In the decade since Earl Warren became Chief Justice of the United States, the Court over which he presides has embarked on three major enterprises of social reform—a number higher than the historical average for comparable periods, to say the least. The reforms in question are the desegregation of public schools; the restriction, and even, it would seem, the abolition of practices of religious observance in those schools; and the reapportionment of state legislatures and perhaps also of Congress. The earliest of these reforms, desegregation, has survived an extremely serious political challenge, and now promises to move forward along well-defined lines toward a goal that is widely acceptable and nowhere misunderstood. The second, concerning school prayers, seems likely to face an equally serious challenge, which it may fail to survive. Though still in need of more than a little clarification, the goal of this reform is at any rate capable of being proclaimed in principled terms and of commanding deliberate adherence as well as opposition. The third reforming enterprise, reapportionment, differs from desegregation and school prayers—and is, indeed, not quite like anything else in the recorded experience of the Supreme Court—in that the decision that launched it a year ago (Baker v. Carr) has evoked a speedy, ample, and largely favorable response, although nobody understands what the decision means or can know where it will, or was intended to, lead.
It is, in retrospect, not difficult to appreciate why the response to Baker v. Carr has been what it has. Malapportionment has for years been the cynosure of assorted muckrakers—a common, catch-all grievance, the supposed cause of urban rot and of the obsolescence of federalism, the real reason, for all we know, why “wealth accumulates, and men decay”; and so something of an embarrassment even to those who benefit from it. The agitators for reform were straining at the starting gate. The Supreme Court fired a starting gun, and the reformers were off. But they cannot long continue to run without knowing where. The curious failure of the Court to announce any comprehensible goal is therefore a matter of some consequence. It means that we ought not be overly impressed by the initial response, and that, of the Warren Court’s three reforming enterprises, this one, which began in a sense most hopefully, is perhaps the least likely to play a substantial and enduring role in shaping our society.
* * *
Baker v. Carr (March 1962) marked a sharp break with precedent. Previously the Court had turned aside all attacks on state or congressional apportionments and on kindred political arrangements either by declining to take jurisdiction or by very briefly and inscrutably indicating that the system under attack did not violate any constitutional principle. This was, on the whole, an unsatisfactory situation. The Court’s position, never precisely or comprehensively stated, was subject to grave misapprehension. Why had the Court approved certain apportionments, while neither approving nor disapproving others? Could the refusal to approve or disapprove be construed as itself a form of approval? Moreover, the Court also treated obsolete apportionments—those that are the simple product of indecision and inertia—no differently than deliberately contrived malapportionments. Normally, in respect of other matters, legislative inertia and the express legislative will are not dealt with indiscriminately in the Supreme Court. The sum of it was that, willy-nilly, the Court was helping to entrench the status quo.
The decision in Baker v. Carr—the fact of the decision and also some passages in the opinions of the Justices—cleared the air. The previous decisions, and all intimations that existing apportionments, obsolete or fresh, are necessarily constitutional were swept away. The case itself concerned, not a recent wilful malapportionment, but rather a sixty-year-old apportionment in Tennessee that had been rendered obsolete and oppressive by sheer inaction in the face of radical shifts of population. The decision made it clear that this situation could not be regarded as constitutionally valid, and predictably enough, the Tennessee legislature responded by acting at long last. Elsewhere as well, whether under pressure of pending lawsuits, or even before any were filed, the most widespread immediate result of Baker v. Carr was a flurry of legislative activity looking to fresh apportionments.
Seen as I have just described it, the decision in Baker v. Carr was welcome and unexceptionable. The trouble is that Mr. Justice Brennan’s majority opinion went beyond an emphasis on the age of the Tennessee apportionment and held—or at least was taken to have held—that the constitutionality of all apportionments, recent or ancient, is henceforth to be passed on as a matter of course, in much the same fashion in which, say, racial practices or criminal procedures are generally brought to ultimate judgment in the Supreme Court. This reading of the decision rests on the following passage in Justice Brennan’s opinion:
Nor need the appellants, in order to succeed in this action, ask the Court to enter upon policy determinations for which judicially manageable standards are lacking. Judicial standards under the Equal Protection clause are well-developed and familiar, and it has been open to courts since the enactment of the Fourteenth Amendment to determine, if on the particular facts they must, that a discrimination reflects no policy, but simply arbitrary and capricious action.
Taking this passage as prescribing their function, numerous lower federal and state courts have during the past year struck down or threatened to strike down fairly recent apportionments on the ground that they were somehow irrational and unconstitutional. Some courts have even set their hand or threatened to set their hand to the task of themselves constructing new and constitutional apportionments. But the standard of judgment enunciated by Mr. Justice Brennan, which lower courts have purported to apply in all directions, and which the Supreme Court itself will be called upon to apply as appeals in apportionment cases mount on its docket—this standard, though it falls soothingly on the ear, is almost wholly meaningless. And because this standard is meaningless, judicially-directed reapportionment is unlikely to work in the way its enthusiastic supporters anticipate.
The requirement of rationality invoked by Justice Brennan is, to be sure, a familiar principle in American constitutional law. It stems from a celebrated paragraph of John Marshall’s in McCulloch v. Maryland (1819), written with reference to the powers of Congress, but applicable as well to judicial review of acts of state legislatures:
We admit, as all must admit, that the powers of the government are limited, and that its limits are not to be transcended. But we think the sound construction of the Constitution must allow to the national [or state] legislature that discretion, with respect to the means by which the powers it confers are to be carried into execution, which will enable that body to perform the high duties assigned to it, in the manner most beneficial to the people. Let the end be legitimate, let it be within the scope of the Constitution, and all means which are appropriate, which are plainly adapted to that end, which are not prohibited, but consist with the letter and spirit of the Constitution, are constitutional.
The test was restated just before the turn of the century in a famous paper entitled, “The Origin and Scope of the American Doctrine of Constitutional Law,” by the constitutional scholar, James Bradley Thayer. A statute is to be declared unconstitutional, Thayer said, “when those who have the right to make laws have not merely made a mistake, but have made a very clear one—so clear that it is not open to rational question.” The Supreme Court, in Thayer’s view, is to be “the ultimate arbiter of what is rational and permissible.”
More concretely, Marshall and Thayer were talking about a method of constitutional construction—that is, a method of interpreting the written document—which would result in the free exercise by the representative institutions of broad and effective powers of government, but not unlimited ones. Thus—in an illustration of Marshall’s—the federal power to establish a postal system being conceded because it is one of those enumerated in the Constitution, it follows rationally that there should be power to punish interference with the movement of the mails. However, it may not follow rationally that because the federal government is empowered to “regulate commerce among the several states,” it is also empowered to establish a uniform law of divorce: we should have to know more of the relation of divorces to the operation of a national economy before we could call this connection rational.
There is also a second, related sort of application of the test of rationality. In exercising any of its powers, government—state or federal—will be permitted to make only those choices that rest on an intelligible and plausible view of reality; only those choices, as Brandeis said, “which an informed, intelligent, just-minded, civilized man could rationally favor.” And as in the first set of illustrations, what is here in question is not really a requirement of rationality so much as a requirement of the absence of demonstrable irrationality. For, obviously, there are myriad legislative choices that are not rationally compelled, as indeed the very word choice implies. Thus, however disagreeable one might find such action, it would still be rational—given what is known of Communist doctrine—for a legislature to act on the belief that a man who belongs to the Communist party at the very least does not look with disfavor upon the violent overthrow of existing American institutions. But it is irrational in the light of common experience—and has been held so by the courts—for a legislature also to conclude that once a man is a Communist, he will always remain one, no matter what present disclaimers he may enter. Or again, a state may rationally conclude that members not only of the Communist party but also of organizations shown to have close connections with that party are in some sense disloyal. But it is surely irrational to conclude, and has been held so by the courts, that persons who joined and left an organization before it had any connection with the Communist party, or who joined not knowing about the connection, have given similar evidence of any kind of disloyalty.
The test of rationality is not, of course, invulnerable to certain lines of philosophical attack. I am not here concerned to defend it, aside from remarking that it has found a substantial practical utility in the working of the American constitutional scheme. But I am concerned with the obvious omission in this test—namely, the question, not whether a legislative choice is rational (whatever the tacit value judgment that may be implicit in that word), but whether it is good. Confronted, for example, with the issue of whether the federal government should be able to punish interference with the free movement of mails, it is quite satisfactory to most of us to be told that the power is one that can be rationally deduced from the undoubted authority of the federal government to run a post office. This is, as Thayer called it, a question of “mere power.” But when it comes to excluding members of a Communist-front organization—let us assume for the moment that we know what that means—from government employment or from the Bar, it does not in the same fashion suffice to be informed that the action is rational. One wants to know also whether it is good, whether, in other words, it is consistent with principles of political freedom. The question is not merely one of power; it is also a question of liberty. It is possible to answer that legislatures should be free to make the choice of values that is involved, just as they are free to make other choices that are not irrational. But such a plenary legislative freedom has never quite characterized our constitutional system. By means of a complex, not unrestricted, and ultimately not uncontrolled process, the Supreme Court has participated in the making of many such fundamental value choices for our society.
Returning to the apportionment problem, one can now perhaps make quite plain that application of the rationality test, as suggested by Mr. Justice Brennan, is very nearly meaningless. Since it is conceded all around that the power to apportion—to create constituencies and otherwise regulate the manner in which public officials and legislators are elected—is a legitimate function of government, the question becomes whether this or that apportionment, restricting the electorate in this or that fashion, or favoring this or that portion of it with more or less influence, is rational—or rather, whether it is ir rational; the question, as always, is not whether reason compels the legislative choice, but whether reason is repelled by it. Now most, if not all, malapportionments favor rural interests over urban, allocate more strength proportionately to sparsely populated areas than to densely populated ones, and make other smaller discriminations within these large ones. This may be undesirable, but who can say that it is irrational? Is it more irrational than a farm policy that favors farmers or an anti-trust policy that favors small enterprise? Are such policies more irrational, indeed, than a policy that forces the majority to listen to ideas that it fears and hates, or to tolerate the possibility that its young will read Henry Miller? Mr. Justice Clark, concurring specially in Baker v. Carr, looked for an internal consistency in the apportionment system in Tennessee, and finding none, held the system to be irrational. But whence is the premise derived that an apportionment must not only be rational in the particular choices that it makes, but must proceed toward a single coherent objective?
The upshot is that, while there may be exceptions, particularly in highly obsolete apportionments, the conventional test of rationality cannot generally lead anywhere in this field-—no more than it could lead anywhere in the field of racial discrimination. No doubt, given the present state of scientific knowledge, a legislature which acted on the assumption that Negroes are inherently less intelligent than whites, or that Jews are less likely to be honest than gentiles, would be held to have acted irrationally. There is no foundation for such an assumption in the common experience of informed, rational men, although there might be one in common prejudice. But for a legislature to conclude that racial tension is an existing fact of life, and that it is therefore best to segregate the races; or for it to conclude that most Negroes are not likely to go on to a university and that it is therefore not economical to maintain one for them—these are certainly not irrational acts. The Constitution is held to forbid them because they are bad in principle, because it is morally reprehensible for government to treat persons in this unequal fashion, no matter what the rational reason for doing so. Is there a similar principle that can be applied in the field of apportionment to restrict otherwise rational legislative choices?
This is the question that the Court not only left unanswered in Baker v. Carr, but actually went out of its way to obscure. One does not expect a full and complete answer to such a question to spring into life from a Court’s first opinion on a problem, but neither does one expect the question to be willfully obfuscated. In any event, if it be true (as it was not with respect to racial discrimination and as it is not even with respect to school prayers) that no viable principle limiting a legislature’s discretion to apportion rationally as it sees fit can as yet be worked out, then it follows that despite all the current furious activity and heightened expectations, we cannot look for any enduring result from this particular enterprise in judicially-directed reform.
The problem of apportionment seems to call forth in many people the most abstract notions concerning the nature and Functioning of representative democratic government. The root idea, apparently, is the one stated by President Kennedy at a news conference on March 29, 1962, shortly after the decision in Baker v. Carr. “Quite obviously,” said the President, “the right to fair representation, that each vote count equally is, it seems to me, basic to the successful operation of a democracy.” The President went on to remark, forgetting perhaps where he once sat, that there “is no sense of a Senator representing five million people sitting next to a Senator representing ten thousand. . . .” But what does this imply? One-person-one-vote? And how? Proportional representation? At-large elections? Is districting to be allowed? What is the use of a second House, or must we all become unicameral, like Nebraska? It is very well to maintain that the federal Senate and the federal Electoral College are the product of an historical compromise, which is explained by particular historical circumstances and which is not applicable to the composition of the House of Representatives or of state legislatures. But are we to believe that our federal government lacks a feature that is “basic to the successful operation of a democracy?”—that, indeed, it contains an inconsistent feature, and is therefore not a successfully operating democracy?
What the President said—and others equally sophisticated are capable of professing their devotion to such simplicities—describes some sort of town-meeting democracy, but not anything that has ever existed in this country on any larger scale, and certainly not anything that exists today. If the Supreme Court has undertaken to make reality conform to the one-person-one-vote-principle, it has set for itself the most enormous of conceivable tasks, on which one would expect it to be busily engaged for the next hundred years. Nevertheless, if a one-person-one-vote democracy is the goal toward which we should strive, then it would be the function of the Court to define it and set it before us, even though a hundred years might be needed to achieve it. The goal of racial equality, set before us in the School Segregation Cases, is not about to be attained tomorrow either, but we adhere to it anyway as one of whose worthiness we are now certain. Should we similarly adhere to the goal of voting equality? Is it what we are now certain we want? Is the true meaning of representative democracy to be found within it?
The principle of equality of representation proceeds from the premise that the representative institutions are something like animated voting machines, engineered to register decisions made by the electorate. These institutions are seen as a rather poor substitute for a decision-making process that works by direct vote of the people. As delegates of the people, they are authorized to dispose of unforeseen matters that were not settled at a prior election, but these decisions are to be ratified or rejected at the next election, and if possible sooner. And above all, as the French say, not too much zeal. Legislatures should meet infrequently, and terms of office should be very short.
The initiative and referendum movement, which engaged the enthusiasm of Progressives in the first decade or so of this century, but which can scarcely be said to have left a significant mark on our political system, operated on some such set of assumptions. And, indeed, on no other premise is equality of representation supportable as the overriding, all-important principle. For if the representative institutions are seen as properly exercising a relatively independent, deliberate decision-making function, elections and the vote become matters of a different order of importance. On this premise, the problem of democracy becomes one of access to, participation in, influence on the process of decision, and only ultimately and in necessarily attenuated fashion one of ensuring at election time the legislature’s fidelity to the popular will.
This, of course, is how modern political science views the problem: the equal-vote premise ignores all that we have learned in a generation of fresh inquiry and reflection. And it ignores not only the laws of probability, but also—in a highly developed, pluralistic society—all the laws of possibility. We have come to realize that elections cannot be occasions when the electorate disposes of the common run of issues that governments are confronted with. But is this to say that elections are practically meaningless, a show-trial of democracy, a spectacle, an emotional binge, a morality play, or what-have-you? There have been those who have come close to arguing as much, particularly when in the throes of defending that plainly counter-majoritarian institution, the Supreme Court—which is a point of some little irony. The truth, however, is that some elections do sometimes turn on issues, although few and general, and no one knows ahead of time which issues they may turn on or how. This very ignorance is a factor in the process by which democratic government makes decisions, and we can therefore say that elections do influence decisions in some indeterminate measure. But the heart of democratic government and the morality which distinguishes it from other systems lie elsewhere: the heart of the matter is that democratic government rests on consent. And the secret of consent is only in part a matter of control, only in part the reserve power of a majority to rise up against decisions that displease it. Perhaps more importantly, it is the sense shared by all that their interests were spoken for in the decision-making process, no matter how the result turned out. Government by consent requires that no segment of society should feel alienated from the institutions that govern. This means that the institutions must not merely represent a numerical majority—which is a shifting and uncertain quantity anyway—but must reflect the people in all their diversity, so that all the people may feel that their particular interests and even prejudices, that all their diverse characteristics, were brought to bear on the decision-making process.
We have thus, and ought to have, majoritarian government in the sense that an essentially numerical majority has, and knows it has, the reserve power to turn out some discredited decision-makers, putting in others who will in future resist what has displeased it. But we also have, and ought to have, reflective government, to which general consent is given, and so government which is legitimate and stable. To achieve both of these aims, we have constructed not one but a number of institutions, and their constituencies are different. It is commonly true, both locally and nationally, that the executive—which has, of course, great powers of initiative as well as approval in the legislative process—is the instrument through which numerical majorities express and make effective such concrete wishes as they may be able to generate. It is also true that the legislature carries the burden of reflecting the diversities of population, and that it could not under any circumstances perform this function if it were built strictly on the one-person-one-vote principle. And where no such function is performed—where no reflective institutions have effective power that is shared in some measure by all groups, interests, factions, and other configurations in society—what we are accustomed to call democracy does not obtain. No matter how freely a strong majoritarian executive like General de Gaulle may be elected—although in his instance it is not even true that he is freely elected by a majority—and no matter how frequently he may put major decisions to a free majoritarian vote by referendum, the result will not be democratic government so long as no reflective institutions share power.
Supporters of reapportionment argue that the reflective character of our legislatures can be maintained through districting, but that districts need not be unequal in size. Yet the people’s diversities are not all conveniently contained in equal contiguous districts; many interests and other configurations, far from being so clustered, would form permanent minorities throughout any number of equal districts. Are there enough Silk Stockings in New York, for example, to elect a Congressman from an equal, compact, contiguous silk-stocking district? It would take some map-making ingenuity to achieve an adequately reflective legislature on the equal vote principle. Thus, as Professor Phil C. Neal, of the University of Chicago Law School, has recently pointed out in an important paper on the subject, a requirement of equal districting would very likely prove self-defeating, for it is possible to achieve all the malapportionment in the world by careful gerrymandering of perfectly equal districts. Those who have been complaining might then have a magnificently equal vote that would never elect anyone. Proportional representation is a mathematically perfect way out, but as has been demonstrated time and again, it is a method that so fragments legislative institutions as to make them incapable of coherent and responsible government. Our institutions are over-fragmented as is; their capacity for inaction is one of the ills attributed to malapportionment. We tend to cure it through the two-party system, and justly complain that the cure is imperfect. Proportional representation invariably leads to a multi-party system, thus forcing the legislature to engage in the initial coalition-building process which needs to have taken place before the legislature is formed, if effective government is to ensue.
It is far from clear, then, to quote Professor Neal, that the one-person-one-vote idea ‘“suggests a principle which even the interests represented by the urban plaintiffs [in Baker v. Carr and like cases] could consider satisfactory under all circumstances, and even less clear that it is a principle which assures fair treatment to other interests equally entitled to fair treatment.” It also follows from our analysis, however, that democratic government requires some institution to be available as an effective voice of the majority, and the executive, including the federal President, fulfills that function in our system. To be sure, the Electoral College seems to deflect the majoritarian character of the federal Presidency. Actually, however, not since 1888 have we elected a minority President in a two-party contest—and then Benjamin Harrison lost the popular vote by about the same tiny margin (although, of course, it was greater relatively) by which John F. Kennedy won it. In circumstances of such even balance, a toss of the coin, the Electoral College, or, as recently in Minnesota, a three-judge recount, serve—one about as well as another—to make a decisive choice and render it legitimate. Moreover, the Electoral College works in practice to strengthen the majoritarian influence, because its distortion of a true majority counterbalances precisely the distortion to be found in the malapportioned legislature. A President views his constituency as being weighted to the places where the major population centers are. Even Harrison won because he carried New York.
All governors are straight-out majoritarian. Until recently, Georgia constituted an exception, but no longer. The state provisionally abandoned the county-unit system in the 1962 election under pressure from a lower federal court, and on March 18, the Supreme Court made it final. Unfortunately, however, the Court, in an opinion by Mr. Justice Douglas, offered no tenable principle in support of its conclusion that election of governors and other state-wide officials by the county-unit method is unconstitutional. While the result of the case can be justified in terms of the analysis I have put forward, it is given the benefit of no coherent explanation in the Court’s opinion, save only by invocation of the one-person-one-vote principle, which is untenable as a general proposition, and which evidently does not, as a general proposition, have the assent of all members of the Court’s majority.
It remains to ask whether we have evolved or can see emerging some operative principle other than equal representation which is capable of general application. Neither the Supreme Court in the two cases—Baker v. Carr and the county-unit case—it has decided, nor any of the other courts that have been busily tackling the problem has come within shouting distance of such a principle. All we have been given are plays on words, plays on statistics, and meaningless figures arbitrarily picked out of thin air. One court, for example, held with a straight face that the ratio of population disparities between districts must be no more than 4—1. And why not, a commentator has asked, 5—1 or 3—1? Indeed, why not? Well, the answer may be, because the court said 4—1, not 5 or 3, and somebody has got to decide these things. But when things are to be decided on a hunch or out of thin air, because there is no other way, it is a fixed characteristic of our system that we let the political institutions do it. The political institutions reflect the people, even if imperfectly, and they include a governor responsive to a majority. Malapportioned though these institutions may be, they are more easily subject to correction and more nearly responsible and responsive than courts—which do not suffer from malapportionment, because they are neither apportioned nor representative. It is an irony to which I alluded earlier that the super-democrats should look to the unrepresentative courts for an arbitrary decision that they resent when it is made by a faulty representative legislature, acting in concert with a majoritarian governor.
Courts are fit to render judgment on questions of principle, which we do not, in our tradition, relegate to the political market-place, to be disposed of by one of those not irrational, but often intuitive or otherwise unverifiable choices that result from the play of conflicting interests. It may be that some principle applicable to apportionments which goes beyond the requirement merely that the executive be majoritarian can be worked out. If so, the first wisdom is to look at the reality of our political arrangements, the reality of our allocations of power to govern, and not at paper provisions and statistical nightmares. What does it mean to juggle ratios or to bewail the fact that 20 per cent of a state’s population can elect a majority of its legislature, X per cent of the population of the United States can elect the President, and X-10 per cent can elect the Senate? These are not facts; such things never happen. The malapportioned federal Senate, for example, is more nearly majoritarian than the House, because Minneapolis and even Atlanta are coming to resemble New York and Cleveland more than they do their neighboring counties, and in state-wide elections, Minneapolis and Atlanta count heavily, and count even beyond the mere numbers of their people. If a tenable principle is ever to emerge, it will proceed from an understanding of the realities of power, of the role of parties and how they are run, of the role of money and of various relevant skills, of the state-wide influence of urban home-rule governments and of various groups and factions, and so forth and so on. When we know, as the title of Professor Robert A. Dahl’s recent book puts it, “who governs,” we may begin to be in a position to lay down a constitutional principle that tells us who should govern. Until then, our imperfect representative institutions are better fitted than the courts to tinker with the system.
The political arena is messier than the judicial, to be sure, but that is where all of us who feel under- or misrepresented should be, exerting every ounce of power and influence we can command. We may be right, but it may turn out, especially if we lose, that we were after all wrong. It is quite possible, for example, that malapportionment has been held responsible for too many evils, and that the responsiveness of our legislatures has been more gravely impaired by antiquated committee and seniority systems, and by other institutional habits that are seldom brought to light. Let a major political effort, led by majoritarian executives, be tried in that direction. So far as Congress is concerned, it may also be that it fails to act when it does because it is truly representative. The country, we should not forget, was very closely divided in 1960. This may be a deplorable fact, but it is scarcely undemocratic. As to state legislatures, urban rot, and the obsolescence of federalism, it may be that the real causes of our discontent are fiscal. The federal government has pre-empted the sources of public finance. State governments subsist in large measure on federal handouts. It is therefore somewhat less than surprising to find them relatively inert and irresponsible. Moreover, there is a finite amount of political energy and even attention in the people, and what there is the federal government, again, pre-empts more and more.
It is therefore question-begging to ask how we can force reapportionment, and a consequent solution of our problems, on a minority entrenched through malapportionment. The supposed minority may not be a minority at all; malapportionment may not be what entrenches it; and our problems may be rooted elsewhere anyway. If, nevertheless, we beg these questions, and assume a minority entrenched in the legislature through malapportionment, the answer is that the remedy lies with the majoritarian executive, whom we can influence, and whose own bargaining power can very properly be heightened by federal judicial holdings striking down obsolete apportionments and forcing legislatures to act affirmatively. The courts may be in no position to do more, but this much they can do.