are for New York only, her report contains two recommendations that have general interest and applicability. One is to wipe out county lines in creating districts; the other to “weight” the legislators’ vote. Both are an attempt to get “maximum equality of representation.”

The present use of counties as the basis for representation is one of the prime obstacles to establishing districts of equal population, because counties vary greatly in population, yet insist on separate county representation. In political folklore, counties are compared with states. Just as the federal government has two houses, one based on population and the other on political unit, so too it is argued should states have two chambers, one based on people, the other on counties.

This philosophy was given recent expression by Anne Arundel Circuit Court Judge O. Bowie Duckett this May 24 when he ordered reapportionment of one Maryland house on population. “I feel,” he said, “that one of the houses can be Constitutionally based upon area and geographical location regardless of population or eligible voters. Such an arrangement protects the minorities. It prevents hasty, although popular, legislation at the time. It is based upon history and reason and helps to protect the republican form of government guaranteed to the states by Article 4 of Section 4 of the United States Constitution. It preserves the checks and balances in the state government, which has worked so well under the federal. Moreover, there would be little advantage in a bicameral legislature, if the composition and qualifications of the members were similar.”

Opponents of this “federal plan” are generally agreed that to identify counties with states is to confuse the creator with its creation, and historic necessity with democratic principle. The 13 original states created the United States; the counties were created by the states. The states preceded the federal government; the counties were sub-divisions of the states. The federal government can not whimsically abolish, divide, merge the separate states; but the states have done and can do all these things to counties. A group in control of the US Senate could not eternally perpetuate its control by wiping out or merging a few states; but a state could do this with its counties. “The United States Senate is both irrelevant and improper as a model for representation within a state,” concludes Professor Paul T. David, University of Virginia, because “a state is not a federal union of sovereign counties.”

The federal model was in its origin part of a great compromise, a necessary evil, made historically imperative as part of the price of union. To elevate this onetime necessity into a current principle is to substitute analogy for argument, “with quite deliberate intent,” argues Professor David, “as a facile means of practicing an invidious discrimination against city people.”

The strongest argument against basing one house on space rather than people, however, flows from the practical consequences of such a system. “Countyism” allows a minority, normally a rural-based and conservative minority, to control at least one house. This gives this minority a veto power in state government, a veto exercised normally on behalf of static, passive government against creative, active government. In order for “metropolitan 20th Century America” to enact legislation responsive to its needs, the 69 percent of our population living in urban areas requires a potential majority in both houses. But to block any such action, a small-town i9th-Century-minded minority need control only one house. Since the distinguishing political feature of our state legislatures is their inclination toward reaction by inaction the control of one house is almost as bad as the control of both houses by conservative minorities.

All this does not mean that there is not good reason for establishing districts that give voice to sections of the citizenry with distinct geography-based interests, attitudes, traditions and forms. Indeed, existing county lines may be the most convenient and appropriate units. To recognize meaningful regionalism and grant it representation, however, does not require an inflation of power for these districts that would convert somewhat artificial, often archaic, administrative county units into the dominant source of political influence. Let the counties have their voice; but let each voice be no louder than the number of people for whom it speaks!

Bicameral or Bimoral?

Both houses can be based on population without surrendering the positive values of a bicameral system, with its checks and balances. These virtues can be preserved by making one house more numerous and the other less; by electing men to one house from smaller, perhaps more homogeneous districts, while electing men to the other house from larger, more heterogeneous districts; by electing one house for two years and the other for six years, with a full turn-over periodically in one and a partial turn-over in the other. Our legislatures can be bicameral without being bimoral.

Dr. Silva’s second proposal suggests that “a fixed number of Assemblymen might be assigned to each county without regard to population, but the Assemblyman’s vote in the Assembly might then be weighted according to the population of the district he represents.” A similar proposal was made at the Michigan convention to offset the 80-20 area-population plan. Dr. Robert H. Engle, voicing a mounting opinion among political scientists, describes the plan as follows: “Each representative’s vote on legislative decisions of the legislature be the same as (weighted by) the vote he received when he was elected. This would be new in legislatures but it has long been done in business when corporation shareholders vote the number of shares they own or hold proxies for; in labor union federations delegates vote the membership of their constituent union; and in Illinois, by state law, political party committeemen vote the number of party ballots in the preceding primary. Electronic roll-calls, now installed in over half the states with adjustments, can count 100,000 votes almost as quickly as one. This would make every elector in the state equal in representation to every other elector regardless of the size or location of his district or the imperfections of the district apportionment.”

The chief virtue of the “weighting” plan for legislative reapportionment is its simplicity:

1. County lines could be used, with each county represented by one or more persons, but with each person casting a vote weighted by the number of people he represents.

2. By basing the “weight” on votes cast, rather than population, it is possible to keep an up-to-date apportionment, without waiting for the decennial census.

3. Since the system does not call for drawing any district lines at all (existing counties are used), the entire process could be placed in the hands of a clerk who has mastered fifth grade math, and who could be ordered by a state or federal court to do his work correctly and on time.

There are, of course, many variations on the weighting theme. Two men could be elected from each district—the top candidate and the runner-up—with each of them casting a vote proportionate to the votes cast for the candidates. Another device would base a legislator’s weighting on the number of votes cast for him minus the number of votes cast for the runner up. By still another system, the percentage of the total vote cast for the winning candidate could be divided into population to provide a mixed voting-population factor in weighting.

By using fixed county lines, this system would eliminate the bothersome problem of creating districts that are compact and contiguous, since these counties generally are. The state legislatures could, of course, decide to carve up their own sacred cow to create new county boundaries for partisan appetites. But should state legislatures do so, new criteria could be established to clip the claws of the gerrymander. Ernest C. Roeck, Jr., of Rutgers University, has made an initial exploration into standards for measuring “compactness,” by the simple expedient of drawing a circle around the outermost points of a district and then measuring the ratio between the area of the district and of the enclosing circle. By fixing a figure on permissible ratio, a state could inhibit the gerrymander.

Actually if one of the “weighting plans” is used—namely, basing the weight on the vote of the winner minus the vote of the loser—a gerrymander would be completely useless. The only drawback on this proposal seems to be the fact that it appears too true to be good.

New formulas will undoubtedly be forthcoming. The National Municipal League is working on its model clause for state constitutions. The Twentieth Century Fund summoned a conference of scholars in New York to prepare a series of research monographs and recommendations. The Advisory Commission on Inter-governmental Relations—an official body composed of spokesmen from all the major levels and branches of government—instructed its staff this spring to prepare a report and proposals.

There will be many answers on what to do. The next question will be how to do it.

The state legislatures are not objective instruments for this mission. They have proved it for decades and are continuing to do so. In Georgia, following Baker v Carr, the legislature under gubernatorial lash, rushed to reapportionment to head off federal court intervention. The result was unsatisfactory and upset by the court that set down its own standards. In Tennessee, following court order, the legislature reapportioned unsatisfactorily and was given a reprieve until next year. In Wisconsin, the order of the court allowed the legislature until 5 p.m. on Monday of the following week to reapportion; the bill became a political football passed by a Republican legislature and vetoed by a Democratic Governor. On July 3 the federal court appointed a special master to draft a reapportionment plan. John A. Hannah, chairman of the Michigan Convention and head of the state university, noted that to allow legislatures to apportion themselves is like allowing students to grade their own papers.

State conventions are not much better at doing this job than are the legislatures. In many states, it is the legislature that must call the convention. In most states, the convention, in its apportionment, mimics the legislature. In a state like New York, the voters are allowed once every 20 years to summon a convention by referendum. And if the voters are asleep that year, as they were in the Empire State in 1957, then prevailing injustice must continue for another two decades.

To Hasten Reform

Initiative and referendum offer channels for reform in the 20 states where these procedures are available. The Supreme Court made special note of the fact that it felt required to act in the Tennessee case because that state did not have initiative and referendum. Some informed opinion even believed that the Supreme Court would not act in a state where there was a viable system of popular action.

But even this path is thorny. In 1926 Massachusetts courts found that the initiative was not applicable to reapportionment. Before that a similar course was followed by the Missouri Supreme Court. In 1956, an initiative statute on reapportionment was passed in Washington, but in 1957 the state legislature undid the deed and was upheld by a five to four decision in the state Supreme Court.

One of the by-products of Baker v Carr may be its clearing the path of initiative should it be used in the future to reapportion. A legislature or court that overruled such an initiative might well be inviting the Supreme Court to intervene.

For the 50 states as a whole, the most available and the most vigorous instrument to hasten or compel reform is the court. (The Supreme Court in Baker felt it was the only instrument.) The lower courts, in particular, have displayed an imaginative truculence in devising sanctions to speed action. In Vermont, the court acted directly: redistricted two seats by switching counties around. To make this stick, the judge barred county clerks from printing ballots until the reapportionment was carried out as he recommended. This action was in line with the opinion of Supreme Court Justice Clark who suggested that it might be wise “to start with existing assembly districts, consolidate some of them, and award the seats thus released to those counties suffering the most egregious discrimination.” In Alabama, a three-judge federal court ordered reapportionment with the caveat that if the legislature failed to do so, the court would itself carry through a provisional districting along lines of Clark’s plan. In Mississippi, State Chancellor W. T. Horton ordered the state legislature to reapportion, adding that “should our legislative bodies not act timely, this court will reapportion them, and enter a decree enjoining the state and county election commissions and commissioners from holding elections or electing any Senator or Representative…” In Wisconsin, a three-judge federal court on June 27 ordered the legislature to a complete reapportionment by 5 p.m. July 2—or the court would appoint a master to do the job. The Colorado Supreme Court unveiled a battery of missiles on June 14 in a series of show cause orders, including a call on the Secretary of State, George Baker, to show cause within seven days why he should not be restrained from holding the election and a call on the State Treasurer, Tim Armstrong, to show cause why he should not be required to cut off the pay of the General Assembly.

Two draconic measures are available to the courts should they prove constitutional. One is to order an election at large in a state. This is less a cure than shock therapy. Should it come in a state where a governing party is also a popular majority, the medicine may be worse than the malady. The other is to order the use of the weighted vote. This is an untried instrument but it enjoys the virtue of serving both as immediate sanction and as a constitutional pattern for the future.

The final problem is to find a formula that will not necessitate repeated court intervention. The first step appears to be a constitutional provision on reapportionment that leaves as little room for discretion as possible. This would specify the formula for both apportionment and districting, use fixed lines for districts wherever possible and a mathematical equation for apportionment of seats to districts. The second step would be to place enforcement in the hands of an apportioning agency, other than the legislature. The third step would allow judicial review in the state and ultimately the federal courts.

Gus Tyler was director of the Political Department of the International Ladies Garment Workers Union. This article originally ran in the July 16, 1962 issue of the magazine.