The Castle Lectures in Ethics, Politics, and Economics

The Castle Lectures in Ethics, Politics, and Economics

00dahlFM.i_x 11/27/01 4:37 PM Page i

00dahlFM.i_x 11/27/01 4:37 PM Page ii

How Democratic Is the

AmericanConstitution?

Robert A. Dahl

Yale University Press / New Haven & London

00dahlFM.i_x 11/27/01 4:37 PM Page iii

Published with assistance from the foundation established in memory ofPhilip Hamilton McMillan of the Class of 1894, Yale College.

Copyright © 2001 by Yale University.All rights reserved.This book may not be reproduced, in whole or in part, includingillustrations, in any form (beyond that copying permitted by Sections 107 and 108 of the U.S. Copyright Law and except by reviewers for the public press), without written permission from the publishers.

Designed by James J. Johnson and set in New Caledonia and Bulmer types by Integrated Publishing Solutions.Printed in the United States of America.

Library of Congress Cataloging-in-Publication Data

Dahl, Robert Alan, 1915–How democratic is the American Constitution?/

Robert A. Dahl.p. cm. — (The Castle lecture series)

Includes bibliographical references and index.ISBN 0-300-09218-0 (cloth : alk. paper)1. Constitutional law—United States. 2. Constitutional history—

United States. 3. Democracy. I. Title. II. Series.KF4550 .D34 2001342.73’02—dc21 2001004323

A catalogue record for this book is available from the British Library.

The paper in this book meets the guidelines for permanence anddurability of the Committee on Production Guidelines for BookLongevity of the Council on Library Resources.

10 9 8 7 6 5 4 3 2 1

00dahlFM.i_x 11/27/01 4:37 PM Page iv

Parts of this book were given as a series

of Castle Lectures in Yale’s Program in

Ethics, Politics, and Economics, deliv-

ered by Robert Dahl at Yale University

in 2000.

The Castle Lectures, endowed by Mr.

John K. Castle, honor his ancestor, the

Reverend James Pierpont, one of Yale’s

founders. Given by prominent public

figures, the lectures are intended to

promote reflection on the moral foun-

dations of society and government and

to enhance understanding of ethical is-

sues facing individuals in our complex

modern society.

00dahlFM.i_x 11/27/01 4:37 PM Page v

00dahlFM.i_x 11/27/01 4:37 PM Page vi

Contents

Acknowledgments ix

CHAPTER 1. Introduction: Fundamental Questions 1

CHAPTER 2. What the Framers Couldn’t Know 7

CHAPTER 3. The Constitution as a Model: An American Illusion 41

CHAPTER 4. Electing the President 73

CHAPTER 5. How Well Does the Constitutional System Perform? 91

CHAPTER 6. Why Not a More Democratic Constitution? 121

CHAPTER 7. Some Reflections on the Prospects for a More Democratic Constitution 141

Appendix A: On the Terms “Democracy” and “Republic” 159

Appendix B: Tables and Figures 163

Notes 173

Index 191

00dahlFM.i_x 11/27/01 4:37 PM Page vii

00dahlFM.i_x 11/27/01 4:37 PM Page viii

Acknowledgments

THE INVITATION TO DELIVER THE CASTLE LECTURES AT

Yale provided me with an incentive for bringing intofocus views about the American Constitution that I hadgradually formed over many years. Though I had setforth some of my arguments in various essays and bookchapters, others remained largely implicit or undevel-oped until I drafted the lectures. This book embodiesthe substance of the Castle Lectures, slightly revisedand enlarged, that I delivered in the early fall of 2000.

For extending to me the invitation to deliver thelectures, I am indebted to Geoffrey Garrett, directorof the Program in Ethics, Politics, and Economics. Toboth him and Ian Shapiro I want to express my appre-ciation for their warm endorsement of the subject Iproposed for my lectures.

00dahlFM.i_x 11/27/01 4:37 PM Page ix

For her invaluable research assistance I am in-debted to Jennifer Smith.

For their helpful queries, corrections, or contribu-tions I also want to thank Wendell Bell, Kai Erikson,Fred Greenstein, Steven Hill, Malcolm Jewell, JosephLaPalombara, Rogers Smith, and, at Yale UniversityPress, Ali Peterson for her sensitive editing and LarisaHeimert for her skillful and energetic help in conduct-ing the manuscript through the publishing process.

Finally, let me take this opportunity to express mythanks to those members of the audience whose ques-tions and comments enabled me to discover aspects ofmy presentation that would benefit, in this publishedversion, from greater clarification or more extensivetreatment.

x ac k n ow l e d g m e n t s

00dahlFM.i_x 11/27/01 4:37 PM Page x

chapter 1

Introduction:Fundamental Questions

MY AIM IN THIS BRIEF BOOK IS NOT TO PROPOSE

changes in the American Constitution but tosuggest changes in the way we think about

our constitution. In that spirit, I’ll begin by posing asimple question: Why should we Americans upholdour Constitution?

Well, an American citizen might reply, it has beenour constitution ever since it was written in 1787 by agroup of exceptionally wise men and was then ratifiedby conventions in all the states.1 But this answer onlyleads to a further question.

To understand what lies behind that next question,I want to recall how the Constitutional Convention thatmet in Philadelphia during the summer of 1787 wasmade up. Although we tend to assume that all thirteen

01dahl.001_006 11/27/01 4:38 PM Page 1

states sent delegates, in fact Rhode Island refused toattend, and the delegates from New Hampshire didn’tarrive until some weeks after the Convention opened.As a result, several crucial votes in June and July weretaken with only eleven state delegations in attendance.Moreover, the votes were counted by states, and al-though most of the time most state delegations agreedon a single position, on occasion they were too dividedinternally to cast a vote.

My question, then, is this: Why should we feelbound today by a document produced more than twocenturies ago by a group of fifty-five mortal men, actu-ally signed by only thirty-nine, a fair number of whomwere slaveholders, and adopted in only thirteen statesby the votes of fewer than two thousand men, all ofwhom are long since dead and mainly forgotten?2

Our citizen might respond that we Americans arefree, after all, to alter our constitution by amendmentand have often done so. Therefore our present consti-tution is ultimately based on the consent of those of usliving today.

But before we accept this reply, let me pose an-other question: Have we Americans ever had an op-portunity to express our considered will on our consti-tutional system? For example, how many readers ofthese lines have ever participated in a referendum thatasked them whether they wished to continue to begoverned under the existing constitution? The answer,of course, is: none.

2 i n t ro d u c t i o n

01dahl.001_006 11/27/01 4:38 PM Page 2

Our citizen might now fall back on another line ofargument: Why should we change a constitution thathas served and continues to serve us well?

Although this is surely a reasonable line of argu-ment, it does suggest still another question: By whatstandards does our constitution serve us well? In par-ticular, how well does our constitutional system meetdemocratic standards of the present day? I’ll turn tothis question in the next chapter.

And if our constitution is as good as most Ameri-cans seem to think it is, why haven’t other democraticcountries copied it? As we’ll see in Chapter 3, everyother advanced democratic country has adopted a con-stitutional system very different from ours. Why?

If our constitutional system turns out to be uniqueamong the constitutions of other advanced democraticcountries, is it any better for its differences, or is itworse? Or don’t the differences matter? I’ll explorethis difficult question in the fourth chapter.

Suppose we find little or no evidence to supportthe view that our constitutional system is superior tothe systems of other comparable democratic countries,and that in some respects it may actually performrather worse. What should we conclude?

As one part of an answer, I am going to suggestthat we begin to view our American Constitution asnothing more or less than a set of basic institutionsand practices designed to the best of our abilities forthe purpose of attaining democratic values. But if an

i n t ro d u c t i o n 3

01dahl.001_006 11/27/01 4:38 PM Page 3

important democratic value is political equality, won’tpolitical equality threaten the rights and liberties weprize? In Chapter 5, I’ll argue that this view—fa-mously defended by Tocqueville, among others—isbased on a misunderstanding of the relationship be-tween democracy and fundamental rights.

Yet the question remains: if our constitution is insome important ways defective by democratic stan-dards, should we change it, and how? As I said, myaim here is not so much to suggest changes in the ex-isting constitution as to encourage us to change theway we think about it, whether it be the existing one,an amended version of it, or a new and more demo-cratic constitution. That said, in my final chapter I’llcomment briefly on some possible changes and on theobstacles to achieving them.

� � �

BEFORE TURNING TO THESE QUESTIONS, I NEED TO DIS-pose of two matters. One is purely terminological. Indiscussing the formation of the constitution at theConvention in 1787, I shall refer to the delegates asthe Framers, not, as is more common, the FoundingFathers. I do so because many of the men who reason-ably might be listed among the Founding Fathers—including such notables as John Adams, Samuel Adams,Tom Paine, and Thomas Jefferson—were not at theConvention. (By my count, only eight of the fifty-five


01dahl.001_006 11/27/01 4:38 PM Page 4

delegates to the Convention had also signed the Dec-laration of Independence.)

The second matter is both terminological and sub-stantive. Some readers may argue that the FoundingFathers (including the Framers) intended to create arepublic, not a democracy. From this premise, accord-ing to a not uncommon belief among Americans, it fol-lows that the United States is not a democracy but arepublic. Although this belief is sometimes supportedon the authority of a principal architect of the Consti-tution, James Madison, it is, for reasons I explain inAppendix A, mistaken.

But even more important, the conclusion does notfollow from the premise. Whatever the intentions of theFramers may have been, we would hardly feel boundby them today if we believed that they were morally,politically, and constitutionally wrong. Indeed, morethan two centuries of experience demonstrates thatwhenever a sufficiently large and influential numberof Americans conclude that the views of the Framerswere wrong, they will change the constitution. Even ifthe Framers did not intend their constitution to abol-ish slavery, when later generations concluded that slav-ery could no longer be tolerated and must be abol-ished, they changed the constitution to conform withtheir beliefs.

Even if some of the Framers leaned more towardthe idea of an aristocratic republic than a democraticrepublic, they soon discovered that under the leader-

i n t ro d u c t i o n 5

01dahl.001_006 11/27/01 4:38 PM Page 5

ship of James Madison, among others, Americans wouldrapidly undertake to create a more democratic repub-lic, and in doing so they would begin almost immedi-ately to change the constitutional system the Framershad created.


01dahl.001_006 11/27/01 4:38 PM Page 6

chapter 2

What the Framers Couldn’t Know

W ISE AS THE FRAMERS WERE, THEY WERE

necessarily limited by their profound igno-rance.

I say this with no disrespect, for like many others Ibelieve that among the Framers were many men of ex-ceptional talent and public virtue. Indeed, I regardJames Madison as our greatest political scientist andhis generation of political leaders as perhaps our mostrichly endowed with wisdom, public virtue, and devo-tion to lives of public service. In the months and weeksbefore the Constitutional Convention assembled “onMonday the 14th of May, A.D. 1787. [sic] and in theeleventh year of the independence of the United Statesof America, at the State-House in the city of Philadel-

02dahl.007_040 11/27/01 4:39 PM Page 7

phia,”1 Madison studied the best sources as carefullyas a top student preparing for a major exam.2 But evenJames Madison could not foresee the future of theAmerican republic, nor could he draw on knowledgethat might be gained from later experiences with de-mocracy in America and elsewhere.

It is no detraction from the genius of Leonardoda Vinci to say that given the knowledge available inhis time he could not possibly have designed a work-able airplane—much less the spacecraft that nowbears his name. Nor, given the knowledge available in1903, could the Wright brothers have built the Boeing707. Although like many others I greatly admire Ben-jamin Franklin, I recognize that his knowledge of elec-tricity was infinitesimal compared with that of a first-year student in electrical engineering—or, for thatmatter, the electrician who takes care of my occasionalwiring problems. In fact, on that famous first experi-ment with the kite, Franklin was lucky to have escapedalive. None of us, I expect, would hire an electricianequipped only with Franklin’s knowledge to do ourwiring, nor would we propose to make a trip fromNew York to London in the Wright brothers’ aircraft.Leonardo, Franklin, the Wright brothers were greatinnovators in their time, but they could not draw onknowledge that was still to be accumulated in theyears and centuries to come.

The knowledge of the Framers—some of them,certainly—may well have been the best available in

8 w h at t h e f r a m e r s co u l d n ’ t k n ow

02dahl.007_040 11/27/01 4:39 PM Page 8

1787. But reliable knowledge about constitutions appropriate to a large representative republic was, atbest, meager. History had produced no truly relevantmodels of representative government on the scale the United States had already attained, not to mentionthe scale it would reach in the years to come. As much as many of the delegates admired the Britishconstitution, it was far from a suitable model. Norcould the Roman Republic provide much of a guide.The famous Venetian Republic, illustrious though ithad been, was governed by a hereditary aristocracy offewer than two thousand men and was already totter-ing: a decade after the Convention an upstart Corsicanwould knock it over in a featherweight military attack.Whatever knowledge the delegates could gain fromhistorical experience was, then, only marginally rele-vant at best.

Leaping into the Unknown

Among the important aspects of an unforeseeable fu-ture, four broad historical developments would yieldsome potential knowledge that the Framers necessar-ily lacked and that, had they possessed it, might wellhave led them to a different constitutional design.

First, a peaceful democratic revolution was soon toalter fundamentally the conditions under which theirconstitutional system would function.

w h at t h e f r a m e r s co u l d n ’ t k n ow 9

02dahl.007_040 11/27/01 4:39 PM Page 9

Second, partly in response to that continuing revo-lution, new democratic political institutions would fun-damentally alter and reconstruct the framework theyhad so carefully designed.

Third, when democratization unfolded in Europeand in other English-speaking countries during the twocenturies to come, constitutional arrangements wouldarise that were radically different from the Americansystem. Within a generation or two, even the Britishconstitution would bear little resemblance to the onethe Framers knew—or thought they knew—and inmany respects admired and hoped to imitate.

Fourth, ideas and beliefs about what democracy re-quires, and thus what a democratic republic requires,would continue to evolve down to the present day andprobably beyond. Both in the way we understand themeaning of “democracy” and in the practices and insti-tutions we regard as necessary to it, democracy is not a static system. Democratic ideas and institutions asthey unfolded in the two centuries after the AmericanConstitutional Convention would go far beyond theconceptions of the Framers and would even transcendthe views of such early democrats as Jefferson andMadison, who helped to initiate moves toward a moredemocratic republic.

I shall consider each of these developments inlater chapters. But first I want to indicate some of thepractical limitations on what the Framers could rea-sonably achieve.


02dahl.007_040 11/27/01 4:39 PM Page 10

What the Framers Couldn’t Do

The Framers were not only limited by, so to speak,their inevitable ignorance. They were also cruciallylimited by the opportunities available to them.

We can be profoundly grateful for one crucial re-striction: the Framers were limited to considering onlya republican form of government. They were con-strained not only by their own belief in the superiorityof a republican government over all others but also bytheir conviction that the high value they placed on re-publicanism was overwhelmingly shared by Americancitizens in all the states. Whatever else the Framersmight be free to do, they well knew that they could notpossibly propose a monarchy or a government ruled byan aristocracy. As the Massachusetts delegate ElbridgeGerry put it, “There was not a one-thousand part of ourfellow citizens who were not against every approach to-ward monarchy.”3 The only delegate who was recordedby Madison4 as looking with favor on monarchy wasAlexander Hamilton, whose injudicious expression ofsupport for that heartily unpopular institution may havegreatly reduced his influence at the Convention, as itwas to haunt him later.5 Hardly more acceptable was anadaptation of aristocratic ideas to an American constitu-tion. During the deliberations about the Senate, Gou-verneur Morris of Pennsylvania explored the possibilityof drawing its members from an American equivalentof the British aristocracy.6 But it soon became obvious

11

02dahl.007_040 11/27/01 4:39 PM Page 11

that the delegates could not agree on just who theseAmerican aristocrats might be, and in any case theywell knew that the overwhelming bulk of American cit-izens would simply not tolerate such a government.

A second immovable limit was the existence of thethirteen states, with still more states to come. A consti-tutional solution that would be available in most of thecountries that were to develop into mature and stabledemocracies—a unitary system with exclusive sover-eignty lodged in the central government, as in Britainand Sweden, for example—was simply out of the ques-tion. The need for a federal rather than a unitary re-public was therefore not justified by a principle ad-duced from general historical experience, much lessfrom political theory. It was just a self-evident fact. IfAmericans were to be united in a single country, it wasobvious to all that a federal or confederal system wasinescapable. Whether the states would remain as fun-damental constituents was therefore never a seriousissue at the Convention; the only contested questionwas just how much autonomy, if any, they would yieldto the central government.7

The delegates had to confront still another stub-born limit: the need to engage in fundamental com-promises in order to secure agreement on any consti-tution at all. The necessity for compromise and theopportunities this gave for coalitions and logrollingmeant that the Constitution could not possibly reflecta coherent, unified theory of government. Compro-mises were necessary because, like the country at large,


02dahl.007_040 11/27/01 4:39 PM Page 12

members of the convention held different views onsome very basic issues.

Slavery. One, of course, was the future of slav-ery. Most of the delegates from the five southern stateswere adamantly opposed to any constitutional provi-sion that might endanger the institution. Although thedelegates from the other seven states were hardly ofone mind about slavery, it was perfectly obvious to themthat the only condition on which coexistence would beacceptable to the delegates from the southern stateswould be the preservation of slavery. Consequently, ifthese delegates wanted a federal constitution theywould have to yield, no matter what their beliefs aboutslavery. And so they did. Although some delegates whosigned the final document abhorred slavery, they nev-ertheless accepted its continuation as the price of astronger federal government.

Representation in the Senate. Another conflict ofviews that could not be settled without a one-sidedcompromise resulted from the adamant refusal of thedelegates from the small states to accept any constitu-tion that did not provide for equal representation inthe Senate. The opponents of equal representation in-cluded two of the most illustrious members of theConvention, James Madison and James Wilson, whowere also among the chief architects of the Constitu-tion. Both men bitterly opposed what seemed to theman arbitrary, unnecessary, and unjustifiable limit onnational majorities. As Alexander Hamilton remarkedabout this conflict: “As states are a collection of indi-


02dahl.007_040 11/27/01 4:39 PM Page 13

vidual men which ought we to respect most, the rightsof the people composing them, or the artificial beingsresulting from the composition. Nothing could be morepreposterous or absurd than to sacrifice the former tothe latter. It has been sd. that if the smaller States re-nounce their equality, they renounce at the same timetheir liberty. The truth is it is a contest for power, notfor liberty. Will the men composing the small States beless free than those composing the larger.”8

Let me give you a flavor of the elevated discussionthat preceded the victory of the small states. Here isGunning Bedford of Delaware on June 30:

The large states dare not dissolve the Confederation. Ifthey do the small ones will find some foreign ally ofmore honor and good faith, who will take them by thehand and do them justice.

To which Rufus King of Massachusetts replied:

I cannot sit down, without taking some notice of the lan-guage of the honorable gentleman from Delaware. . . .It was not I who with a vehemence unprecedented inthis House, declared himself ready to turn his hopesfrom our common Country, and court the protection ofsome foreign hand. . . . I am grieved that such a thoughthas entered into his heart. . . . For myself whatevermight be my distress, I would never court relief from aforeign power.9

Faced with the refusal of the small states to ac-cept anything less, Madison, Wilson, Hamilton, andthe other opponents of equal representation finally ac-


02dahl.007_040 11/27/01 4:39 PM Page 14

cepted compromise of principle as the price of a con-stitution. The solution of equal representation was not,then, a product of constitutional theory, high prin-ciple, or grand design. It was nothing more than apractical outcome of a hard bargain that its opponentsfinally agreed to in order to achieve a constitution.10

Incidentally, this conflict illustrates some of thecomplexities of voting coalitions at the ConstitutionalConvention, for the faction opposed to equal repre-sentation in the Senate included four strange bed-fellows: Madison, Wilson, Hamilton, and GouverneurMorris. Although all four generally supported movesto strengthen the federal government, Madison andWilson usually endorsed proposals that leaned towarda more democratic republic, while Hamilton and Mor-ris tended to support a more aristocratic republic.

Undemocratic Elements in the Framers’ Constitution

It was within these limits, then, that the Framers con-structed the Constitution. Not surprisingly, it fell farshort of the requirements that later generations wouldfind necessary and desirable in a democratic republic.Judged from later, more democratic perspectives, theConstitution of the Framers contained at least sevenimportant shortcomings.

Slavery. First, it neither forbade slavery nor em-powered Congress to do so. In fact, the compromise


02dahl.007_040 11/27/01 4:39 PM Page 15

on slavery not only denied Congress the effectivepower to prohibit the importation of slaves before180811 but it gave constitutional sanction to one of themost morally objectionable byproducts of a morally re-pulsive institution: the Fugitive Slave laws, accordingto which a slave who managed to escape to a free statehad to be returned to the slaveholder, whose propertythe slave remained.12 That it took three-quarters of acentury and a sanguinary civil war before slavery wasabolished should at the least make us doubt whetherthe document of the Framers ought to be regarded asholy writ.

Suffrage. Second, the constitution failed to guar-antee the right of suffrage, leaving the qualifications of suffrage to the states.13 It implicitly left in place the exclusion of half the population—women—as wellas African Americans and Native Americans.14 As weknow, it took a century and a half before women wereconstitutionally guaranteed the right to vote, and nearlytwo centuries before a president and Congress couldovercome the effective veto of a minority of states inorder to pass legislation intended to guarantee the vot-ing rights of African Americans.

Election of the president. Third, the executivepower was vested in a president whose selection, ac-cording to the intentions and design of the Framers,was to be insulated from both popular majorities andcongressional control. As we’ll see, the Framers’ maindesign for achieving that purpose—a body of presi-dential electors composed of men of exceptional wis-


02dahl.007_040 11/27/01 4:39 PM Page 16

dom and virtue who would choose the chief executiveunswayed by popular opinion—was almost immedi-ately cast into the dustbin of history by leaders sympa-thetic with the growing democratic impulses of theAmerican people, among them James Madison him-self. Probably nothing the Framers did illustrates moresharply their inability to foresee the shape that politicswould assume in a democratic republic. (I shall saymore about the electoral college in a later chapter.)

Choosing senators. Fourth, senators were to bechosen not by the people but by the state legislatures,for a term of six years.15 Although this arrangementfell short of the ambitions of delegates like Gouver-neur Morris who wanted to construct an aristocraticupper house, it would help to ensure that senatorswould be less responsive to popular majorities andperhaps more sensitive to the needs of property hold-ers. Members of the Senate would thus serve as acheck on the Representatives, who were all subject topopular elections every two years.16

Equal representation in the Senate. The attemptto create a Senate that would be a republican versionof the aristocratic House of Lords was derailed, as wehave seen, by a prolonged and bitter dispute over anentirely different question: Should the states be equallyrepresented in Congress or should members of bothhouses be allocated according to population? This ques-tion not only gave rise to one of the most disruptive is-sues of the Convention, but it resulted in a fifth unde-mocratic feature of the constitution. As a consequence


02dahl.007_040 11/27/01 4:39 PM Page 17

of the famous—or from a democratic point of view, in-famous—“Connecticut Compromise” each state was,as we have seen, awarded the same number of sena-tors, without respect to population. Although thisarrangement failed to protect the fundamental rightsand interests of the most deprived minorities, somestrategically placed and highly privileged minorities—slaveholders, for example—gained disproportionatepower over government polices at the expense of lessprivileged minorities. (I shall come back to this ele-ment in the constitution in a later chapter.)

Judicial power. Sixth, the constitution of theFramers failed to limit the powers of the judiciary todeclare as unconstitutional laws that had been prop-erly passed by Congress and signed by the president.What the delegates intended in the way of judicial re-view will remain forever unclear; probably many dele-gates were unclear in their own minds, and to the ex-tent that they discussed the question at all, they werenot in full agreement. But probably a majority ac-cepted the view that the federal courts should rule onthe constitutionality of state and federal laws in casesbrought before them. Nevertheless, it is likely that a substantial majority intended that federal judgesshould not participate in making government laws andpolicies, a responsibility that clearly belonged not tothe judiciary but to the legislative branch. Their oppo-sition to any policy-making role for the judiciary isstrongly indicated by their response to a proposal inthe Virginia Plan that “the Executive and a convenient


02dahl.007_040 11/27/01 4:39 PM Page 18

number of the National Judiciary, ought to compose acouncil of revision” empowered to veto acts of the Na-tional Legislature. Though this provision was vigor-ously defended by Madison and Mason, it was voteddown, 6 states to 3.17

A judicial veto is one thing; judicial legislation isquite another. Whatever some of the delegates mayhave thought about the advisability of justices sharingwith the executive the authority to veto laws passed byCongress, I am fairly certain that none would havegiven the slightest support to a proposal that judgesshould themselves have the power to legislate, tomake national policy. However, the upshot of theirwork was that in the guise of reviewing the constitu-tionality of state and congressional actions or inac-tions, the federal judiciary would later engage in whatin some instances could only be called judicial policy-making—or, if you like, judicial legislation.18

Congressional power. Finally, the powers of Con-gress were limited in ways that could, and at times did,prevent the federal government from regulating orcontrolling the economy by means that all moderndemocratic governments have adopted. Without thepower to tax incomes, for example, fiscal policy, not tosay measures like Social Security, would be impos-sible. And regulatory actions—over railroad rates, airsafety, food and drugs, banking, minimum wages, andmany other policies—had no clear constitutional au-thorization. Although it would be anachronistic tocharge the Framers with lack of foresight in these


02dahl.007_040 11/27/01 4:39 PM Page 19

matters,19 unless the constitution could be altered byamendment or by heroic reinterpretation of its provi-sions—presumably by what I have just called judiciallegislation—it would prevent representatives of latermajorities from adopting the policies they believedwere necessary to achieve efficiency, fairness, and se-curity in a complex post-agrarian society.

Enlightened as the Framers’ constitution may havebeen by the standards of the eighteenth century, fu-ture generations with more democratic aspirationswould find some of its undemocratic features objec-tionable—and even unacceptable. The public expres-sion of these growing democratic aspirations was notlong in coming.

Even Madison did not, and probably could not,predict the peaceful democratic revolution that wasabout to begin. For the American revolution was soonto enter into a new and unforeseen phase.

The Framers’ Constitution Meets Emergent Democratic Beliefs

We may tend to think of the American republic and itsconstitution as solely the product of leaders inspiredby extraordinary wisdom and virtue. Yet without a citi-zenry committed to republican principles of govern-ment and capable of governing themselves in accor-dance with those principles, the constitution wouldsoon have been little more than a piece of paper. As


02dahl.007_040 11/27/01 4:39 PM Page 20

historical experience would reveal, in countries wheredemocratic beliefs were fragile or absent, constitutionsdid indeed become little more than pieces of paper—soon violated, soon forgotten.

The American democratic republic was not cre-ated nor could it have been long maintained by lead-ers alone, gifted as they may have been. It was they, tobe sure, who designed a framework suitable, as theythought, for a republic. But it was the American people,and the leaders responsive to them, who ensured thatthe new republic would rapidly become a democraticrepublic.

The proto-republican phase. The ideas, practices,and political culture necessary to sustain a republicangovernment were by no means unfamiliar to Ameri-cans. Unlike some countries that have moved almostovernight from dictatorship to democratic forms, andoften soon thereafter to chaos and back to dictator-ship, by 1787 the Americans had already accumulateda century and a half of experience in the arts of gov-ernment.

The long colonial period had provided opportuni-ties to both leaders and many men of ordinary rank to become acquainted with the requirements of self-government, both in the direct form of a town meetingand through electing representatives to the coloniallegislatures.20 We easily forget that although in its twofamous opening paragraphs the Declaration of Inde-pendence laid down some new and audacious claims,in the rest of that document—the part few people


02dahl.007_040 11/27/01 4:39 PM Page 21

bother to read today—the authors mainly protestedagainst the British king for violating rights that, withsome exaggeration, they had previously enjoyed asEnglishmen.

The republican phase. The next phase, creating apopular republic, had begun with the astounding dec-laration on July 4, 1776, “that all Men are createdequal.” The Declaration marks the beginning of a se-ries of events that went much further than simplygaining independence from Britain. In what the histo-rian Gordon Wood has called the “greatest Utopianmovement in American history,”21 the Declaration alsotriggered a democratic revolution in beliefs, practices,and institutions—or better, an evolution—that has con-tinued ever since. The two decades since independ-ence had provided still more, and deeper, experiencein the practices of self-government. Nor was this expe-rience limited to a tiny minority. In some of the thir-teen states, a fairly high proportion of adult males hadacquired the franchise.22

Toward a democratic republic. The lengthy colo-nial and post-independence experience provided asturdy foundation for the efforts that Americans nowundertook in the next phase of the revolution, whenthe new republic was transformed into a more demo-cratic republic. To be sure, at the end of the eigh-teenth century few Americans were ready to concedethat the principles of the Declaration, much less dem-ocratic citizenship, applied to everyone.23 It would taketwo more centuries of evolution in democratic beliefs


02dahl.007_040 11/27/01 4:39 PM Page 22

before most Americans would be inclined to agreethat the famous claim in the Declaration might be re-phrased: not just “all men,” but “all persons are cre-ated equal.”

Yet always keeping in mind the huge and persistentexceptions, by the standards prevailing elsewhere inthe world the extent of equality among Americans wasextraordinary. Alexis de Tocqueville, who observedAmericans during his year’s visit in 1831–32, openedhis famous work with these words:

Among the novel objects that attracted my attentionduring my stay in the United States, nothing struck memore forcibly than the general equality of conditions. Ireadily discovered the prodigious influence which thisprimary fact exercises on the whole course of society, bygiving a certain direction to public opinion, and a cer-tain tenor to the laws; by imparting new maxims to thegoverning powers, and peculiar habits to the governed.

I speedily perceived that the influence of this factextends far beyond the political character and the lawsof the country, and that it has no less empire over civilsociety than over the Government. . . .

The more I advanced in the study of American so-ciety, the more I perceived that the equality of condi-tion is the fundamental fact from which all others seemto be derived, and the central point at which all my ob-servations constantly terminated.24

During the three decades before Tocqueville ar-rived, under the leadership of Jefferson, Madison, andothers, supporters of a more democratic republic hadalready made some changes. The seismic shift from


02dahl.007_040 11/27/01 4:39 PM Page 23

the views of the Framers and the Federalists is sym-bolized by the changing name of the party that wonboth the presidency and Congress in the election thatJefferson called—as have later historians—the Revo-lution of 1800. To defeat the Federalists, win the elec-tion, and gain control of the new government, Jeffer-son and Madison had created a political party thatthey appropriately named the Democratic-RepublicanParty. By 1832, with Andrew Jackson as its winningcandidate, the Democratic-Republican party becamethe Democratic Party, plain and simple.25 The namehas stuck ever since.

Conservative delegates among the Framers—laterthe core of the Federalist Party—had feared that if or-dinary people were given ready access to power theywould bring about policies contrary to the views and in-terests of the more privileged classes, which, as theconservative delegates viewed their interests, were alsothe best interests of the country. These conservativefears were soon confirmed. Within a decade the emi-nent Federalist leaders were pushed aside and the Fed-eral Party became a minority party. A generation laterhad seen the demise of both the party and its leaders.

If these changes justified some of the pessimismabout popular majorities of many of the Framers, theirpessimism proved unjustified in another important re-spect. A substantial number of the Framers believedthat they must erect constitutional barriers to popularrule because the people would prove to be an unrulymob, a standing danger to law, to orderly government,


02dahl.007_040 11/27/01 4:39 PM Page 24

and to property rights. Contrary to these pessimisticappraisals, when American citizens were endowed withthe rights and opportunities to support demagoguesand rabble rousers, they chose instead to support law,orderly government, and property rights. White maleAmericans were, after all, mainly farmers who ownedtheir own land; or, where farm land was not easilyavailable because most of it had already been occu-pied, they could count on the ready availability of goodfarm land farther west—often obtained, to be sure, at the expense of its earlier inhabitants, the NativeAmericans.

White Americans in vast numbers bought westernland and settled down on their own farms. “Two-thirdsof the landless white men of Virginia moved West inthe 1790s. . . . Between 1800 and 1820, the trans-Appalachian population grew from a third of a millionto more than two million.”26 In foreseeing a democraticrepublic based on a citizen body consisting predomi-nantly of independent farmers, mainly property ownerscultivating their own lands, Jefferson reflected the re-ality of his time.27 Outside the South, and even in thesouthern piedmont, a predominant number of Ameri-can citizens were free farmers who stood to benefitfrom an orderly government dependent on their votes.

Ordinary citizens also revealed strong beliefs indemocratic values and procedures. Presented with theopportunity to do so, they would choose leaders whocultivated democratic values and procedures. Just suchan opportunity was soon presented by four acts passed


02dahl.007_040 11/27/01 4:39 PM Page 25

in 1798 by the Federalists, who were alarmed not onlyby the seemingly subversive activities of France butalso by the rapidly growing influence of boisterous, ir-reverent, and sometimes libelous opponents in thenew Republican party. In particular, the Federalistsemployed one of these new laws, the Sedition Act, inan effort to silence Republican critics. Notable amongthe fourteen who were prosecuted was a bombastic andsomewhat unsavory Republican congressman, the Irishimmigrant Mathew Lyon, whose only memorable con-tribution to American history was his conviction forsedition, which carried a fine of a thousand dollars—ahuge amount in those days—and four months in jail.28

To the Republicans, the Sedition Act was a flagrant vio-lation of the newly adopted First Amendment. Afterthey gained the presidency and control of Congress inthe election of 1800, the Sedition Act was allowed tolapse, despite the vigorous efforts of the Federalists.

Democratic Changes to the Framers’ Constitution: Amendments

The fate of the Alien and Sedition Acts symbolizes alarger change at work in the country. The democraticrevolution, fitful and uncertain though it would for-ever remain, not only helped to democratize the for-mal constitution itself by amendments, it generatednew democratic political institutions and practiceswithin which the constitutional system would operate.


02dahl.007_040 11/27/01 4:39 PM Page 26

The constitutional system that has emerged is nolonger that of the Framers, nor is it one they had in-tended to create.

The Bill of Rights. To be sure, the first ten amend-ments to the Constitution—the Bill of Rights—cannotbe attributed to the democratic revolution that fol-lowed the Convention. They resulted instead from de-mands within the Convention itself by delegates whogenerally favored a more democratic system than theircolleagues could then accept. Among the most influen-tial of these was George Mason, who wrote the Virginiaconstitution and its Declaration of Rights. Respondingto the insistent demands of Mason and several others,as well as to similar voices outside the Convention,Mason’s fellow Virginian, James Madison, drafted tenamendments that were ratified in 1789–90 by elevenstates, more than a sufficient number for their adop-tion. (Incidentally, the two laggards, Georgia and Con-necticut, finally did come around—but not until 1939!)Thus, for all practical purposes the Bill of Rights was a part of the original constitution. In any case, theamendments have proved to be a veritable cornucopiaof expanding rights necessary to a democratic order.29

Other Amendments

As I have mentioned, the most profound violation ofhuman rights permitted by the original constitution,slavery, was not corrected until the adoption of the


02dahl.007_040 11/27/01 4:39 PM Page 27

Thirteenth, Fourteenth, and Fifteenth Amendmentsbetween 1865 and 1870. In 1909 the Sixteenth Amend-ment in 1913 gave Congress the power to enact in-come taxes. The election of U.S. senators by state legis-latures finally gave way to direct election with theadoption of the Seventeenth Amendment in 1913.Women were finally guaranteed the right of suffrage infederal and state elections with the passage of theNineteenth Amendment in 1919. Although the effortto add an Equal Rights Amendment failed, the Four-teenth Amendment was later interpreted to provide a constitutional basis for eliminating discriminationagainst women as well as certain minorities whosemembers suffered from discriminatory practices. Theiniquitous poll tax that had continued to bar AfricanAmericans from voting in some southern states was fi-nally forbidden in 1964 by the Twenty-Fourth Amend-ment. Finally, in a move toward a more inclusive elec-torate, in 1971 the Twenty-Sixth Amendment reducedthe voting age to eighteen.

In this halting fashion, the democratic revolutionbelatedly worked its way through the Constitution toovercome the veto power of long-entrenched minoritiesand to eliminate some of the most flagrantly undemo-cratic features of the constitution. As Alan Grimes ob-served some years ago, of the twenty-six (now twenty-seven) amendments to the constitution, “Twenty-oneamendments may be said to affirm either the principleof democratic rights or that of democratic processes.”30


02dahl.007_040 11/27/01 4:39 PM Page 28

Democratic Changes in Political Practices and Institutions

The constitution of the Framers was changed not onlyby formal amendments. It was also fundamentally al-tered by political practices and institutions that theFramers did not foresee, even though they were un-avoidable—indeed, highly desirable—in a democraticrepublic.

Political parties. Perhaps the most important ofthese was the political party. The Framers feared anddetested factions, a view famously expressed by Madisonin Federalist No. 10.31 Probably no statement has beenso often cited to explain and justify the checks againstpopular majorities that the Framers attempted to buildinto the constitution. It is supremely ironic, therefore,that more than anyone except Jefferson, it was Madisonwho helped to create the Republican Party in order todefeat the Federalists. Although the system would notsettle down for some years, Jefferson and Madisonhelped to inaugurate the competitive two-party systemthat has pretty much remained in place ever since.

Which suggests other questions. Despite the claimof every political party everywhere in the world that ittruly represents the general interest, aren’t politicalparties really “factions” in Madison’s sense? So did theFramers fail after all to prevent government by fac-tions? And did they succeed only in making it moredifficult for a majority faction to prevail—that is, aparty reflecting the interests of a majority coalition?

29

02dahl.007_040 11/27/01 4:39 PM Page 29

Whatever the best answers to these hard ques-tions, it cannot be denied that partisan politics trans-formed the constitution. Despite their familiarity withthe role of the Tories and Whigs in Britain and nascentparties in their own legislatures, the Framers did notfully foresee that in a democratic republic politicalparties are not only possible, they are also inevitableand desirable. As Jefferson and Madison soon came torealize, without an organized political party to mobi-lize their voters in the states and their fellow support-ers in the Congress, they could not possibly overcomethe entrenched political domination of their politicaladversaries, the Federalists. The democratic rights in-corporated in the Bill of Rights made parties possible;the need to compete effectively made them inevitable;the ability to represent citizens who would otherwisenot be adequately represented made them desirable.

Today we take for granted that political parties andparty competition are essential to representative de-mocracy: we can be pretty sure that a country whollywithout competitive parties is a country without democ-racy. If the Framers had been aware of the central im-portance of political parties to a democratic republic,would they have designed their constitution differently?They might well have. At the very least they would nothave created the absurdity of an electoral college.

The electoral college. In an outcome the Framershad made possible by their defective design of theelectoral college, the election of 1800 produced a tiebetween Jefferson and his running mate, Aaron Burr.


02dahl.007_040 11/27/01 4:39 PM Page 30

From the time the final results were known in lateDecember 1800, the deadlock in the electoral collegepersisted, despite many attempts at persuasion andcompromise, until February 17, 1801, when shifts andabstentions by a number of state delegations gave Jef-ferson the presidency.32 Ironically, the very institutionthat the Framers hoped would insulate the election ofthe president from partisan politics was its first victim.Although a similar fiasco was prevented in the futureby the Twelfth Amendment in 1804, even with theamendment the electoral college was converted by par-tisan politics into nothing more than a rather peculiarand ritualized way of allocating the votes of the statesfor president and vice president. Yet the electoral col-lege still preserved features that openly violated basicdemocratic principles: citizens of different states wouldbe unequally represented, and a candidate with thelargest number of popular votes might lose the presi-dency because of a failure to win a majority in the elec-toral college. That this outcome was more than a theo-retical possibility had already occurred three timesbefore it was displayed for all the world to see in theelection of 2000. I’ll come back to the democraticshortcomings of the electoral college in a later chapter.

The Democratic Revolution: What Madison Learned—and Taught

James Madison arrived in Philadelphia in 1787, a fewmonths past his thirty-sixth birthday. He was already


02dahl.007_040 11/27/01 4:39 PM Page 31

far from a political neophyte, having been elected atthe age of twenty-five to the Virginia constitutionalconvention where, with George Mason, he helped todraft the Virginia Declaration of Rights and the newstate constitution. He then became successively a mem-ber of the Virginia legislature (though he failed to bereelected because, it was said, he refused to treat thevoters to the customary rum punch), a delegate to theContinental Congress, and again a member of the Vir-ginia legislature. In the months before the Constitu-tional Convention opened, he drafted the outline ofthe proposal that would be presented in the openingdays of the Convention and that would come to beknown as the Virginia Plan. (We shall see something ofits contents in the next chapter.)

Yet, experienced as he was, like his fellow delegatesMadison brought to the Convention limited knowl-edge of the institutions and practices that a more fullydemocratized republic would require. Before his deathin 1836 at the age of eighty-five, nearly half a centuryafter the Convention, Madison could have looked backon a rich body of experience that would have shapedhis constitutional views in many ways.

Following the Convention, he was elected to theU.S. House of Representatives where he drafted andintroduced the first ten amendments to the Constitu-tion—the Bill of Rights. With Jefferson he soon be-came a leader of the opposition to Federalist policiesand ideas. As we have seen, they formed and led the


02dahl.007_040 11/27/01 4:39 PM Page 32

opposition party, the Democratic Republicans. AfterJefferson’s election, Madison became secretary ofstate. He then succeeded Jefferson in the presidency.By the time he left that office in 1817, his views aboutdemocratic political institutions were probably as wellinformed as those of any person then alive.

However that may be, the Madison of seventy in1821 was no longer the Madison of thirty-six in 1787.Among other changes, the Madison of 1821 wouldhave trusted popular majorities—American popularmajorities, anyway—far more than the Madison of1787. The mature and experienced Madison of 1821might therefore have done less to check majority ruleand more to facilitate it. Let me offer several pieces ofevidence, one from a time early in his awakening tothe requirements of a democratic republic, the othersfrom his reflections in old age.

I have already alluded to the first: the basic alter-ation in his views about “factions,” or what the two distinguished historians of Federalism describe as“Madison Revises The Federalist.”33 Madison’s views inFederalist No. 10, influenced by his reading of DavidHume, are cited endlessly: the dangers of factions, thethreat from majorities united on principles contrary tothe general interest, political parties as at best a neces-sary evil. But these were not his more mature views.

In January 1792, less than five years after the closeof the Convention, Madison begins to publish a seriesof essays in The Gazette, an opposition newspaper


02dahl.007_040 11/27/01 4:39 PM Page 33

published by Philip Freneau. The first is entitled “OnParties.” In “every political society,” he writes, “partiesare unavoidable.” To combat their dangers, Madisonoffers five proposals that might well serve us better inour own time than the anti-majoritarian biases dis-played in Federalist No. 10. Whatever dangers politi-cal parties may pose can be overcome

“By establishing political equality among all.”

“By withholding unnecessary opportunities from a few,to increase the inequality of property by an immoder-ate, and especially unmerited, accumulation of riches.”

“By the silent operation of the laws, which, without vio-lating the rights of property, reduce extreme wealth to-wards a state of mediocrity, and raise extreme indigencetoward a state of comfort.”

“By abstaining from measures which operate differentlyon different interests, and particularly favor one inter-est, at the expense of another.”

“By making one party a check on the other, so far as theexistence of parties cannot be prevented, nor theirviews accommodated.”34

“If this is not the language of reason,” he went on tosay, “it is that of republicanism.”

Nearly thirty years later (around 1821), when he ispreparing his notes on the constitutional debates forpublication, he records some of his later reflections.As to the right of suffrage, he remarks that his obser-vations at the Convention “do not convey the speaker’s[Madison’s] more full and matured view of the sub-


02dahl.007_040 11/27/01 4:39 PM Page 34

ject.” “The right of suffrage,” he now insists, “is a fun-damental Article in Republican Constitutions.” Healso makes explicit his view of political parties: “Nofree Country,” he says, “has ever been without parties,which are a natural offspring of Freedom.” But politi-cal parties and a broad suffrage may create a conflictover property. “An obvious and permanent division ofevery people is into the owners of the Soil, and theother inhabitants.” Consequently, if the suffrage is ex-tended to citizens who are not freeholders, a majoritymight threaten the property rights of the freeholders.

Madison then considers a number of possible solu-tions to this problem, of which the first would be to re-strict the suffrage to “freeholders, and to such as holdan equivalent property.” He rejects this solution withan observation that might well have been a centralprinciple of the Second Phase of the American Revo-lution. “The objection to this regulation,” he writes, “isobvious. It violates the vital principle of free Govt. thatthose who are to be bound by laws, ought to have avoice in making them. And the violation wd. be morestrikingly unjust as the lawmakers became the minor-ity.” A second option is “confining the right of suffragefor one branch to the holders of property, and for theother Branch to those without property.” But to do so“wd. not in fact be either equal or fair.” Nor prudent:“The division of the State into the two Classes . . .might lead to contests & antipathies not dissimilar tothose between the Patricians and Plebeians at Rome.”

After examining other possibilities, he concludes:


02dahl.007_040 11/27/01 4:39 PM Page 35

Under every view of the subject, it seems indispensablethat the Mass of Citizens not be without a voice, inmaking the laws which they are to obey, & in chusingthe Magistrates, who are to administer them, and if theonly alternative be between an equal & universal rightof suffrage for each branch of the Govt. and a confine-ment of the entire right to a part of the Citizens, it isbetter that those having the greater interest at stakenamely that of property & persons both, should be de-prived of half their share in the Govt. than, that thosehaving the lesser interest, that of personal rights only,should be deprived of the whole.35

The older Madison is also more favorable to ma-jority rule. Like most of his contemporaries, Madisonbelieves that “all power in human hands is liable to beabused.” But taking that assumption as axiomatic to-gether with the need for government, the relevantquestion becomes: what kind of government is best?His answer remains unchanged:

In Governments independent of the people, the rightsand views of the whole may be sacrificed to the views ofthe Government. In Republics, where the people gov-ern themselves, and where, of course, the majority gov-ern, a danger to the minority arises from opportunitiestempting a sacrifice of their rights to the interest, realor supposed, of a majority. No form of government,therefore, can be a perfect guard against the abuse ofpower. The recommendation of the republican form is,that the danger of abuse is less than any other.36

What has changed is his greater confidence in ma-jority rule. Compared with its alternatives at least, the


02dahl.007_040 11/27/01 4:39 PM Page 36

mature Madison is confident that majority rule, in thewords of Marvin Meyers, promises the “least imper-fect government.”37

“[E]very friend to Republican Government,” hewrites in 1833, “ought to raise his voice against thesweeping denunciation of majority Governments as themost tyrannical and intolerable of all Governments.”

It has been said that all Government is an evil. It wouldbe more proper to say that the necessity of any govern-ment is a misfortune. This necessity however exists; andthe problem to be solved is, not what form of govern-ment is perfect, but which of the forms is least imper-fect; and here the general question must be between arepublican Government in which the majority rule theminority, and a government in which a lesser number orthe least number rule the majority.

The result . . . is, that we must refer to the moni-tory reflection that no government of human deviceand human administration can be perfect; that thatwhich is the least imperfect is therefore the best gov-ernment; that the abused of all other governments haveled to the preference of republican government as thebest of all governments, because the least imperfect;that the vital principle of republican government is thelex majoris parties, the will of the majority.38

� � �

I HAVE LITTLE DOUBT THAT IF THE AMERICAN CONSTI-tutional Convention had been held in 1820, a very dif-ferent constitution would have emerged from the de-liberations—although, I hasten to add, we can never


02dahl.007_040 11/27/01 4:39 PM Page 37

know what shape that constitution might have taken.We can be reasonably sure, however, that the dele-gates would have attempted to provide more supportfor, and fewer barriers to, a democratic republic.

As to the undemocratic features of the constitutioncreated in 1787, let me suggest four conclusions.

First, the aspects of the constitution that are mostdefective from a democratic point of view do not nec-essarily all reflect the intentions of the Framers, inso-far as we may surmise them. Though the flaws aretraceable to their handiwork, they are in some casesflaws resulting from the inability of these superbly tal-ented craftsmen to foresee how their carefully craftedinstrument of government would work under the chang-ing conditions that were to follow—and most of all,under the impact of the democratic revolution in whichAmericans were, and I hope still are, engaged.

Second, some of the undemocratic aspects of theoriginal design also resulted from the logrolling andcompromises that were necessary to achieve agree-ment. The Framers were not philosophers searchingfor a description of an ideal system. Nor—and we maybe forever grateful to them for this—were theyphilosopher kings entrusted with the power to rule.They were practical men, eager to achieve a strongernational government, and as practical men they madecompromises. Would the country have been better offif they had refused to do so? I doubt it. But in anycase, they did compromise, and even today the consti-tution bears the results of some of their concessions.


02dahl.007_040 11/27/01 4:39 PM Page 38

I’ll have more to say on that point in my next chapter.Third, undemocratic aspects that were more or

less deliberately built into the constitution overesti-mated the dangers of popular majorities—Americanpopular majorities, at any rate—and underestimatedthe strength of the developing democratic commit-ment among Americans. As a result, in order to adaptthe original framework more closely to the require-ments of the emerging democratic republic, with thepassage of time some of these aspects of the originalconstitution were changed, sometimes by amendment,sometimes, as with political parties, by new institu-tions and practices.

Finally, though the defects seem to me serious andmay grow even more serious with time, Americans arenot much predisposed to consider another constitu-tion, nor is it clear what alternative arrangementswould serve them better.

As a result, the beliefs of Americans in the legiti-macy of their constitution will remain, I think, in con-stant tension with their beliefs in the legitimacy ofdemocracy.

For my part, I believe that the legitimacy of theconstitution ought to derive solely from its utility as aninstrument of democratic government—nothing more,nothing less. In my last chapter, I’ll reflect further onthe meaning of that judgment.


02dahl.007_040 11/27/01 4:39 PM Page 39

02dahl.007_040 11/27/01 4:39 PM Page 40

chapter 3

The Constitution as a Model:An American Illusion

MANY AMERICANS APPEAR TO BELIEVE THAT OUR

constitution has been a model for the rest ofthe democratic world.1 Yet among the coun-

tries most comparable to the United States and wheredemocratic institutions have long existed without break-down, not one has adopted our American constitutionalsystem. It would be fair to say that without a single ex-ception they have all rejected it. Why?

Before I explore that question, I need to clarify twomatters. As you may have noticed, rather than speakingsimply of “the constitution,” I’ve sometimes used thephrase “the constitutional system.” I do so because Iwant to include in a constitutional system an importantset of institutions that may or may not be prescribed inthe formal constitution itself: these are its electoral

03dahl.041_072 11/27/01 4:39 PM Page 41

arrangements. As we’ll see, electoral systems can in-teract in crucial ways with the other political institu-tions and thereby determine the way they function.

Also, I’ve just referred to the countries where democracy is oldest and most firmly established. Wecould call them the older democracies, the maturedemocracies, the stable democratic countries, and soon, but I’ll settle on “the advanced democratic coun-tries.” Whatever we choose to call them, in order tocompare the characteristics and performance of theAmerican constitutional system with the characteris-tics and performance of the systems in other demo-cratic countries, we need a set of reasonably compa-rable democratic countries. In short, we don’t want tocompare apples and oranges—or good apples and rot-ten apples.

I’ve noticed that we Americans often assure our-selves of the superiority of our American political sys-tem by comparing it with political systems in countriesruled by nondemocratic regimes or in countries thatsuffer from violent conflict, chronic corruption, fre-quent chaos, regime collapse or overthrow, and thelike. On voicing or hearing criticism of political life inthe United States, an American not infrequently adds,“Yes, but just compare it with X!,” a favorite X beingthe Soviet Union during the Cold War and, after itscollapse, Russia. One could easily pick more than ahundred other countries with political systems that byalmost any standard are unquestionably inferior to ourown. But comparisons like this are absurdly irrelevant.

42 t h e co n st i t u t i o n a s a m o d e l

03dahl.041_072 11/27/01 4:39 PM Page 42

To my mind, the most comparable countries arethose in which the basic democratic political institu-tions have functioned without interruption for a fairlylong time, let’s say at least half a century, that is, since1950. Including the United States, there are twenty-two such countries in the world.2 (See Appendix B, Ta-bles 1 and 2.) Fortunately for our purposes, they arealso comparable in their relevant social and economicconditions: not a rotten apple in the bunch. Not sur-prisingly, they are mostly European or English speak-ing, with a few outliers: Costa Rica, the only LatinAmerican country; Israel, the only Middle Easterncountry; and Japan, the only Asian country.

When we examine some of the basic elements inthe constitutional structures of the advanced demo-cratic countries, we can see just how unusual the Amer-ican system is. Indeed, among the twenty-two older de-mocracies, our system is unique.3

Federal or Unitary

To begin with, among the other twenty-one countrieswe find only six federal systems, in which territorialunits—states, cantons, provinces, regions, Länder—are endowed by constitutional prescription and prac-tice with a substantial degree of autonomy and withsignificant powers to enact legislation. As in the UnitedStates, in these federal countries the basic territorialunits, whether states, provinces, or cantons, are not

t h e co n st i t u t i o n a s a m o d e l 43

03dahl.041_072 11/27/01 4:39 PM Page 43

simply legal creatures of the central government withboundaries and powers that the central governmentcould, in principle, modify as it chooses. They arebasic elements in the constitutional design and in thepolitical life of the country.

As with the United States, so too in these other fivecountries federalism was not so much a free choice asa self-evident necessity imposed by history. In most,the federal units—states, provinces, cantons—existedbefore the national government was fully democra-tized. In the extreme case, Switzerland, the consti-tuent units were already in place before the SwissConfederation itself was formed from three Alpinecantons in 1291, five centuries before America wasborn. Throughout the following seven centuries theSwiss cantons, now twenty in number,4 have retained arobust distinctiveness and autonomy. In the outlier,Belgium, federalism followed long after a unitary gov-ernment had been imposed on its diverse regionalgroups. As the brilliant period of Flemish painting,weaving, commerce, and prosperity in the sixteenthand seventeenth centuries reminds us, profound terri-torial, linguistic, religious, and cultural differences be-tween the predominantly Flemish and Walloon areasexisted long before Belgium itself became an inde-pendent country in 1830. Despite the persistent cleav-ages between the Flemish and Walloons, however,federalism did not arrive until 1993 when the three regions—Wallonia, Flanders, and Brussels—were fi-nally given constitutional status. I should point out


03dahl.041_072 11/27/01 4:39 PM Page 44

that the deep divisions between Walloons and Flemishcontinue to threaten the survival of Belgium as a singlecountry.

The second and third features follow directly fromthe existence of federalism.

Strong Bicameralism

A natural, if not strictly necessary, consequence of fed-eralism is a second chamber that provides special rep-resentation for the federal units. To be sure, unitarysystems may also have, and historically all have had, asecond chamber. However, in a democratic countrywith a unitary system, the functions of a second cham-ber are far from obvious. The question that was posedduring the American constitutional convention is boundto arise: Exactly whom or whose interests is a secondchamber supposed to represent? And just as the Fram-ers could provide no rationally convincing answer, sotoo as democratic beliefs grow stronger in democraticcountries with unitary governments, the standard an-swers become less persuasive—in fact, so unpersua-sive to the people of the three Scandinavian countriesthat they have all abolished their second chambers.Like the state of Nebraska, Norway, Sweden, andDenmark also seem to do quite nicely without them.Even in Britain, the gradual advance of democraticbeliefs created an inexorable force opposed to the his-torical powers of the House of Lords. As early as 1911


03dahl.041_072 11/27/01 4:39 PM Page 45

the Liberals wiped out the power of the Lords to veto“money bills” passed by the Commons. The continu-ing advance of democratic beliefs during the past cen-tury led in 1999 to the abolition of all but ninety-twohereditary seats, whose occupants would be elected byhereditary peers.5 The future of that ancient chamberremains in considerable doubt.

By the end of the twentieth century, then, astrongly bicameral legislature continued to exist inonly four of the advanced democratic countries, all ofthem federal: in addition to the United States, thesewere Australia, Germany, and Switzerland. Their exis-tence poses a question: What functions can and shoulda second chamber perform in a democratic country?And in order to perform its proper functions, if any,how should a second chamber be composed? As thedeliberations of the Parliamentary Commission on thefuture of the House of Lords indicate, these questionsadmit of no easy answer. It would not be surprising,then, if Britain ends up with no real second chamber atall, even if a ghostly shade of the upper house persists.

Unequal Representation

A third characteristic of federal systems is significantunequal representation in the second chamber. By un-equal representation I mean that the number of mem-bers of the second chamber coming from a federalunit such as a state or province is not proportional to


03dahl.041_072 11/27/01 4:39 PM Page 46

its population, to the number of adult citizens, or tothe number of eligible voters. The main reason, per-haps the only real reason, why second chambers existin all federal systems is to preserve and protect un-equal representation. That is, they exist primarily toensure that the representatives of small units cannotbe readily outvoted by the representatives of largeunits. In a word, they are designed to construct a bar-rier to majority rule at the national level.

To make this clear, let me extend the range of theterm unequal representation to include any systemwhere, in contrast to the principle of “one person onevote,” the votes of different persons are given unequalweights. Whenever the suffrage is denied to some per-sons within a system, we might say that their votes arecounted as zero, whereas the votes of the eligible citi-zens are counted as one. When women were deniedthe vote, a man’s vote effectively counted for one, awoman’s for nothing, zero. When property require-ments were required for the suffrage, property ownerswere represented in the legislature, those below theproperty threshold were not: like women their “votes”counted for zero. Some privileged members of Parlia-ment, like Edmund Burke, referred to “virtual repre-sentation,” where the aristocratic minority repre-sented the best interests of the entire country. But thebulk of the people who were excluded easily sawthrough that convenient fiction, and as soon as theywere able to they rejected these pretensions andgained the right to vote for their own M.P.s. In nine-


03dahl.041_072 11/27/01 4:39 PM Page 47

teenth-century Prussia, voters were divided into threeclasses according to the amount of their propertytaxes. Because each class of property owners was givenan equal number of votes irrespective of the vast dif-ference in numbers of persons in each class, a wealthyPrussian citizen possessed a vote that was effectivelyworth almost twenty times that of a Prussian worker.6

To return now to the United States: as the Ameri-can democratic credo continued episodically to exertits effects on political life, the most blatant forms ofunequal representation were in due time rejected. Yet,one monumental though largely unnoticed form of un-equal representation continues today and may wellcontinue indefinitely. This results from the famousConnecticut Compromise that guarantees two sena-tors from each state.

Imagine a situation in which your vote for yourrepresentative is counted as one while the vote of afriend in a neighboring town is counted as seventeen.Suppose that for some reason you and your friendeach change your job and your residence. As a resultof your new job, you move to your friend’s town. Forthe same reason, your friend moves to your town.Presto! To your immense gratification you now dis-cover that simply by moving, you have acquired six-teen more votes. Your friend, however, has lost sixteenvotes. Pretty ridiculous, is it not?

Yet that is about what would happen if you lived on the western shore of Lake Tahoe in California andmoved less than fifty miles east to Carson City, Nevada,


03dahl.041_072 11/27/01 4:39 PM Page 48

while a friend in Carson City moved to your commu-nity on Lake Tahoe. As we all know, both states areequally represented in the U.S. Senate. With a popula-tion in 2000 of nearly 34 million, California had twosenators. But so did Nevada, with only 2 million resi-dents. Because the votes of U.S. senators are countedequally, in 2000 the vote of a Nevada resident for theU.S. Senate was, in effect, worth about seventeen timesthe vote of a California resident. A Californian whomoved to Alaska might lose some points on climate,but she would stand to gain a vote worth about fifty-four times as much as her vote in California.7 Whetherthe trade-off would be worth the move is not for meto say. But surely the inequality in representation itreveals is a profound violation of the democratic ideaof political equality among all citizens.

Some degree of unequal representation also existsin the other federal systems. Yet the degree of unequalrepresentation in the U.S. Senate is by far the most extreme. In fact, among all federal systems, includingthose in more newly democratized countries—a totalof twelve countries—on one measure the degree ofunequal representation in the U.S. Senate is exceededonly by that in Brazil and Argentina.8

Or suppose we take the ratio of representatives inthe upper chamber to the populations of the federalunits. In the United States, for example, the two sena-tors from Connecticut represent a population of slightlyabove 3.4 million, while the two senators from its neigh-bor New York represent a population of 19 million:


03dahl.041_072 11/27/01 4:39 PM Page 49

a ratio of about 5.6 to 1. In the extreme case, the ratioof over-representation of the least populated state,Wyoming, to the most populous state, California, is justunder 70 to 1.9 By comparison, among the advanceddemocracies the ratio runs from 1.5 to 1 in Austria to40 to 1 in Switzerland. In fact, the U.S. disproportionis exceeded only in Brazil, Argentina, and Russia.10

On what possible grounds can we justify this ex-traordinary inequality in the worth of the suffrage?

A brief digression: rights and interests. A commonresponse is to say that people in states with smallerpopulations need to be protected from federal lawspassed by congressional majorities that would violatetheir basic rights and interests. Because the people instates like Nevada or Alaska are a geographical minor-ity, you might argue, they need to be protected fromthe harmful actions of national majorities. But this re-sponse immediately raises a fundamental question. Isthere a principle of general applicability that justifiesan entitlement to extra representation for some indi-viduals or groups?

In searching for an answer, we need to begin withan eternal and elementary problem in any governmen-tal unit:11 whether the unit is a country, state, munici-pality, or whatever, virtually all of its decisions will in-volve some conflict of interests among the people ofthe relevant political unit. Inevitably, almost any gov-ernmental decision will favor the interests of some cit-izens and harm the interests of others. The solution tothis problem, which is inherent in all governmental


03dahl.041_072 11/27/01 4:39 PM Page 50

units, is ordinarily provided in a democratic system bythe need to secure a fairly broad consent for its deci-sions by means, among other things, of some form ofmajority rule. Yet if decisions are arrived at by majorityrule, then the possibility exists, as Madison and manyothers have observed, that the interests of any minor-ity will be damaged by a majority. Sometimes, fortu-nately, mutually beneficial compromises may be found.But if the interests of a majority clash irreconcilablywith those of a minority, then the interests of that mi-nority are likely to be harmed.

Some interests, however, may be protected fromthe ordinary operation of majority rule. To a greater orlesser degree, all democratic constitutions do so.

Consider the protections that all Americans enjoy,not just in principle but substantially in practice aswell. First, the Bill of Rights and subsequent amend-ments provide a constitutional guarantee that certainfundamental rights are protected whether a citizenlives in Nevada or California, Rhode Island or Massa-chusetts, Delaware or Pennsylvania. Second, an im-mense body of federal law and judicial interpretationbased on constitutional provisions enormously extendsthe domain of protected rights—probably far beyondanything the Framers could have foreseen. Third, theconstitutional division of powers in our federal systemprovides every state with an exclusive or overlappingdomain of authority on which a state may draw inorder to extend even further the protections for theparticular interests of the citizens of that state.


03dahl.041_072 11/27/01 4:39 PM Page 51

The basic question. Beyond these fundamentaland protected rights and interests, do people in thesmaller states possess additional rights or interests thatare entitled to protection from policies supported bynational majorities? If so, what are they? And on whatgeneral principle can their special protection be justi-fied? Surely they do not include a fundamental rightto graze sheep or cattle in national forests or to extractminerals from public lands on terms that were setmore than a century ago. Why should geographical lo-cation endow a citizen or group with special rights andinterests, above and beyond those I just indicated, thatshould be given additional constitutional protection?

If these questions leave me baffled, I find myselfin good company. “Can we forget for whom we areforming a government?” James Wilson asked at theConstitutional Convention. “Is it for men, or for theimaginary beings called States?” Madison was equallydubious about the need to protect the interests ofpeople in the small states. “ Experience,” he said, “sug-gests no such danger. . . . Experience rather taught acontrary lesson. . . . The states were divided into dif-ferent interests not by their differences in size, but byother circumstances.”12

Two centuries of experience since Madison’s timehave confirmed his judgment. Unequal representationin the Senate has unquestionably failed to protect thefundamental interests of the least privileged minorities.On the contrary, unequal representation has some-times served to protect the interests of the most privi-


03dahl.041_072 11/27/01 4:39 PM Page 52

leged minorities. An obvious case is the protection ofthe rights of slaveholders rather than the rights of theirslaves. Unequal representation in the Senate gave ab-solutely no protection to the interests of slaves. On thecontrary, throughout the entire pre–Civil War periodunequal representation helped to protect the interestsof slave owners. Until the 1850s equal representationin the Senate, as Barry Weingast has pointed out, gavethe “the South a veto over any policy affecting slavery.”Between 1800 and 1860 eight anti-slavery measurespassed the House, and all were killed in the Senate.13

Nor did the Southern veto end with the Civil War. Afterthe Civil War, Senators from elsewhere were compelledto accommodate to the Southern veto in order to securethe adoption of their own policies. In this way theSouthern veto not only helped to bring about the end ofReconstruction; for another century it prevented thecountry from enacting federal laws to protect the mostbasic human rights of African Americans.

So much for the alleged virtues of unequal repre-sentation in the Senate.

Suppose for a moment we try to imagine that weactually wanted the constitution to provide special pro-tection to otherwise disadvantaged minorities by giv-ing them extra representation in the Senate. What mi-norities most need this extra protection? How wouldwe achieve it? Would we now choose to treat certainstates as minorities in special need of protection sim-ply because of their smaller populations? Why wouldwe want to protect these regional minorities and not


03dahl.041_072 11/27/01 4:39 PM Page 53

other, far weaker minorities? To rephrase James Wil-son’s question in 1787: Should a democratic govern-ment be designed to serve the interests of “the imagi-nary beings called States,” or should it be designedinstead to serve the interests of all its citizens consid-ered as political equals?

As I have said, the United States stands out amongtwenty-two comparable democratic countries for thedegree of unequal representation in its upper chamber.Of the half dozen that have federal systems and anupper house designed to represent the federal units,none come even close to the United States in the ex-tent of its unequal representation in its upper house.

We begin to see, then, that our constitutional sys-tem is unusual. As we continue our exploration we shalldiscover that it is not merely unusual. It is one of a kind.

Strong Judicial Review of National Legislation

Not surprisingly, other federal systems among the olderdemocracies also authorize their highest national courtsto strike down legislation or administrative actions bythe federal units—states, provinces, and the like—that are contrary to the national constitution. The casefor the power of federal courts to review state actionsin order to maintain a federal system seems to mestraightforward, and I accept it here. But the authorityof a high court to declare unconstitutional legislationthat has been properly enacted by the coordinate con-


03dahl.041_072 11/27/01 4:39 PM Page 54

stitutional bodies—the parliament or in our system theCongress and the president—is far more controversial.

If a law has been properly passed by the law-making branches of a democratic government, whyshould judges have the power to declare it unconstitu-tional? If you could simply match the intentions andwords of the law against the words of the constitution,perhaps a stronger case could be made for judicial re-view. But in all important and highly contested cases,that is simply impossible. Inevitably, in interpreting theconstitution judges bring their own ideology, biases,and preferences to bear. American legal scholars havestruggled for generations to provide a satisfactory ra-tionale for the extensive power of judicial review thathas been wielded by our Supreme Court. But thecontradiction remains between imbuing an unelectedbody—or in the American case, five out of nine jus-tices on the Supreme Court—with the power to makepolicy decisions that affect the lives and welfare ofmillions of Americans. How, if at all, can judicial re-view be justified in a democratic order? I’ll discussthat question in my last chapter.

Meanwhile, let me return to another aberrant as-pect of the American constitutional system.

Electoral Systems

Earlier I explained that I wanted to use the term con-stitutional system because some arrangements that are


03dahl.041_072 11/27/01 4:39 PM Page 55

not necessarily specified in a country’s constitutionaldocument interact so strongly with the other institu-tions that we can usefully regard them as a part of thecountry’s constitutional arrangements. In that spirit, wemight want to reflect on the peculiarities of our elec-toral system, which, natural as it may seem to us, is of aspecies rare to the vanishing point among the advanceddemocratic countries. Closely allied with it is an equallyrare bird, our much revered two-party system.

To be sure, our electoral system was not the doingof the Framers, at least directly, for it was shaped lessby them than by British tradition. The Framers simplyleft the whole matter to the states and Congress,14

both of which supported the only system they knew,one that had pretty much prevailed in Britain, in thecolonies, and in the newly independent states.

The subject of electoral systems is fearfully com-plex and for many people fearfully dull as well. I shalltherefore employ a drastic oversimplification, but onesufficient for our purposes. Let me simply divide elec-toral systems into two broad types, each with a variantor two. In the one we know best, typically you can castyour vote for only one of the competing candidates,and the candidate with the most votes wins. In theusual case, then, a single candidate wins office by gain-ing at least one more vote than any of his or her oppo-nents. We Americans tend to call this one-vote margina plurality; elsewhere, to distinguish it from an ab-solute majority it may be called a relative majority. To describe our system, American political scientists


03dahl.041_072 11/27/01 4:39 PM Page 56

sometimes employ the cumbersome expression “singlemember district system with plurality elections.” I pre-fer the British usage: on the analogy of a horse racewhere the winner needs only a fraction of a nose-length to win, the British tend to call it the “first-past-the-post” system.

If voters were to cast their ballots in the same pro-portion in every district, the party with the most voteswould win every seat. In practice, as a result of varia-tions from district to district in support for candidates,a second party generally manages to gain some seats,although its percentage of seats will ordinarily besmaller than its percentage of votes. But the represen-tation of third parties usually diminishes to the vanish-ing point. In short, first-past-the-post favors two-partysystems.

The main alternative to first-past-the-post is pro-portional representation. As the name implies, propor-tional representation is designed to ensure that votersin a minority larger than some minimal size—say, 5percent of all voters—will be represented more or lessin proportion to their numbers. For example, a groupconsisting of 20 percent of all voters might win prettyclose to 20 percent of the seats in the parliament.Consequently, countries with proportional represen-tation systems are also very likely to have multipartysystems in which three, four, or more parties are rep-resented in the legislature. In short, although the rela-tionship is somewhat imperfect, in general a countrywith first-past-the-post is likely to have a two-party


03dahl.041_072 11/27/01 4:39 PM Page 57

system and a country with proportional representationis likely to have a multiparty system.

In the most common system of proportional repre-sentation, each party presents voters with a list of itscandidates; voters cast their votes for a party’s candi-dates; each party is then awarded a number of seatsroughly in proportion to its overall share of the vote.Countries with a list system may also permit voters toindicate their preferences among the party’s candi-dates. The party’s seats are then filled by the candi-dates who are most preferred by the voters. Twelve ofthe twenty-two advanced democratic countries employthe list system of proportional representation, and an-other six use some variant of it. (See Appendix B,Table 3.)

Of the four countries without proportional repre-sentation, France avoids one of the defects of single-member districts by providing that in parliamentarydistricts where no candidate receives an absolute ma-jority of votes, a second election will be held in whichthe two candidates with the highest number of votescompete. This run-off, two-round, or double-ballot sys-tem, as it is variously called, thereby ensures that allthe members have been elected by a majority of thevoters in their constituency.

This leaves the three oddballs with first-past-the-post, a plurality system in single member districts:Canada, the United Kingdom, and the United States.Even in the United Kingdom, the original source onwhich the Americans drew, the traditional system was


03dahl.041_072 11/27/01 4:39 PM Page 58

replaced by proportional representation in the 1999elections to the newly created legislative bodies inScotland and Wales. Four parties won seats in theScottish Parliament, and four too in the Welsh Assem-bly. What is more, the Independent Commission onthe Voting System set up by the Labor Party in 1997 torecommend an alternative to first-past-the-post pro-posed in its report a year later that members of theHouse of Commons be elected by means of a propor-tional representation system—a hybrid, to be sure,but one that would ensure greater proportionality be-tween votes and seats in that ancient house.15 It is al-together possible that one day not far off, Britain will be added to the list of proportional representationcountries, leaving only Canada and the United Statesamong the advanced democracies with first-past-the-post.

Although few Americans know much about experi-ence in the other advanced democratic countries withproportional representation and multiparty systems,they seem to have strong prejudices against both. Un-willing to conceive of an alternative to first-past-the-post and under pressure to ensure fairer representa-tion for minorities in state legislatures and Congress,our legislatures and federal courts in recent years havesometimes gerrymandered weirdly shaped districts. . .well, yes, rather like a salamander. But neither legisla-tures nor courts seem willing to give serious thoughtto some form of proportional representation as quitepossibly a better alternative.


03dahl.041_072 11/27/01 4:39 PM Page 59

The extent to which we take first-past-the-post forgranted was clearly revealed in 1993, when it was dis-covered that a well-qualified candidate to head theCivil Rights Division of the Department of Justice hadwritten an article in a law journal suggesting that arather sensible system of proportional representationmight be worth considering as a possible solution to theproblem of securing more adequate minority repre-sentation.16 From the comments the author’s innocentheresy generated, you might have thought that she hadburned the American flag on the steps of the SupremeCourt. Her candidacy, naturally, was stone dead.

First-past-the-post was the only game in town in1787 and for some generations thereafter. Like the locomotive, proportional representation had not yetbeen invented. It was not fully conceived until themid-nineteenth century when a Dane and two Eng-lishmen—one of them John Stuart Mill—provided asystematic formulation. Since then it has become thesystem overwhelmingly preferred in the older democ-racies.

After more than a century of experience with otheralternatives, isn’t it time at last to open our minds tothe possibility that first-past-the-post may be just finefor horse races but might not be best for elections in alarge and diverse democratic country like ours? Mightwe not also want to consider the possible advantagesof a multiparty system?

I do not say that we should necessarily make thesechoices. But should we not at least give them serious


03dahl.041_072 11/27/01 4:39 PM Page 60

consideration? Shouldn’t we ask ourselves this ques-tion: What kind of electoral and party systems wouldbest serve democratic ends?

Party Systems

Nearly a half-century ago, a French political scientist,Maurice Duverger, proposed what came to be calledDuverger’s Law: first-past-the-post electoral systemstend to result in two-party systems. Conversely, pro-portional representation systems are likely to producemultiparty systems.17 Although the causal relation maybe more complex than my brief statement of Du-verger’s Law suggests,18 a country with a proportionalrepresentation system is likely to require coalition gov-ernments consisting of two or more parties. In a coun-try with a first-past-the-post electoral system, however,a single party is more likely to control both the execu-tive and the legislature. Thus in countries with propor-tional representation–multiparty systems and coalitiongovernments, minorities tend to be represented moreeffectively in governing. By contrast, in countries withfirst-past-the-post and two-party systems, the govern-ment is more likely to be in the hands of a single partythat has gained a majority of seats in the parliamentand the most popular votes, whether by an outrightmajority, or more commonly, a plurality. To distinguishthe two major alternatives, I’ll refer to the propor-tional representation–multiparty countries as “propor-


03dahl.041_072 11/27/01 4:39 PM Page 61

tional” and countries with first-past-the-post electoralsystems and only two major parties as “majoritarian.”19

Where does the United States fit in? As usual: inneither category. It is a mixed system, a hybrid, nei-ther predominantly proportional nor predominantlymajoritarian. (See Appendix B, Table 4.) I am going toreturn to the American hybrid in Chapter 5, but threebrief observations may help to put it in perspectivehere. First, the Framers had no way of knowing aboutthe major alternatives to first-past-the-post, much lessfully understanding them. Second, since the Framers’time most of the older and highly stable democraticcountries have rejected first-past-the-post and optedinstead for proportional systems. Third, our mixed de-sign contributes even further to the unusual structureof our constitutional system.

Our Unique Presidential System

As we make our way through the list of countries thatshare some constitutional features with the UnitedStates, the list, short to begin with, diminishes even fur-ther. By the time we reach the presidency the UnitedStates ceases to be simply unusual. It becomes unique.

Among the twenty-two advanced democracies, theUnited States stands almost alone in possessing a singlepopularly elected chief executive endowed with im-portant constitutional powers—a presidential system.Except for Costa Rica, all the other countries govern


03dahl.041_072 11/27/01 4:39 PM Page 62

themselves with some variation of a parliamentary sys-tem in which the executive, a prime minister, is chosenby the national legislature. In the mixed systems ofFrance and Finland, most of the important constitu-tional powers are assigned to the prime minister, butan elected president is also provided with certain pow-ers—chiefly over foreign relations. This arrangementmay lead, as in France, to a president from one majorparty and a prime minister from the opposing party, asituation that with a nice Gallic touch the French call“cohabitation.” Yet even allowing for the French andFinnish variations, none of the other advanced demo-cratic countries has a presidential system like ours.

Why is this? The question breaks down into sev-eral parts. Why did the Framers choose a presidentialsystem? Why didn’t they choose a parliamentary sys-tem? Why have all the other advanced democraticcountries rejected our presidential system? Why havethey adopted some variant of a parliamentary systeminstead, or as in France and Finland a system that ispredominantly parliamentary with an added touch ofpresidentialism?

To answer these questions in detail would go be-yond our limits here. But let me sketch a brief answer.

Before I do so, however, I want to admonish younot to cite the explanation given in the Federalist Papers. These were very far from critical, objectiveanalyses of the constitution. If we employ a dictionarydefinition of propaganda as “information or ideas me-thodically spread to promote or injure a cause, nation,


03dahl.041_072 11/27/01 4:39 PM Page 63

etc.,” then the Federalist Papers were surely propa-ganda. They were written post hoc by partisans—Alexander Hamilton, John Jay, and James Madison—who wanted to persuade doubters of the virtues of theproposed constitution in order to secure its adoptionin the forthcoming state conventions. Although theywere very fine essays indeed, and for the most partmuch worth reading today, they render the work ofthe convention more coherent, rational, and compel-ling than it really was. Ironically, by the way, the taskof explaining and defending the Framers’ design forthe presidency was assigned to Hamilton, who hadsomewhat injudiciously remarked in the Conventionthat as to the executive, “The English model was theonly good one on this subject,” because “the heredi-tary interest of the king was so interwoven with that ofthe nation. . . and at the same time was both suffi-ciently independent and sufficiently controuled [sic],to answer the purpose.” He then proposed that the ex-ecutive and one branch of the legislature “hold theirplaces for life, or at least during good behavior.”20 Per-haps as a result of these remarks, Hamilton seems tohave had only a modest influence in the Conventionon that matter or any other.

How it came about. What is revealed in the mostcomplete record of the Convention21 is a body floun-dering in its attempts to answer an impossibly difficultquestion: How should the chief executive of a republicbe selected, and what constitutional powers should beassigned to the executive branch? The question was im-


03dahl.041_072 11/27/01 4:39 PM Page 64

possibly difficult because, as I emphasized in the pre-vious chapter, the Framers had no relevant model ofrepublican government to give them guidance. Mostof all, they lacked any suitable model for the executivebranch. To be sure, they could draw on the sacred doc-trine of “separation of powers.” Not surprisingly, thereferences to that doctrine recorded in Madison’s noteswere all positive. And up to a point, its implicationswere obvious: a republic would need an independentjudiciary, a bicameral legislature consisting of a popularhouse and some kind of second chamber to check thepopular house, and an independent executive.

But how was the independent executive to be cho-sen? How independent of the legislature and of thepeople should he be? How long should his term of of-fice be? (“He” is, of course, the language of Article IIand, like most Americans until recently, the only waythe Framers could conceive of the office.) The Britishconstitution was a helpful model for the Framers insome respects. But as a solution to the problem of theexecutive, it utterly failed them. Despite the respect ofthe delegates for many aspects of the British constitu-tion, a monarchy was simply out of the question.22

Even so, they might have chosen a democratic ver-sion of the parliamentary system, as the other evolvingEuropean democracies were to do. Although they wereunaware of it, even in Britain a parliamentary systemwas already evolving. Why then didn’t the Framerscome up with a republican version of a parliamentarysystem?


03dahl.041_072 11/27/01 4:39 PM Page 65

Well, they almost did. It has been too little empha-sized, I think, that the Framers actually came veryclose to adopting something like a parliamentary sys-tem. What is more, it is far from clear, to me at least,why they rejected it and ended up instead with a pres-idential system.23 One obvious solution—even moreobvious to us today than it would have been in 1787—was to allow the national legislature to choose the ex-ecutive. In fact, throughout most of the Conventionthis was their favored solution. Right off the bat onJune 2, only two weeks after the Convention opened,the Virginia delegation, which contained some of thebest minds and most influential delegates, proposedthat the national executive should be chosen by the na-tional legislature. In Madison’s notes, the subsequentcourse of that proposal and the alternatives to it hasleft a fascinating and often mystifying trail.

The meandering trail they pursued, as best I canreconstruct it, looks something like this.24 On threeoccasions—July 17, July 24, and July 26—the dele-gates vote for the selection of the president by “the na-tional legislature,” the first time by a unanimous vote,the last by a vote of 6–3. With one exception everyother alternative is defeated by substantial majorities:in a puzzling detour on July 19, with Massachusetts di-vided, they vote 6–3 for electors appointed by thestate legislatures. On July 26, their favored solution,election by the national legislature, is forwarded to aCommittee on Detail. On August 6 the committee dulyreports in favor of election by the national legislature.


03dahl.041_072 11/27/01 4:39 PM Page 66

On August 24 two other alternatives fail once again. Anew committee to consider the issue reports back onSeptember 4. By now the delegates are eager to windup a convention that has already gone on for threemonths. In contradiction to the recommendation ofthe previous committee, however, this one recom-mends that the executive be chosen by electors ap-pointed by the state legislatures. Two days later, withnine states in favor and only two opposed, the impa-tient delegates adopt this solution.

Well, not exactly. What they adopt actually statesthat: “Each state shall appoint, in such manner as thelegislature thereof may direct, a number of electors,equal to the whole number of Senators and representa-tives to which the State may be entitled in Congress.”Whatever the Framers intend by these words, they willoffer a huge opportunity for the democratic phase ofthe American revolution to democratize the presidency.

Ten days after they agree on this provision, theconstitution is signed and the Convention adjourns.

What this strange record suggests to me is a groupof baffled and confused men who finally settle on a so-lution more out of desperation than confidence. Asevents were soon to show, they had little understand-ing of how their solution would work out in practice.

So the question remains with no clear answer: Why,finally, did they fail to adopt the solution they hadseemed to favor, a president elected by the Congress, asort of American version of a parliamentary system? Thestandard answer no doubt has some validity: they feared


03dahl.041_072 11/27/01 4:39 PM Page 67

that the president might be too beholden to Congress.And all the other alternatives seemed to them worse.

Among these alternatives was election by thepeople, which had been twice rejected overwhelm-ingly. Yet it was this twice-rejected solution, electionby the people, that was quickly adopted de facto dur-ing the democratic phase of the American revolution.

How their solution failed. Perhaps in no part oftheir work did the Framers fail more completely todesign a constitution that would prove acceptable to ademocratic people. As I have mentioned, their hopefor a group of electors who might exercise their inde-pendent judgments about the best candidate to fill theoffice came a cropper following the election of 1800.But as I shall describe in the next chapter, more wasstill to come. If the election of 1800 first revealed howinappropriate the electoral college was in a democraticorder, the presidential election of 2000, two centurieslater, dramatized for all the world to witness the con-flict between the Framers’ constitution and the demo-cratic ideal of political equality.

Ironically, had they adopted the Virginia Plan andplaced the choice of the chief executive in the hands ofthe legislature, as would become the practice in parlia-mentary systems, the Framers would have put a bitmore distance between the people and the presidentthan their solution provided in practice. Here again, in1787 they could not anticipate a constitutional designthat was yet to evolve fully in Britain and, even later, inother countries on the path to democracy.


03dahl.041_072 11/27/01 4:39 PM Page 68

The continuing democratic revolution would bringabout an even more profound change in the presi-dency. However deftly Jefferson steered the Congressas he rode the tide of the democratic revolution, henever publicly challenged the standard view that theonly legitimate representative of the popular will wasthe Congress, not the president. Nor did any of hissuccessors, Madison, Monroe, John Quincy Adams, laydown such a claim.

Andrew Jackson did just that. In justifying his useof the veto against Congressional majorities, as theonly national official who had been elected by all thepeople and not just by a small fraction, as were Sena-tors and Representatives, Jackson insisted that he alonecould claim to represent all the people. Thus Jacksonbegan what I have called the myth of the presidentialmandate: that by winning a majority of popular (andpresumably electoral) votes, the president has gaineda “mandate” to carry out whatever he had proposedduring the campaign.25 Although he was bitterly at-tacked for this audacious assertion, which not all laterpresidents supported, it gained credibility from its re-assertion by Lincoln, Cleveland, Theodore Roosevelt,and Wilson and was finally nailed firmly in place byFranklin Roosevelt.

Whatever we may think of the validity of theclaim—I am inclined to think it is little more than amyth created to serve the political purposes of ambi-tious presidents—it is simply one part of a transforma-tion of the presidency in response to democratic ideas


03dahl.041_072 11/27/01 4:39 PM Page 69

and beliefs that has produced an office completely dif-ferent from the office that the Framers thought theywere creating, vague and uncertain as their intentionsmay have been.

And a good thing, too, you may say. But if you ap-prove of the democratization of the presidency—or, as Iwould prefer to say, its pseudo-democratization—aren’tyou suggesting in effect that the constitutional systemshould be altered to meet democratic requirements?

Why other countries became parliamentary de-mocracies. There is still one more reason why theFramers didn’t choose a parliamentary system. Theyhad no model to inspire them. One hadn’t yet been in-vented.

The British constitutional system they knew, and insome respects admired, was already on its way to his-tory’s attic of abandoned or failed constitutions. Al-though no one saw it clearly in 1787, even at the timeof the Convention the British constitution was under-going rapid change. Most important, the monarch wasswiftly losing the power to impose a prime minister onthe parliament. The contrary assumption was gainingstrength: that a prime minister must receive a vote ofconfidence from both houses of parliament, and thathe must resign if and whenever he lost their confi-dence. But this profound change in the British consti-tution did not become fully manifest until 1832, toolate for the Framers to see its possibilities.

In addition, there was the problem of a monarch.How could a country have a parliamentary system


03dahl.041_072 11/27/01 4:39 PM Page 70

without a symbolic head of state who would performceremonial functions, symbolize the unity of the coun-try, and help to confer legitimacy on the parliament’schoice by anointing him as prime minister? After theevolution of a parliamentary system in Britain, in duetime monarchies also helped the Swedes, the Danes,and the Norwegians—and much later Japan andSpain—to move to a parliamentary system that themonarchy helped to legitimize. But in 1787 the fulldevelopment of parliamentary democracy in countrieswith a monarchy was still a long way off. For Ameri-cans, a monarch, even a ceremonial monarch, was com-pletely out of the question. So why didn’t they split thetwo functions, ceremonial and executive, by creating atitular head of state to serve in the place of a ceremo-nial monarch, and a chief executive, the equivalent ofa prime minister, to whom executive functions wouldbe assigned? Although that arrangement may seemobvious enough to us now, for the Framers in 1787 itwas even more distant than the system that was gradu-ally evolving in Britain, the country they knew best. Itwas not until after 1875 and the installation of theThird Republic in France that the French evolved asolution that would later be adopted in many other de-mocratizing countries: a president elected by the par-liament, or in some cases by the people, who serves asformal head of state, and a prime minister chosen byand responsible to the parliament, who serves as theactual chief executive. But for the Framers this inven-tion, which now seems obvious enough to us, was al-


03dahl.041_072 11/27/01 4:39 PM Page 71

most as far off and about as difficult to imagine, per-haps, as a transcontinental railroad.

Without intending to do so, then, the Framers cre-ated a constitutional framework that under the drivingimpact of the continuing American Revolution woulddevelop a presidency radically different from the onethey had in mind. In time American presidents wouldgain office by means of popular elections—a solutionthe Framers rejected and feared—and by combiningthe functions of a head of state with those of a chiefexecutive the president would be the equivalent ofmonarch and prime minister rolled into one.

I can’t help wondering whether the presidencythat has emerged is appropriate for a modern demo-cratic country like ours.

� � �

SO: AMONG THE OLDER DEMOCRACIES OUR CONSTITU-tional system is not just unusual. It is unique.

Well, you might say, being unique isn’t necessarilybad. Perhaps our constitutional system is better for it.

Better by what standards? Is it more democratic?Does it perform better in many ways? Or worse?

These questions are by no means easy to answer—probably impossible to answer with finality. But beforeturning to them, we need to take one more look at thatanomalous vestige of the Framers’ work, the electoralcollege.


03dahl.041_072 11/27/01 4:39 PM Page 72

chapter 4

Electing the President

ON THE NIGHT OF NOVEMBER 7, 2000, A DRAMA

opened in the United States that absorbed theattention of millions of people until the cur-

tain came down six weeks later. The nation was fo-cused once again on an anomalous institution that hadoriginated in the Framers’ search for a suitable way toelect the new republic’s chief executive. This was theelectoral college, by means of which the presidencywas won—not for the first time and perhaps not thelast—by a candidate with fewer votes than his rival.1

As we saw in the previous chapter, the Framerswere baffled by how to conceive of the executive in a republic. How should a republican executive bechosen? During the final debates over the electoralcollege, James Wilson commented: “This subject has

04dahl.073_090 11/27/01 4:40 PM Page 73

greatly divided the House, and will also divide thepeople out of doors. It is in truth the most difficult ofall on which we have had to decide.”2 Three monthsafter the Convention had adjourned, when his memo-ries were still fresh, he made the same observation tohis fellow Pennsylvanians who had assembled to ratifythe new constitution: “The convention, sir, were per-plexed with no part of this plan, so much as with themode of choosing the president of the United States.”3

Every solution seemed worse than the rest. Thearrangement they finally cobbled together at the lastminute was adopted more out of desperation, perhaps,than out of any great confidence in its success. So whydid the delegates finally give their approval to theelectoral college? Probably the best answer to ourquestion would be: the Framers settled on an electoralcollege because they had run out of alternatives.

How the Electoral College Came About

We have discussed how during the three summermonths the delegates considered and rejected the mostobvious possibilities. From the scanty information avail-able, it is possible to add a few more facts.

On August 6, the Committee on Detail brings in adraft of the Constitution that calls for the election ofthe president by Congress. This proposal finds no tak-ers. As late as August 24, despite mounting pressures

74 e l e c t i n g t h e p r e s i d e n t

04dahl.073_090 11/27/01 4:40 PM Page 74

to complete their work and adjourn, the delegates votedown every plausible alternative. Even a motion thatthe president “shall be chosen by electors” fails to gaina majority of votes. Unable to agree, on the last day ofAugust the delegates turn the problem of choosingthe president over to yet another committee, this onedrawn from each of the eleven state delegations. Fourdays later the committee offers one of the solutions thedelegates had already rejected earlier: “Each State shallappoint in such manner as its Legislature may direct, anumber of electors equal to the whole number of Sen-ators and members of the House of Representatives towhich the State may be entitled in the Legislature.”4

The delegates, no doubt weary of their task andeager to finish, find one remaining flaw. The commit-tee has proposed that in case of a tie vote in the elec-toral college, the choice will be made by the Senatefrom the five highest candidates. Objection! Many del-egates believe, it appears, that the competition for thepresidency will usually lead to more than two leadingcandidates—three, four, five, more. If their conjec-tures are correct, then the election of the presidentwill frequently, perhaps regularly, fall to the Senate, abody they expect to be highly powerful and ratheraristocratic. Wilson forcefully puts the case againstconferring that power on the Senate. Combining thisadditional power “with other parts of the plan,” hesaid, he was “obliged to consider the whole as having adangerous tendency to aristocracy; as throwing a dan-

e l e c t i n g t h e p r e s i d e n t 75

04dahl.073_090 11/27/01 4:40 PM Page 75

gerous power into the hands of the Senate.”5 Othersagree, and by a lopsided vote the delegates relocatethe power in the hands of the more popular branch.

Thus was the electoral college born.

Why the Electoral College?

To return now to our question. Why did the Framerssettle on this solution? The usual answer runs like this:they wanted to remove the choice of the Presidentfrom the hands of popular majorities and to place theresponsibility in the hands of a select body of wise,outstanding, and virtuous citizens—as they clearly sawthemselves, a cynic might add. The main source of thestandard view seems to be the Federalist No. 68, writ-ten by Hamilton. “The immediate election should bemade by men most capable of analyzing the qualitiesadapted to the station, and acting under circumstancesfavorable to deliberation, and to a judicious combina-tion of all the reasons and inducements which wereproper to govern their choice.”6 Nothing, of course,could be further from the reality of the electoral col-lege as it swiftly developed. And even Hamilton ap-pears to have misunderstood the mechanics of theelectoral college in one respect: He assumed that “thepeople of each State shall choose . . . [the] electors.”7

But what the Constitution actually provided, as we justsaw, was that the power to determine how the electorswere to be chosen was assigned to the state legisla-


04dahl.073_090 11/27/01 4:40 PM Page 76

tures. And most states did not initially assign that re-sponsibility to the people.

The fullest available explanation for the commit-tee’s proposal is one offered by Gouverneur Morris, amember (and probably a highly influential member)of the committee, who ran through the now familiarobjections to alternative solutions:

Congress: “The danger of intrigue & faction if the ap-pointmt. [sic] Should be made by the Legislature.”“No body appeared to be satisfied with an appoint-ment by the Legislature.” “The indispensable neces-sity of making the Executive independent of the Leg-islature.”

The People: “Many were anxious [i.e., in fear]8 even foran immediate choice by the people.”

Cabals and Corruption: “As the Electors would vote atthe same time throughout the U.S. and at so great adistance from each other, the great evil of cabalwould be avoided. It would be impossible to corruptthem.”9

Nothing here of Hamilton’s later rationalization,though something like it might have been on theminds of some delegates.

Failure

No part of the constitution revealed the flaws in its de-sign more quickly than the provision for the electoralcollege. Within a dozen years, the election of 1800 hadalready displayed two of its defects. The more serious


04dahl.073_090 11/27/01 4:40 PM Page 77

at the time, but easily corrected, was the failure of theFramers to distinguish adequately between electing thepresident and electing the vice-president.10 At a con-gressional caucus in May 1800, the Republicans hadunanimously agreed on the nominations of ThomasJefferson as president and Aaron Burr as vice presi-dent. But later on in the electoral college, the voteswere split among five candidates: Jefferson and Burrtied for president with 73 votes each; of two Federal-ist candidates, the incumbent, President John Adams,gained 65, and Charles Cotesworth Pinckney, 64; JohnJay, the governor of New York, received 1 vote. As theConstitution prescribed, the deadlock was referred tothe House. After thirty-six ballots taken over a week’stime, Jefferson finally prevailed with the votes of tenof the sixteen states.

What had not been foreseen by the Framers—a tiebetween the top two candidates—was now made ob-vious.11 The solution was equally obvious. The TwelfthAmendment, requiring separate ballots for presidentand vice president, was swiftly adopted, in time for theelection of 1804.12

Although this failure in the Framers’ design of theelectoral college was easily corrected, the other thatwas clearly revealed by the election of 1800 remained.It remains still. That presidential election shatteredwhatever hopes the delegates to the Convention mayhave entertained that the electoral college would serveas an independent body free of the supposed vices ofpopular election. Party politics—partisan politics, if you


04dahl.073_090 11/27/01 4:40 PM Page 78

will—had transformed the electors into party agents, arole that except for an aberrant member would con-tinue to be played. The privilege of serving as an elec-tor would typically be awarded not to leading citizenswilling to express their independent judgments, as theFramers may have assumed, but to party loyalists, andusually minor ones at that. Thus the development ofpolitical parties and party loyalties turned the elabo-rate machinery of the electoral college into little morethan a way of counting votes.

What is more, in a change that I’ll come back to ina moment, the democratizing phase of the Americanrevolution speedily put the choice of the presidentwhere the Framers had explicitly refused to lodge it: inthe hands of the people (that is, the white male voters).

So endeth the aristocratic pretensions of the elec-toral college.

Inherent Democratic Defects

Even when the electors were chosen by popular elec-tion, however, three undemocratic features remainedinherent features of the electoral college.

Popular votes versus electoral votes. First, the can-didate with the greatest number of popular votes—aplurality or even an outright majority—might not re-ceive a majority of electoral votes and thus might failto be chosen president. Four presidential elections—including that of 2000—have led to just such an out-


04dahl.073_090 11/27/01 4:40 PM Page 79

come. In the most famous instance, the election of1876, a candidate with an outright majority of popularvotes lost the presidency. The crisis began when SamuelJ. Tilden, the Democratic candidate, won 51 percentof the popular votes and yet failed to gain a clear anduncontested majority of votes in the electoral collegeagainst his Republican opponent, Rutherford B. Hayes.In the complex political shenanigans that ensued,13

Southern Democrats in Congress extorted from Hayesa promise to withdraw federal troops from the South,and they promised in turn to respect Negro rights(which, needless to say, they never did). Hayes wasthen awarded 185 electoral votes to Tilden’s 184. In thewords of one account: “The country acquiesced. Thusended a crisis that could have resulted in civil war.”14

In an additional three elections in which the win-ning candidate gained fewer popular votes than hisrival, none of the candidates gained a majority of popu-lar votes. In these cases, which include the election of2000, the votes cast for third party candidates deprivedboth major party candidates of a popular majority.15

Winning with a minority of popular votes. In fact,winning the presidency with only a minority of popu-lar votes has been a fairly common occurrence. In atotal of eighteen elections, candidates have gained thepresidency without winning a majority of popularvotes. (See Appendix B, Figure 1.) Overall, then, inone out of every three presidential elections the high-est office in the land has been awarded to a candidatechosen by a minority of voters. In a close election


04dahl.073_090 11/27/01 4:40 PM Page 80

where third party candidates pick up some votes, as inthe election of 2000, this outcome is highly probable.

Losing though preferred by a majority. In somecases where no candidate receives a majority of popu-lar votes, if voters’ second choices had been taken intoaccount (as they can be in some electoral systems), orif a run-off had occurred between the two highest can-didates, it is altogether possible that the outcomemight have gone the other way. If voters had beenprovided with an opportunity to express their secondchoices in the contested election of 2000, most votesof the major third party candidate, Ralph Nader,would likely have gone to Vice President Gore, inwhich case Gore would have gained the presidency.

Unequal representation of voters. To these de-fects in the electoral college we can add yet one more.Because each state is entitled to “a number of electorsequal to the whole number of Senators and Represen-tatives” from that state, unequal representation in theSenate plays itself out once again. Although the effectsare somewhat diluted by the inclusion of Representa-tives in the total, the inequality in the weight of votesin the electoral college is still strong. The vote of aWyoming resident, for example, is worth almost fourtimes the vote of a California resident in the electoralcollege. Where the number of residents for each elec-tor runs from 165 thousand to a little more than 300thousand in the ten smallest states, in the ten largest itranges from 586 thousand in Georgia to 628 thousandin California. The ten smallest states each choose two


04dahl.073_090 11/27/01 4:40 PM Page 81

to three times as many electors as they would if astate’s electors were strictly in proportion to its popu-lation.16 (See Appendix B, Figure 2.)

A Remediable Defect: Winner Take All

The democratic deficiencies inherent in the electoralcollege were compounded by another item that was de-liberately tacked on. For the first elections the states ex-perimented with a variety of methods for choosing theelectors. An extreme example was Massachusetts, which“altered its system of selecting electors no fewer thanseven times in the first ten presidential elections, oftento suit short-term partisan interests.”17 The two majoroptions were to lodge the choice of electors in the legis-lature or to give the choice to the people; popularchoice in turn could be made either in districts—oneelector per district—or at large, with the winner takingall the state’s electoral votes. Giving the legislature thepower to choose the delegates ran sharply contrary tothe democratizing currents of the time, and popularelection soon predominated. By 1832 only South Car-olina continued to lodge the choice with the legislature,a practice it finally abandoned during the Civil War.18

But of the two systems of popular election, winner-take-all soon predominated as political leaders concludedthat by concentrating all the state’s electoral votes in asingle slate, they could enhance their weight in the elec-toral college and thus their influence on the elections.


04dahl.073_090 11/27/01 4:40 PM Page 82

If the winner-take-all system provides some stateswith strategic advantages in presidential elections, italso has at least three disadvantages. First, it reducesthe incentives of a presidential candidate to competefor votes in “safe” states that are clearly going to betaken by one of the two major party candidates. As aconsequence, candidates tend to compete most heav-ily for votes in “swing” states that could reasonably beexpected to go either way in the election. Second, itseverely reduces the incentives of potential third-partycandidates to undertake the costly task of running forpresident, since they cannot realistically hope to winany votes at all in the electoral college. Finally, forboth reasons it may weaken the incentives of manyvoters in “safe” states to go to the polls: Why bother tovote if you know that a majority of voters in your statewill, in effect, choose the entire slate of presidentialelectors?

Unlike the problems in the electoral college aris-ing from features that are constitutionally prescribed,winner-take-all, as the history of its development clearlydemonstrates, can be altered by state legislatures.

Should We Alter It or Abolish It?

In view of its many deficiencies, what should be donewith the electoral college? Before turning to that ques-tion, let me consider a common objection to changingit. As we saw, the constitutional allocation of electors


04dahl.073_090 11/27/01 4:40 PM Page 83

gives a considerable advantage to a voter in a less pop-ulated state and correspondingly reduces the influ-ence of a voter in a larger state. This marked inequal-ity in representation is often defended, as with themore extreme case of the Senate itself, on the groundthat the small states need and are entitled to protectionfrom the large states.

Entitlement. Why should the interests of votersin the small states be entitled to extra protection? Theobjections I raised in the previous chapter about theSenate are also germane to the electoral college, so I’llrepeat them here:

James Wilson at the Constitutional Convention:“Can we forget for whom we are forming a govern-ment? Is it for men [“persons,” we would want to saytoday] or for the imaginary beings called States?”

James Madison: “Experience suggests [that] . . .the states were divided into different interests not bytheir differences in size, but by other circumstances.”

Beyond the protections provided by the Bill ofRights, the constitutional division of powers in ourfederal system, and an immense body of legislativeand judicial protections for basic rights, do people inthe less populated states possess additional rights orinterests that are entitled to special protection bymeans of unequal representation? If so, specificallywhat are they?

Is there a principle of general applicability thatjustifies an entitlement to extra representation forsome individuals or groups? If so, what is it?


04dahl.073_090 11/27/01 4:40 PM Page 84

If we were to formulate a general principle andapply it fairly, would not those most entitled to protec-tion be the least-privileged minorities—rather thanpeople who happen to live in the smaller states?

I am not aware of any convincing answers to thesequestions and objections.

Need. Behind the belief that small states needspecial protection against the larger states may lurk animage of the oversized bully intimidating his weakerfellows. But once again the concrete issue comes downto one of legitimate rights and interests. If the presi-dent were elected by popular vote, would the legiti-mate rights and interests of citizens in the small statessuffer from neglect or abuse? The belief that theywould suffer rests on the assumption that presidentialcandidates would have little incentive to compete forthe votes of citizens in the small states, and as a resulttheir interests would tend to be neglected in nationalpolicy-making.

But this assumption seems to me mistaken. In asystem of direct election where every citizen’s votesare given equal weight, presidential candidates will beeven more eager than they are now to win votes wher-ever they might be available; and the closer they ex-pect the election to be, the more eagerly they willsearch out those votes. It is true, of course, that inter-est groups with many members are likely, as they arewith the electoral college, to secure greater considera-tion than interest groups with few members. But let usimagine two relatively small, similarly sized groups of


04dahl.073_090 11/27/01 4:40 PM Page 85

potential voters, one concentrated in a small state andthe other in a large state; both groups, let us assume,possess interests and express demands that are equallycompatible with a candidate’s general program. Otherthings being roughly comparable, under a system ofpopular election the incentives of a presidential can-didate to win the votes of the two groups would beequally strong, and the geographical location of eachgroup would be largely irrelevant, conspicuously so inthis age of television.

I see no reasonable grounds on which to concludethat the legitimate rights, interests, and demands ofvoters in small states should be privileged or that theywould be unjustifiably impaired if the president werepopularly elected.

So what should we do? And what can we do?19

What Should We Do About the Electoral College?

From a democratic perspective, the most desirablechange would be a constitutional amendment thatwould replace the electoral college with the direct elec-tion of the president by popular vote; if no candidatewere to receive more than 50 percent of the popularvote, a run-off election between the top two candi-dates would be held shortly thereafter.20

A second possibility is a constitutional amendmentthat would maintain the electoral college but requirethat a state’s electoral votes be allocated to candidates


04dahl.073_090 11/27/01 4:40 PM Page 86

in direct proportion to their share of the popular votesin the state.

Finally, even without a constitutional amendment,in response to a ground swell of popular opinion, statelegislatures might choose the second solution—andthus return to the district system that, as we saw, wasadopted in a number of states in the earliest presiden-tial elections.

What Can We Do?

Evidence from surveys indicates that a substantial ma-jority of Americans would support a reform of theelectoral college.21 In 1989, a proposal to amend theConstitution in order to abolish the electoral collegeand provide instead for the direct popular election ofthe president passed the House of Representativeswith overwhelming support—338–70, or 83 percentof the votes in the House.22

Yet desirable and popular as a reform might be, themost realistic answer to our second question—whatcan we do?—is: Probably not much. By one count, overseven hundred proposals have been introduced in theHouse to modify or abolish the electoral college. Nonehave succeeded. As might be expected, the graveyard ofconstitutional amendments altering the electoral collegeis the Senate—which, as we have seen, is the citadel ofunequal representation. When the proposal for directelection that had passed the House with 83 percent of


04dahl.073_090 11/27/01 4:40 PM Page 87

the votes in 1989 reached the floor of the Senate a yearlater, it ran into a filibuster—unlimited debate by oppo-nents of a measure they do not want to come to a vote.Under Senate rules, to shut off the debate and get onwith the vote would have required the support of 60Senators—60 percent of the total. The motion to enddebate actually did win a majority of votes—54 out of100; but it failed to gain the 60 percent that the rules re-quired.23 And even if the proposed amendment hadcome to a vote, it could not have gained the necessarytwo-thirds—67 Senators—required for an amendment.

Thus the requirement that an amendment mustgain the votes of two-thirds of the members of theSenate gives a veto power to Senators from the smallstates, and these Senators may act in concert withother colleagues who foresee a reduction in the influ-ence of their states on the presidency.24

The relative desirability of three possible solutionsto the problems of the electoral college appears to beinversely related to the likelihood of their enactment.Consequently, however deluded the Framers wereabout to the way their last-minute improvisation wouldwork in practice, it seems unlikely that we shall man-age to erase this undemocratic blemish on the Ameri-can constitutional system.

� � �

THE FRAMERS’ ATTEMPT TO INSULATE THE CHIEF EX-ecutive from popular choice provides the most telling


04dahl.073_090 11/27/01 4:40 PM Page 88

example of their failure to provide a constitutional sys-tem that would be appropriate for a democratic repub-lic. The fate of the electoral college illustrates threeaspects of this failure.

First, almost from the beginning the electoral col-lege wholly failed to operate as the Framers had in-tended. It was, so to speak, swiftly subverted by theemerging democratic forces.

Second, even the more democratized electoral col-lege retained features that could and sometimes didlead to undemocratic outcomes.

Finally, the requirements laid down by the Fram-ers for amending the constitution made it extraordi-narily difficult to bring about changes that might besupported by a majority of American citizens.


04dahl.073_090 11/27/01 4:40 PM Page 89

04dahl.073_090 11/27/01 4:40 PM Page 90

chapter 5

How Well Does the ConstitutionalSystem Perform?

L ET ME REPEAT THE QUESTION I RAISED AT THE

beginning: Why should we uphold the Ameri-can Constitution? One response might be: Be-

cause it performs better than any feasible alternative.1

If the unique properties of our constitutional sys-tem enable it to perform better than the systems ofother democratic countries, then it merits our prideand confidence. If these peculiarities don’t matter, per-haps we should ignore them. But if it performs worse,then shouldn’t we begin to consider possible changes?

Questions about the relative performance of dif-ferent constitutional systems are easy to pose but ex-traordinarily difficult to answer responsibly. True, wecan find today, as only a generation ago or more we

05dahl.091_120 11/27/01 4:40 PM Page 91

could not, many good indicators of how different coun-tries’ systems perform in a variety of important ways:from literacy, education, health, and life expectancy topolitical and civil rights, incomes, income distribution,and others. It is, however, not easy to determine the ex-tent to which a country’s constitutional arrangementsinfluence that country’s performance on such matters.As one scientific commonplace puts it: Correlationsdon’t prove causation. If a greater degree of incomeinequality exists in the United States than in most ofour twenty-two established democracies, is this a con-sequence of our unique constitutional system?

As difficult as questions like this are and althoughconstitution-making is still far from an exact science,we do have more knowledge at our disposal todayabout different constitutional systems than the Fram-ers could have dreamed of—indeed, more than anygeneration in history could possibly have assembled.In the years to come we could acquire even betterknowledge, if we are determined to do so.

With due respect for uncertainty, then, I want toassess how well our constitutional arrangements per-form in comparison with those of the other countrieswhere democracy is well established. I’ll use five crite-ria. To what extent, if at all, do constitutional arrange-ments help to:

1. maintain the democratic system;2. protect fundamental democratic rights;3. ensure democratic fairness among citizens;

92 h ow w e l l d o e s t h e co n st i t u t i o na l sy st e m p e r f o r m ?

05dahl.091_120 11/27/01 4:40 PM Page 92

4. encourage the formation of a democratic consensus;and

5. provide a democratic government that is effective insolving problems?

Maintaining Democratic Stability

Do different constitutional arrangements significantlyaffect the chances that a country will preserve its basicdemocratic institutions—that it will, in short, remain ademocracy? This question opens up a vast subject thathas been extensively explored in recent years.

Unfortunately for my purposes here, but fortu-nately for democracy, the experience of our twenty-two democratic countries cannot provide the evidencewe need to answer our question about stability. I se-lected these countries as suitable for comparison withthe United States precisely because they are the onlycountries in the world today that have fully maintainedtheir basic democratic institutions for a half-century ormore. Since democratic institutions have never col-lapsed in any of these countries during that period (orlonger), we lack any basis for comparing their perform-ance in maintaining basic democratic stability. For example, if twenty-two persons of greatly varying butmoderate diets all remain about equally healthy, wecouldn’t draw any conclusions about the effects of theirdiets on their health. So, too, with our twenty-two coun-

h ow w e l l d o e s t h e co n st i t u t i o na l sy st e m p e r f o r m ? 93

05dahl.091_120 11/27/01 4:40 PM Page 93

tries: they have all performed equally well in maintain-ing the existence of their democratic systems.

From our hypothetical study of diets, however, wecould derive one conclusion that is far from trivial:good health is possible within a considerable range ofmoderate diets. Similarly, from the experience of ourtwenty-two countries we can draw at least one impor-tant conclusion: the differences in their constitutionalarrangements evidently have not affected the survivalof their basic democratic institutions. Because all ofthese countries have remained steadily democratic, itfollows that within the rather wide range of constitu-tional variations they exhibit, their differences simplyhaven’t mattered for democratic survival.2

How can we explain this surprising conclusion?Let me offer three general propositions.

First, if the conditions in a country are highly fa-vorable for democracy, constitutional differences likethose that exist among our twenty-two countries willnot affect the stability of the basic democratic institu-tions. To return to the analogy with diet and health:among a group of people who otherwise live healthylives, the variations in their moderate diets won’tmatter much. It would take us too far afield to de-scribe the conditions that favor democratic stability,but we can say that they appear to include such thingsas the effective control by elected leaders over the mil-itary and police, a political culture supportive of dem-ocratic beliefs, and a relatively well-functioning eco-nomic order, among others.


05dahl.091_120 11/27/01 4:40 PM Page 94

Although a country with favorable conditions likethese is likely to maintain its democratic institutionsunder a variety of constitutional possibilities, no con-stitutional system can preserve democracy in a countrywhere these and other favorable conditions are absent.As I said, none of our twenty-two democratic coun-tries have suffered a breakdown of democracy duringthe past half-century. But if we move back to the pre-vious century, we do find that in one of our countrieswhere all the basic democratic institutions (except forinclusive citizenship) had been in place for more thanhalf a century, the constitutional arrangements couldnot prevent a breakdown into a civil war—a conflict,moreover, that resulted in vastly greater casualties thanthe American or the French Revolution. That countrywas the United States. The conditions necessary forpreserving national unity had become so unfavorablethat probably no constitutional arrangements couldhave prevented both secession and civil war. Given theextreme polarization in interests, values, and ways oflife between the citizens of the slave states and thoseof the free states, I cannot imagine any democraticconstitution under which the two sections could havecontinued to coexist peacefully in one country.

But suppose that, unlike our twenty-two countries,we have a country in which some conditions are fa-vorable for democracy while others are unfavorable.Might the particular features of a constitution matterin a country where the underlying conditions makedemocracy rather chancy? It seems possible that in


05dahl.091_120 11/27/01 4:40 PM Page 95

situations of uncertainty, constitutional arrangementsmight just tip the balance one way or the other, towarddemocratic stability or democratic breakdown. Whatdoes the evidence indicate?

A much debated issue is the effect on democraticstability of presidential versus parliamentary systems.Perhaps because they were impressed by Americanstability and power, developing countries have oftenadopted some version of a presidential system. As twoscholars have noted, “A remarkable fact . . . is the ex-tent to which presidentialism is a Third World phe-nomenon.”3 So is the likelihood of breakdown. Are thetwo connected—presidentialism and breakdown? Theanswer is a subject of dispute. Some scholars have con-cluded that in countries where the conditions for a sta-ble democracy are mixed—some favorable, some un-favorable—a presidential system is more likely than aparliamentary system to put a greater strain on demo-cratic survival.4 Others argue, however, that “parlia-mentarism has not fared any better in the third worldthan has presidentialism; arguably, it has fared worse.”5

To explore this controversial question would takeus far outside the purposes of this book, thus I leave it unsettled here but accompanied by four brief ad-monitions: One, the intricate American constitutional system is probably not suitable for export to othercountries. Two, insofar as we Americans can directlyinfluence decisions in newly democratizing countries,we should avoid trying to impose it on them. Three,there is probably no single best constitutional system.


05dahl.091_120 11/27/01 4:40 PM Page 96

And finally, democratic constitutions need to be tai-lored to fit the culture, traditions, needs, and possibili-ties of a particular country.

Protecting Fundamental Rights

If the evidence about democratic stability is inconclu-sive, what does the evidence show about democraticrights? How well do the constitutional systems of dem-ocratic countries protect the rights, opportunities, andduties of both majorities and minorities?

Here we again run into a methodological problem.As I shall explain in more detail in the next chapter,democracy and its fundamental institutions presup-pose the existence of certain fundamental rights, suchas freedom of speech and freedom of the press. Wecan reasonably classify all twenty-two of our countriesas democratic only because, among other things, theyall have maintained a high level of protection for basicdemocratic rights and liberties. As with stability andbreakdown, in selecting our twenty-two countries wehave necessarily excluded any in which massive andpersistent violations of fundamental political rightshave occurred.

Nonetheless, even if we assume that all of themhave maintained political rights at or above the basicthreshold for democracy, we do find some smallervariations. The important point, however, is that thereis no discernible relationship between constitutional


05dahl.091_120 11/27/01 4:40 PM Page 97

systems, broadly defined, and these variations in rightsand freedoms. Freedom House, an independent non-profit organization that since 1973 has provided an-nual evaluations of the conditions of liberty among thecountries of the world, assigns the identical score onpolitical rights to all twenty-two democratic countries.On civil liberties, seven countries — Belgium, CostaRica, France, Germany, Israel, Italy, and the UnitedKingdom— fall just below the top score.6 Yet if we goback to our examination of the variations in broad con-stitutional features that might explain why these sevenfall below the rest, none can account for the differ-ence. Federalism, strong bicameralism, unequal rep-resentation in the upper house, strong judicial review,the electoral and party systems, and parliamentarismor presidentialism: none provide an explanation. Orconsider freedom in print and broadcasting media. At the top, with a near-perfect score in the FreedomHouse evaluations, is Norway: a nonfederal countrywith a parliamentary system, a unicameral parliament,proportional representation, multiple parties, coalitiongovernments, and no judicial review of parliamentaryenactments. Halfway down, just below the UnitedStates, is the Netherlands, another nonfederal countrywith a parliamentary system, proportional representa-tion, multiple parties, coalition governments, and nojudicial review. Why the difference? Or compare fourof the federal countries—Switzerland, Australia, theUnited States, and Germany. Federalism can hardlyaccount for the variations in their scores.7


05dahl.091_120 11/27/01 4:40 PM Page 98

The most relevant conclusion we can draw is thatamong mature democratic countries, where the con-ditions for democracy are generally favorable, differ-ences in rights and liberties cannot be attributed toconstitutional systems. But if not constitutional sys-tems, then what?

The answer will be found, I believe, in differencesin national histories, political cultures, and perceptionsof internal and strategic threats to survival. If this is thecase, then in the end a democratic country cannot de-pend on its constitutional systems for the preservationof its liberties. It can depend only on the beliefs andcultures shared by its political, legal, and cultural elitesand by the citizens to whom these elites are responsive.

Democratic Fairness

How does the American constitutional system com-pare with those of other mature democracies in thefairness with which it treats different citizens? As weall know, the question of fairness or justice has been asource of endless debate among the best minds sinceancient times. Indeed, differences in views about jus-tice seem to be built into the human condition. I, how-ever, want to bypass these perennial controversies andfocus instead on one aspect of fairness that bears di-rectly on the question at hand.

If I may now use the term constitutional system inits broader sense to include electoral arrangements,


05dahl.091_120 11/27/01 4:40 PM Page 99

we can arbitrarily reduce the alternatives to two. Inone, which I’ll call a proportional system, as a result ofproportional representation the percentage of seatswon by a party in the legislature will roughly mirror thepercentage of votes cast for candidates of that party. Inthe other, which I’ll call a majoritarian system,8 candi-dates receiving the most votes in a particular districtwin that district’s single seat, and the other candidates,therefore, win no seats at all. In a proportional system,all minority parties that gain votes above some thresh-old, such as 5 percent, will be represented in the legis-lature. In a majoritarian system, if the candidate ofone party were to win a plurality (relative majority) ofvotes in every district, then that party would win allthe seats. Although such an extreme outcome is only a theoretical possibility, in majoritarian systems theparty with a majority of votes does ordinarily win a disproportionately large number of seats; the secondlargest party gains a disproportionately small numberof seats; and all third parties gain few if any seats.

In an earlier chapter I pointed out that propor-tional representation is likely to produce a multipartysystem and coalition governments; first-past-the-postis likely to produce two dominant parties; and in a par-liamentary system with two dominant parties, theprime minister and cabinet are likely to be drawn froma single party with a majority of seats, as is typically thecase in Britain.

In the debate over the relative desirability of pro-portionality versus majoritarianism,9 virtually no one


05dahl.091_120 11/27/01 4:40 PM Page 100

questions that proportionality is fairer to citizens thanmajoritarianism. Proportionality doesn’t necessarilymean, however, that the principle of majority rule en-tirely stops operating. In the legislature, for example,elected representatives will ordinarily make their deci-sions by majority rule. But because the governing coali-tion will generally include representatives from minor-ity parties, governing majorities are likely to be moreinclusive than in a majoritarian system. Thus, a propor-tional system comes closer than a majoritarian system toproviding equal representation—an equal say—for all.

Advocates of majoritarianism may concede thatproportionality is fairer; but they might argue that amajoritarian system offers two advantages that consid-erably outweigh its unfairness. For one thing, its de-fenders often say, proportionality tends to produce gov-erning coalitions that are more unstable and, therefore,more ineffective than governments in majoritarian sys-tems. Does the experience of the large number of ma-ture democratic countries with proportional systemconfirm that their governments are less effective? In amoment I’ll turn to evidence bearing on this question.But just suppose we were to find that proportional systems are, in general, no less effective than govern-ments in countries with majoritarian systems. On whatgrounds could we then reject proportionality?

We might still reject proportionality if we concludethat having two dominant parties rather than the mul-tiplicity of parties typical of proportional systems helpsto make governments more accountable to voters. Our


05dahl.091_120 11/27/01 4:40 PM Page 101

supposition might run something like this: Two-partymajoritarian systems help voters hold governmentsmore accountable because they are better at simplify-ing and clarifying the alternatives open to voters. Con-sequently, during campaigns and elections, voters canfix responsibility for the decisions and policies that thegovernment has recently adopted. What is more, be-cause voters in proportional systems face a multiplicityof parties and possible governing coalitions, they mayfind it difficult to guess what their vote will actuallymean. Forming a majority coalition in a multipartyparliament can be a tricky business. If a voter’s partywants to be included in the governing coalition, whatcompromises will it be compelled to make in order tofind a place? And what policies will the coalition fi-nally manage to agree on and carry through? In con-trast, because voters in majoritarian systems ordinarilyhave only two realistic choices, they can make moreinformed guesses about the direction the governmentis likely to take under one major party or the other.10

A justification along these lines would providestrong support for majoritarian systems. But as appeal-ing as it is, the majoritarian vision isn’t easily trans-formed into reality. For one thing, in the small num-ber of countries with nominally majoritarian systems,as Powell points out, we find a “persistent refusal ofvoters to deliver majority support for a single party oreven a preelection coalition.” In forty-five electionsfrom 1969 through 1994 in six “predominantly majori-tarian” countries, “only in Australia in 1975 and in


05dahl.091_120 11/27/01 4:40 PM Page 102

France in 1981 did a party or preelection coalition wina clear voter majority.” In short, just as happens inAmerican presidential elections, majoritarianism oftenfails to produce a government that reflects the choicesof a majority of voters. Second, the distortion betweenseats and votes in majoritarian systems sometimes cre-ates a majority of seats for a party that has failed to win even a plurality of votes and thus has actuallycome in second. In these cases, the minority partyamong voters becomes the majority party in the legis-lature. Third, even in majoritarian systems, “in prac-tice, purely two-party politics is a rare phenomenonand often not robust when it appears.” That is, a thirdparty—like the Liberal Democrats in Britain—mayprevent either of the two major parties from gaining amajority of votes, even though one of them may gain amajority of seats.11

Encouraging Consensus

Even if proportionality is fairer than majoritarianism,many Americans will say that the price of fairness istoo high. Any country in which multiple parties com-pete for office, they assume, will surely be divided andcontentious and suffer from government by unstableand ineffective coalitions. How valid is this commonAmerican view?

In direct contradiction to that view, Arend Lijp-hart, the scholar who pioneered the comparative analy-


05dahl.091_120 11/27/01 4:40 PM Page 103

sis of proportionality and majoritarianism in democraticcountries, refers to proportional systems as “consensusgovernments.”12 And rightly so, for experience showsthat even if proportionality cannot always overcomedeep political, social, cultural, or economic cleavages(as in Israel, for example), a proportional system cansometimes help to maintain internal peace, provideopportunities for compromise among opponents, andproduce a broad consensus in favor of not only gov-ernment policies but the country’s political arrange-ments as well.

Let me offer three examples. In the Netherlands,13

religious and ideological differences led to a profounddivision of the country into four basic groups: Prot-estants, Catholics, Liberals, and Socialists. The fourgroups became fairly distinctive subcultures prettymuch walled off from one another by their own insti-tutions, from newspapers and radio to schools, tradeunions, hospitals, marriages, residences, and more.After proportional representation was introducedearly in the twentieth century, each of the groups alsosupported its own separate political party. Not sur-prisingly, in the late nineteenth and early twentiethcenturies these basic cleavages among the four subcul-tures led to severe conflicts over education, the fran-chise, and the rights of labor. By 1910 the disputes hadbecame so intense that leaders of the four groups be-came alarmed for the future of the country. Spurredby their concern, from 1913 to 1917 they not onlymanaged to negotiate acceptable compromises, but


05dahl.091_120 11/27/01 4:40 PM Page 104

they also agreed that the political parties representingthe four groups would all be represented in the cabi-net. In short, they created a consensus government.Despite the persisting cleavages among the four sub-cultures, a highly institutionalized system of full inclu-sion endured for half a century, when demographicchanges and a decline in the intensity of the differ-ences reduced the need for complete inclusion of allfour parties in every cabinet. Even so, and down to thepresent day, Dutch governments have continued toemphasize inclusion and consensus rather than ma-joritarian control of the government.

Or consider Switzerland with its four national languages — German, French, Italian, and the tinyRomansh-speaking population; its two major religions,Protestantism and Catholicism, which were the sourceof sanguinary conflict until the middle of the nine-teenth century; and its two dozen or so cantons, manyof which are internally rather homogeneous in lan-guage and religion. If you were to reflect on the possi-bilities of conflict among these subcultures you mightconclude that, like the Balkans, Switzerland must beforever bubbling over with intense disputes and mighteven be on the verge of national disintegration. But thepragmatism, common sense, and national attachmentsof the Swiss enabled them in 1959 to create a propor-tional system in which representatives of the fourmajor parties representing the different subculturesare usually all included in the executive branch—thefederal council or Bundesrat.


05dahl.091_120 11/27/01 4:40 PM Page 105

It is a reasonable conclusion that majoritarian sys-tems in Switzerland and in the Netherlands would havemade the search for governments based on a broad con-sensus among different subcultures not only extraordi-narily difficult but probably downright impossible.

A very different situation exists in Sweden. A highlyhomogeneous people (until the recent influx of immi-grants, at any rate), Swedes have a long and settledtradition of consensus politics. Although the origins ofthe Swedish parliament go back some centuries, de-mocratization arrived comparatively late. Not until1917 was the power to choose the prime ministershifted from the king to the parliament. In that sense,Swedish democracy dates back to 1917. Proportionalrepresentation had already been introduced in parlia-mentary elections, however, at the beginning of thetwentieth century. Yet neither proportional represen-tation nor democratization diminished the long-stand-ing Swedish tradition of consensus. As a Swedish polit-ical scientist has written:

In the Swedish political tradition . . . “accountability” isseldom mentioned as a value. Instead, legitimacy is pro-moted by another strategy. By sharing power with theparties in opposition and including them in the rule ofthe country, the government is supposed to be regardedas representative for the people as a whole and conse-quently one that all can feel loyal to. To “reach consen-sus,” to “find a common policy,” to “capture the will ofthe people” have been the declared motives of Swedishpoliticians. Representativeness is the central norm ofpolitical culture.14


05dahl.091_120 11/27/01 4:40 PM Page 106

From an American perspective, the result can beunbelievable. In stark contrast to Holland and Switzer-land, Swedish cabinets have frequently been drawnfrom a party or coalition that has actually lacked a ma-jority of seats in parliament. In the past century, “mi-nority governments have been by far the most com-mon. The average parliamentary support enjoyed bygovernments between 1920 and 1994 has been 41.5percent.” You might well wonder how minority gov-ernments could ever get anything accomplished—or,for that matter, remain very long in office. The answerseems to be that in order to gain a broad consensusboth in the parliament and in the country at large,even minority governments negotiate with representa-tives of parties outside the government. In short, evenminority governments govern by consensus.

If you wonder why the Dutch, the Swiss, and theSwedes prefer proportionality to majoritarianism, theanswer is fairly clear: not only does it seem to themmuch fairer but it also helps to achieve and maintain abroad consensus for government policies.

What is more, proportionality can strengthen con-sensus not just for policies but for democracy as well.The reason appears to be that proportionality resultsin fewer losers. To clarify this point let me overstate it:In a majoritarian system the only winners in electionsare the citizens who happen to be in the majority; allthe other citizens, being in the defeated minority, arelosers. By contrast, in proportional systems with con-sensus governments, everyone—well, almost every-


05dahl.091_120 11/27/01 4:40 PM Page 107

one—can win, not everything they hoped for perhapsbut enough to leave them basically satisfied with theirgovernment.

Lest you suppose that these judgments are nothingmore than interesting speculations, let me cite somepersuasive supporting evidence.15

In a 1990 survey of citizens’ views in eleven Euro-pean democracies, respondents were asked how satis-fied they were with “the way democracy works” in theircountry. They also reported how they voted in theircountry’s last national election. Knowing the outcomeof that election in each country, the authors of the studyclassified the respondents as winners or losers. Theeleven countries were then arranged from the most ma-joritarian, Britain, to the most consensual, the Nether-lands. The results, which the authors of the study describe as “robust,” were pretty clear: In the moreconsensual countries, losers were almost as satisfied asthe winners with the way democracy worked in theircountry. By contrast, in the more majoritarian coun-tries, losers were much more likely to be dissatisfied.

To describe the results another way, suppose that inone country 70 percent of the winners but only 40 per-cent of the losers are satisfied with the working ofdemocracy, a difference of 30 percent. In another coun-try, say, 70 percent of the winners and 65 percent of thelosers are satisfied, a difference of only 5 percent. Inthe study of eleven European democracies that I justmentioned, this difference in satisfaction with the way


05dahl.091_120 11/27/01 4:40 PM Page 108

democracy worked decreased steadily from large differ-ences in the most majoritarian countries like Britain—around 25 percent—to almost negligible differencesin the most consensual countries like the Netherlands,where it was less than 5 percent.16 Moreover, these re-sults held up even when variations in such influencesas economic performance, socioeconomic status, andpolitical interest were taken into account.17

In short, if you live in a majoritarian country andyour party comes in second or worse, you’re likely tobe dissatisfied with the way democracy works in yourcountry. But if you live in a democratic country with amore consensual system and your party comes in sec-ond, or third, or maybe even fourth, you’re likely tofeel satisfied with the way democracy works becauseyou know your views will still be represented in thegovernment.

This is all fine, you might say, but can a consensussystem produce effective governments—governmentsthat can solve the problems that concern citizens?Might not majoritarian governments perform more ef-fectively? In particular, hasn’t our American constitu-tional system been just as effective as many consensusgovernments and perhaps even more effective thanmost in getting things done that citizens want? I’ll turnto this question in a moment.

But before I do so, I want to call your attention toone salient aspect of our constitutional system: TheAmerican constitutional system is not majoritarian.


05dahl.091_120 11/27/01 4:40 PM Page 109

The American Hybrid

Whatever its hypothetical advantages might be, themajoritarian vision is not applicable to the Americansystem of government. Although our system is not pro-portional, neither is it majoritarian. Whether by rationalintention or bv an understandable inability to foreseethe consequences, or both, James Madison and his fel-low delegates created a constitutional system that is ahybrid of proportionality and majoritarianism.

Three majorities. It is true that two parties domi-nate our political landscape more completely than inany other established democracy. Yet even when oneparty wins not only the presidency but also majoritiesin both the Senate and House, three different popularmajorities are at work; the composition of each of thethree majorities does not match the others; and theirrepresentatives do not necessarily agree. I would notsay that this feature is necessarily undesirable, but un-desirable consequences do show up when it is com-bined with other aspects of our system.

Divided government. To begin with, one partymay fail to win control of all three branches. Indeed,during the past half-century, control of the presidencyand both houses of Congress by a single party has be-come a rarity. As David Mayhew comments in DividedWe Govern, “Since World War II, divided party con-trol of the American national government has come to seem normal.”18 From 1946 to 2000, the three

110

05dahl.091_120 11/27/01 4:40 PM Page 110

branches have been divided between the two partiesmore than six years out of every ten. Our constitutionnot only permits divided government, it cannot pre-vent divided government. And it provides no way outexcept by elections at fixed intervals—elections thatmay only reproduce the existing divisions or inaugu-rate new ones.

Does divided government matter? In particular, doperiods of divided government make it more difficultfor the three branches to agree on national policies re-quiring legislation—in short, do these periods make astalemate easier? The evidence is mixed. In a land-mark study of the period between the elections of1946 and 1990, David Mayhew found no “relationworth crediting between the incidence of importantlaws and whether party control was unified or di-vided.”19 A subsequent analysis of the period from1947 to 1994, however, concluded that important leg-islation is more likely to be passed in periods of uni-fied government. The difference between divided andunified government is particularly marked when allthree branches are under the unified control of themore activist of the two parties, the Democrats.20

The president: consensual, majoritarian, neither,both? At the apex of this complex structure of politi-cal institutions sits—or stands, as the case may be—the American presidency, an office with no equivalentin any of the other established democracies or, so faras I am aware, in any other democratic country.


05dahl.091_120 11/27/01 4:40 PM Page 111

It is difficult, indeed impossible, to fit the pres-idency into the simple categories of consensual or majoritarian. One obstacle to straightforward classifi-cation is the president’s combination of roles. Most no-tably, whereas in the other older democracies the rolesof prime minister and ceremonial head of state areseparated, in our system they are blended, not onlyconstitutionally but also in popular expectations. Weexpect our president to serve both as chief executiveand as a sort of ceremonial, dignified, American-styleelected monarch and moral exemplar.

The mix of roles was present from the beginning.Although during the early years the vituperation ofpresidents in the press often far exceeded boundariesnow regarded as acceptable, in order to maintain thedignity of the office presidents rarely addressed thegeneral public, except possibly on official occasions;and when they did they rarely employed popular rhet-oric or discussed their own policies. In these respects,if no others, they acted less like politicians than mon-archs or ceremonial heads of state. Indeed, until the1830s, presidential candidates did not make campaignspeeches at all; and until Woodrow Wilson broke acentury-old taboo in 1912, no president had ever“stumped on his own behalf.”21

Beginning with Andrew Jackson, however, presi-dents had already begun to make the audacious claimthat by virtue of their election, they alone representedthe entire people, or at least a majority. Some would


05dahl.091_120 11/27/01 4:40 PM Page 112

even assert that their election endowed them with a“mandate” for their policies. To the extent that theclaim of a mandate was accepted, it would increase theacceptability of a president’s policies by clothing themwith the legitimacy of popular sovereignty.

Despite the frequency with which newly electedpresidents stake a claim to a mandate, the closer youinspect the chain of assumptions that are supposed tosupport the claim, the more fragile the links appear.22

It requires an extraordinary leap of faith to infer theviews of voters from nothing more than the way theycast their votes for president. Although systematicopinion surveys provide a much firmer basis for un-derstanding public attitudes and expectations, theclaim to a presidential mandate based on nothingmore than the outcome of an election antedated sys-tematic surveys by more than a century. And evensince the beginning of systematic opinion surveys inthe 1940s, presidents and their followers (and pundits)have typically based their shaky claim to a mandate onnothing more than the election returns, which couldnot possibly reveal whether presidential policies ac-corded with voters’ preferences. Since the 1940s thatconcordance is more likely to come from scrupulousattention to public opinion surveys than from readingthe tea leaves of the election results.

Presidential claims to represent “the Americanpeople” together with efforts to promote particular na-tional policies are elements in the general mixture of


05dahl.091_120 11/27/01 4:40 PM Page 113

roles that distinguishes the American presidency andmakes it neither simply majoritarian nor simply con-sensual.

This mix of roles seems to be generally acceptedby Americans. We want our presidents, it seems, to besimultaneously shrewd politicians and gifted states-men. We expect them to live in both the real world ofdaily politics and an imaginary world above politics.Most of us understand that to succeed in office a pres-ident must be an active and forceful partisan, a partyleader, and the negotiator and deal-maker who mas-sages, cajoles, bribes, threatens, and coerces Congressto secure both votes and support if promises and poli-cies are to be realized.

But we also expect our president to serve as amoral example for us all, to stand as an icon on whichwe can devoutly project qualities of intelligence, knowl-edge, understanding, compassion, and character farabove those we expect in ordinary beings like our-selves. Because no mortal can meet these exalted stan-dards, we have often savaged a president while he is inoffice and then exalted him in memory. In office, wemay portray a president as a cartoonish bungler. Butafter departing the White House, or this world, we ig-nore the warts and scars and paint an idealized por-trait of a noble and exemplary figure.

Ambivalence toward the presidency is deeply in-grained in our American culture. As children we learnto worship our presidents for their greatness;23 asadults we deride them for failing to achieve the great-


05dahl.091_120 11/27/01 4:40 PM Page 114

ness of their mythic predecessors. In choosing amongcandidates for the presidency we yearn for perfection;yet our only realistic choices are among flawed humanbeings who live with all the moral ambiguities re-quired by the life of politics. In short, the impossiblemix of roles an American president is expected to playplaces a heavy burden not only on the incumbent, butmore important, on American voters.

Accountability. Holding the government account-able for its actions may be an even greater burden forvoters. Where are we to place responsibility for theconduct of our government? When we go to the polls,whom can we hold accountable for the successes and failures of national policies? The president? TheHouse? The Senate? The unelected Supreme Court?Or, given our federal system, the states, where govern-ments are, in their complexity, a microcosm of the na-tional government?

Even for those who spend their lives studying poli-tics, these can be extremely difficult questions to an-swer. I, for one, am inclined to think that comparedwith the political systems of the other advanced demo-cratic countries, ours is among the most opaque, com-plex, confusing, and difficult to understand.

We see, then, that our hybrid system, which is nei-ther majoritarian nor proportional, may possess theadvantages of neither and the defects of both. If it failsto ensure the fairness promised by the proportional vi-sion, it also fails to provide the clear accountabilitypromised by the majoritarian vision.


05dahl.091_120 11/27/01 4:40 PM Page 115

Democratic Effectiveness

To all this you might say: Even if the American hybridmay have some shortcomings when it is viewed in acomparative perspective, isn’t it just as effective asother governments in dealing with the issues that con-cern American citizens?

Once again, we cannot answer this question re-sponsibly without facing up to some severe method-ological problems. Our twenty-two democratic coun-tries differ in so many ways that teasing out the effectsthat we can reasonably attribute to constitutional sys-tems is a pretty formidable task. Take size. The popu-lation of the United States is sixty times larger thanNorway’s, fifty times Denmark’s, thirty-seven timesSwitzerland’s, thirty times Sweden’s, and nearly a thou-sand times larger than Iceland’s, which is somewhatsmaller in population than Tampa, Florida.

Although the effects of population size on demo-cratic political life are extremely difficult to measure,they can hardly be ignored.24

Or consider diversity: in a very general way, diver-sity tends to increase with size.25 Yet can we truly saythat the United States is more diverse than Switzer-land, or than our neighbor Canada?

Add in one more variable: relative affluence. Al-though Norway and Costa Rica are both relatively smallin population—there are around 4.5 million Norwe-gians and around 3.7 million Costa Ricans—Norway’sGNP per capita is fourteen times that of Costa Rica.26

116

05dahl.091_120 11/27/01 4:40 PM Page 116

How much do these differences in size, diversity,and relative affluence affect political life and publicpolicy? Despite the difficulties presented by nationalvariations like these, comparative data can help us gainsome appreciation of the way in which American per-formance compares with that of other advanced dem-ocratic countries.27 When the United States is rankedwith other established democracies on such matters asthe rate of incarceration, the ratio of poor to rich, eco-nomic growth, social expenditures, energy efficiency,foreign aid and the like, its performance is somethingless than impressive. (See Appendix B, Table 5.) Twoareas in which our country ranks highest are hardlyachievements of which we can be proud. On the per-centage of the population we incarcerate, we come outa clear winner, while our ratio of rich to poor is higherthan that of most other countries. We rank in the bot-tom third—and on some measures close to the bottomof the bottom third—on voter turnout, state welfaremeasures, energy efficiency, and the representation ofwomen in the national legislature. What is more, inspite of our good showing on economic growth, we arealmost dead last in our social expenditures. Finally,even though many Americans believe that we are toogenerous in our economic aid to other countries,among nineteen democratic countries we are at thevery bottom.

In his comparison of consensual and majoritariansystems in thirty-six countries, Arend Lijphart con-cluded that “majoritarian democracies do not out-


05dahl.091_120 11/27/01 4:40 PM Page 117

perform the consensus democracies on macroeconomicmanagement and the control of violence—in fact, theconsensus democracies have the slightly better rec-ord—but the consensus democracies do clearly out-perform the majoritarian democracies with regard tothe quality of democracy and democratic representa-tion as well as with regard to what I have called thekindness and gentleness of their public policy orienta-tions.”28

Paraphrasing Lijphart’s conclusion, I find no con-vincing evidence that our hybrid outperforms systemsthat are either more fully consensual systems or morefully majoritarian. On the contrary, compared withother democratic countries our performance appears,on balance, to be mediocre at best.

How much does our performance have to do withour constitutional system? To tease out the extent ofthat connection would be extraordinarily difficult, per-haps impossible, and I am going to leave that task toothers.29

It seems reasonably clear, however, that a constitu-tional system better designed to achieve such demo-cratic goals as the protection of fundamental rights,fair representation, and greater consensus does notnecessarily come at the price of governmental effec-tiveness, much less the stability of the democratic sys-tem itself.

If this is so, then do we not have every reason toundertake a serious and responsible examination of pos-


05dahl.091_120 11/27/01 4:40 PM Page 118

sible alternatives to our present American Constitution?Or, at the very least, isn’t it time—well-past time—thatwe stop thinking of our Constitution as a sacred textand begin to think of it as nothing more, or less, than ameans for achieving democratic goals?


05dahl.091_120 11/27/01 4:40 PM Page 119

05dahl.091_120 11/27/01 4:40 PM Page 120

chapter 6

Why Not a More DemocraticConstitution?

I BEGAN BY POSING THIS QUESTION: WHY SHOULD WE

Americans uphold our Constitution? Let me nowchange the question slightly: What kind of consti-

tution should we feel obliged to uphold?I mean, of course, an American constitution—not

necessarily our present Constitution, but a constitutionthat, after careful and prolonged deliberation, we andour fellow citizens conclude is best designed to serveour fundamental political ends, goals, and values.

The Constitution as National Icon

I am well aware that in expressing reservations aboutthe Constitution, as I have in these essays, I may be

06dahl.121_140 11/27/01 4:40 PM Page 121

judged guilty of casting stones at a national icon.“From the time of the founding fathers,” a historianhas recently remarked, “there [has] been a sacred auraabout the Constitution, manifest in holiday politicalrhetoric.” During the years between the two WorldWars, worship of the Constitution “acquired the trap-pings of a religious cult.”1 This reverential attitude con-tinues. In a telephone survey of one thousand adultU.S. citizens in 1997, 71 percent said they stronglyagreed with the statement that they were proud of theConstitution; another 20 percent somewhat agreed.2

In a 1999 survey, 85 percent said they thought theConstitution is a major reason why “America has beensuccessful during this past century.”3

I don’t dismiss the importance of icons for strength-ening beliefs, religious or political, nor do I dismissthe utility of myth and ritual in helping to foster na-tional cohesion. But a faith that rests on little morethan a general conformity to conventional beliefs is afragile foundation for nationhood—not to say for de-mocracy. I want therefore to suggest an alternative.

The only legitimate constitution for a democraticpeople, it seems to me, is one crafted to serve demo-cratic ends. Viewed from this perspective an Americanconstitution ought to be the best that we can design forenabling politically equal citizens to govern themselvesunder laws and government policies that have beenadopted and are maintained with their rational consent.

This is hardly a novel view. What I am suggesting isthat a constitution derives its legitimacy from a moral

122 w h y n ot a m o r e d e m o c r at i c co n st i t u t i o n ?

06dahl.121_140 11/27/01 4:40 PM Page 122

and political judgment announced to the world morethan two centuries ago. This judgment (slightly modi-fied from the original) asserts:

That all human beings are created equal, that they areendowed by their Creator with certain inalienableRights, that among these are Life, Liberty, and the Pur-suit of Happiness. That to secure these rights, Govern-ments are instituted among a people, deriving their justpowers from the consent of the governed. That when-ever a Form of government becomes destructive of theseends, it is the Right of the people to alter or abolish it,and to institute new government, laying its foundationon such principles and organizing its powers in such aform, as to them shall seem most likely to effect theirSafety and Happiness.

But two questions immediately arise. First, is polit-ical equality a realistic goal? Second, is it really a desir-able goal?4

Is Political Equality a Realistic Goal?

Some of you may dismiss the noble words I have justquoted as obviously false. If anything about equality isself-evident, you might object, it is that human beingsaren’t equal. Whether by genes, birth, luck, achieve-ments, or whatever, we aren’t equal in education, cul-tural endowments, social and communication skills, in-telligence, motor skills, incomes, wealth, the countryin which we live, and so on. Though this objection is a

w h y n ot a m o r e d e m o c r at i c co n st i t u t i o n ? 123

06dahl.121_140 11/27/01 4:40 PM Page 123

commonplace, it wholly misses the point. The men whowrote and adopted the American Declaration of Inde-pendence hardly needed to be reminded of such ele-mentary matters. They knew too much about the waysof the world to make assertions that were obviously con-tradicted by everyday human experience. But, of course,they didn’t mean the Declaration to be understood asa statement of fact. They meant it to be understood asa moral statement. Human equality, they were insist-ing, is a moral and even a religious standard againstwhich it is right and proper to judge a political system.

Yet ideal standards might rise so far beyond humanreach as to be irrelevant. Is political equality so remotefrom human possibilities that we might just as wellforget it?

I need hardly remind you of the enormous andpersistent barriers to political equality and, indeed, tohuman equality in general.5 Consider that elementaland age-old barrier arising from differences in thetreatment of men and women. The authors of the fa-miliar words about equality that I just quoted and thefifty-five delegates to the Second Continental Con-gress who voted to adopt the Declaration in July 1776,were, of course, all men, none of whom had the slight-est intention of extending the suffrage or many otherbasic political and civil rights to women—who by thelaws of that time and for a full century after were thelegal property of their fathers or husbands.

Nor did the worthy supporters of the Declarationmean to include slaves or, for that matter, free persons


06dahl.121_140 11/27/01 4:40 PM Page 124

of African origin, who were a substantial fraction of thepopulation in almost all the colonies that claimed theright to become independent self-governing republics.The principal author of the Declaration, Thomas Jef-ferson, owned several hundred slaves, none of whomhe freed during his life.6 It was not until more thanfour score and seven years later (to borrow a poeticphrase from Lincoln’s Gettysburg Address) that slav-ery was legally abolished in the United States by forceof arms and constitutional enactment. And it took yetanother century before the rights of African Ameri-cans to participate in political life began to be effec-tively enforced in the American South. Now, two gen-erations later, Americans white and black still bear thedeep wounds that slavery and its aftermath inflictedon human equality, freedom, dignity, and respect.

Nor did our noble Declaration mean to include thepeople who for thousands of years had inhabited thelands that Europeans colonized and came to occupy.We are all familiar with the story of how the settlers de-nied homes, land, place, freedom, dignity, and human-ity to these earlier peoples of America, whose descen-dants even today continue to suffer from the effects oftheir treatment throughout several centuries, whentheir most elementary claims to legal, economic, andpolitical—not to say social—standing as equal humanbeings were rejected, often by violence; more recently,this lengthy period has been followed by neglect andindifference.

All this in a country that visitors from Europe, such


06dahl.121_140 11/27/01 4:40 PM Page 125

as Alexis de Tocqueville, portrayed (correctly, I think)as displaying a passion for equality stronger than theyhad ever observed elsewhere.

Yet, despite the fact that throughout human historyequality has often been denied in practice, throughoutthe past several centuries many claims to equality, in-cluding political equality, have come to be much morestrongly reinforced by institutions, practices, and be-havior. Although this monumental historical move-ment toward equality is in some respects worldwide, ithas been most conspicuous in such democratic coun-tries as Britain, France, the United States, the Scandi-navian countries, the Netherlands, and others.

In the opening pages of the first volume of Democ-racy in America, Tocqueville pointed to the inexorableincrease in the equality of conditions among hisFrench countrymen “at intervals of fifty years, begin-ning with the eleventh century.” Nor was this revolu-tion taking place in only his own country: “Whitherso-ever we turn our eyes,” he wrote, “we shall witness thesame continual revolution throughout the whole ofChristendom.” He goes on to say, “The gradual devel-opment of the equality of conditions is . . . a providen-tial fact, and it possesses all the characteristics of a Di-vine decree: it is universal, it is durable, it constantlyeludes all human interference, and all events as well asmen contribute to its progress.”7

We may wish to grant Tocqueville a certain meas-ure of hyperbole in this passage. We may also want to note that in the second volume that he published


06dahl.121_140 11/27/01 4:40 PM Page 126

several years later, he was more troubled by what he viewed as some of the undesirable consequences of democracy and equality. I shall return to his con-cerns in a moment. Even so, he did not doubt that acontinuing advance of democracy and equality was in-evitable. And if we look back today to the changessince his time, like Tocqueville in his own day we maywell be amazed at the extent to which ideas and prac-tices that respect and promote political equality haveadvanced across so much of the world—as, for thatmatter, have some aspects of a broader human equal-ity as well.

As to political equality, consider the incrediblespread of democratic ideas, institutions, and practicesduring the century that just ended. In 1900, 48 coun-tries were fully or moderately independent countries.Of these, only 8 possessed all the other basic institu-tions of representative democracy, and in only 1 ofthese, New Zealand, had women gained the right tovote. Furthermore, these 8 countries contained nomore than 10–12 percent of the world’s population. Atthe opening of our present century, among some 190countries the political institutions and practices ofmodern representative democracy, including universalsuffrage, exist in around 85, at levels comparable tothose in Britain, western Europe, and the UnitedStates. These countries include almost 5 out of every10 inhabitants of the globe today.8

In Britain, the working classes and women wereenfranchised, and more. Men and women of middle,


06dahl.121_140 11/27/01 4:40 PM Page 127

lower-middle, and working-class origins gained accessnot only to the House of Commons and its facilitiesbut to the cabinet and even the post of prime minister.And the hereditary peers in the House of Lords have,after all, at last been sent packing—well, most of themanyway. In the United States, too, women were en-franchised. The Civil Rights Act of 1964, which pro-tected the right of African Americans to vote, did infact become law; the law was actually enforced; andAfrican Americans have become a significant force inAmerican political life. I wish I could say that the mis-erable condition of so many Native Americans hadgreatly changed for the better, but that sad legacy ofhuman injustice remains with us.

Although we must admit to persistent failures andcontinuing obstacles, if we assume that beliefs aboutequality are hopelessly anemic contestants in thestruggle against the powerful forces that generate in-equalities, we cannot account for the enormous gains inhuman equality achieved over the past two centuries.

How Does Greater Political Equality Come About?

In the face of so many obstacles, how does greaterequality—or better, a reduction in some inequalities—ever come about? Although no brief summary can dojustice to an explanation of the historical variations andcomplexities in the process by which changes toward


06dahl.121_140 11/27/01 4:40 PM Page 128

equality take place—and here I have in mind mainlypolitical equality—a summary of the most importantelements would probably run something like this:

Despite fervent efforts by privileged elites to pro-mote views intended to give legitimacy to their supe-rior power and status, together with their own unques-tioning belief in the rightness of their entitlements(think of the Federalists!), many members of subordi-nate groups doubt that the inferior position assignedto them by their self-proclaimed superiors is really jus-tified. James Scott has shown pretty convincingly thatpeople who have been relegated to subordinate statusby history, structure, and elite belief systems are muchless likely to be taken in by the dominant ideologythan members of the upper strata are prone to as-sume.9 Given the open or concealed rejection of theelite ideology by members of the subordinate groups,a change in conditions, whether in ideas, beliefs, gen-erations, structures, resources, or whatever, begins tooffer the subordinate groups new opportunities to ex-press their grievances. And given these new opportu-nities and moved by anger, resentment, a sense of injustice, a prospect of greater individual or group opportunities, group loyalty, or other motives, somemembers of the subordinate groups begin to press forchange by whatever means are available. Some mem-bers of the dominant group begin to support theclaims of the subordinate strata. Privileged insidersally themselves with outsiders. Insiders may do so for avariety of reasons: moral convictions, compassion, op-


06dahl.121_140 11/27/01 4:40 PM Page 129

portunism, fear of the consequences of disorder, dan-gers to property and the legitimacy of the regime aris-ing from widening discontent, and even the real orimagined possibility of revolution.10

So a seismic shift occurs: extension of the fran-chise, legal protection of basic rights, political compe-tition from leaders of hitherto subordinate groups,election to public office, changes in law and policy,and so on. In the United States, Civil Rights Acts werepassed in 1957, 1960, and most crucial of all, 1964.What is more, they were enforced. African Americansbegan to seize their opportunities to vote—and amongother things soon tossed out the police officials whohad violently enforced their subordination. In India,the scheduled castes have begun to vote in substantialnumbers for leaders and parties who are drawn fromtheir own strata and committed to reducing discrimi-nation against them. Though changes toward equalitymay be and typically are incremental, a series of incre-mental changes can, in time, mount to a revolution.

By such processes, then, a certain measure of po-litical equality and democracy have been obtained insome countries despite enormous and persistent ob-stacles to human equality.

Is Political Equality a Justifiable Goal?

Yet even if a greater degree of political equality anddemocracy can be achieved, are these goals really de-


06dahl.121_140 11/27/01 4:40 PM Page 130

sirable? What’s more, are they so desirable that weshould make the constitution of a democratic country—in particular that of the United States—subordinate toachieving these ends?

The desirability of political equality and thus ofdemocracy follows, in my view, from two fundamentaljudgments. One is moral, the other practical.

The moral judgment holds that all human beingsare of equal intrinsic worth; that no person is intrinsi-cally superior in worth to another; and that the goodor interests of each person ought to be given equalconsideration.11 Let me call this the assumption of in-trinsic equality.

Yet if we accept this moral judgment, a deeplytroublesome question immediately arises: Who or whatgroup is the best qualified to decide what the good orinterests of a person really are? Clearly the answer willvary, depending on the situation, the kinds of deci-sions, and the persons involved. But if we restrict ourfocus to the government of a state, then it seems to methat the safest and most prudent assumption wouldrun something like this: Among adults, none are sobetter qualified than others to govern that they shouldbe entrusted with complete and final authority overthe government of the state.12

Although we might reasonably add refinementsand qualifications to this prudential judgment, it is dif-ficult for me to see how a significantly different propo-sition could be defended, particularly if we draw oncrucial historical cases in which substantial numbers of


06dahl.121_140 11/27/01 4:40 PM Page 131

persons were denied equal citizenship. Does anyonetoday really believe that when the working classes,women, and racial and ethnic minorities were excludedfrom political participation, their interests were ade-quately considered and protected by those who wereprivileged to govern over them?

Does Political Equality Threaten Liberty?

Like many desirable goals, political equality mightconflict with—and may indeed do harm to—other im-portant goals, ends, values. If so, shouldn’t our pursuitof political equality be tempered by our justifiable de-sire to attain these other goals?

It is frequently said that equality conflicts with lib-erty and fundamental rights. Like many others, Toc-queville appears to have believed so.

But before I turn to his remarks, I cannot forgoadding that I am amazed by a frequent assertion aboutthe supposed conflict between liberty and equality thatmakes no mention of what would seem to me to be anabsolutely essential requirement of any reasonable dis-cussion about the relation between the two. Wheneverwe talk about liberty, freedom, or rights, are we notobliged to answer the question: Liberty or rights forwhom? When we speak of freedom, liberty, or rights itseems to me essential that we go beyond answeringthe question, “What liberty or right?” An answer tothat question only specifies the domain of liberty. But


06dahl.121_140 11/27/01 4:40 PM Page 132

we are also obliged to answer the question, “Libertyfor whom?”13

Keeping this question in mind, let me return toTocqueville. His view, if I understand him correctly,was roughly this: An equality of condition among apeople helps to make democracy possible, perhapseven inevitable. But the very equality of condition thatmakes democracy possible also carries dangers to lib-erty. Let me paraphrase Tocqueville:

Since the very essence of democratic government is theabsolute sovereignty of the majority, which nothing indemocratic states is capable of resisting, a majority nec-essarily has the power to oppress a minority. Just as aman with absolute power may misuse it, so may a ma-jority. Given an equality of condition among citizens,we may expect that in democratic countries a whollynew species of oppression will arise. Among citizens allequal and alike, the supreme power, the democraticgovernment, acting in response to the will of the major-ity, will create a society with a network of small compli-cated rules, minute and uniform, that none can escape.Ultimately, then, the citizens of a democratic countrywill be reduced to nothing better than a flock of timidand industrious animals, of which the government is theshepherd.14

If I have fairly summarized Tocqueville, howshould we interpret his forecast in the light of subse-quent developments? After all, we have the advantage,as he did not, of two centuries of experience with mod-ern democratic institutions. Some readers have inter-preted these passages in Tocqueville as foreshadowing


06dahl.121_140 11/27/01 4:40 PM Page 133

mass society, while to others Tocqueville expected thatmass democracy would be the seed of twentieth-century authoritarian and totalitarian systems. Yet, ifwe read the passages as a forecast of the way in whichdemocratic countries would tend to evolve, I think weare bound to conclude that Tocqueville was just deadwrong. When we examine the course of democraticdevelopment over the past two centuries, and particu-larly over the century just ended, what we find is apattern of democratic development that stands in totalcontradiction to such a prediction. We find insteadthat as democratic institutions become more deeplyrooted in a country, so do fundamental political rights,liberties, and opportunities. As a democratic govern-ment matures in a country, the likelihood that it willgive way to an authoritarian regime approaches zero.Democracy can, as we all know, collapse into dictator-ship. But breakdowns are extraordinarily rare in ma-ture democracies; they occur instead in countries thatencounter times of great crisis and stress when theirdemocratic institutions are relatively new. Crisis ap-pears to be inevitable in the life of every country. Evenmature democratic countries have had to face wars,economic depression, large-scale unemployment, ter-rorism, and other challenges. But they did not collapseinto authoritarian regimes.

In the twentieth century, on something like sev-enty occasions democracies have given way to nonde-mocratic regimes. Yet with very few exceptions, thesebreakdowns have occurred in countries where demo-


06dahl.121_140 11/27/01 4:40 PM Page 134

cratic institutions were very new—less than a genera-tion old. Indeed, the only clear-cut case of a demo-cratic breakdown in a country where democratic insti-tutions had existed for twenty years or more seems tobe Uruguay in 1973. In the same year, Chile provideda less clear-cut case because of restrictions on the suf-frage that had only recently been lifted. The WeimarRepublic had existed fewer than fourteen years beforethe Nazi takeover. In all three countries the path to col-lapse bore no relation to the Tocquevillean scenario.

Nor, as we know, is that scenario confirmed by theolder or mature democracies. As I indicated in theprevious chapter, we can find some small variationsamong these countries in their protection of basicrights. But they all maintain these rights well abovethe threshold necessary for democracy. Have the fun-damental rights and liberties of citizens grown steadilynarrower or less secure over the past half-century? Ido not see how an affirmative answer to this questioncould be seriously maintained. Much as I admire Toc-queville, on this issue, he, like the Framers, could notforesee the future of democratic government.

Far from being a threat to fundamental rights andliberties, political equality requires them as anchorsfor democratic institutions. To see why this is so, letme once again view democracy as, ideally at least, apolitical system designed for citizens of a state who arewilling to treat one another, for political purposes, aspolitical equals. Citizens might view one another asunequal in other respects. Indeed, they almost cer-


06dahl.121_140 11/27/01 4:40 PM Page 135

tainly would. But if they were to assume that all citi-zens possess equal rights to participate, directly or in-directly through their elected representatives, in mak-ing the policies, rules, laws, or other decisions thatcitizens are expected (or required) to obey, then thegovernment of their state would, ideally, have to satisfyseveral criteria.

Let me list them here without amplification. To befully democratic, a state would have to provide: rights,liberties, and opportunities for effective participation;voting equality; the ability to acquire sufficient under-standing of policies and their consequences; and themeans by which the citizen body could maintain ade-quate control of the agenda of government policiesand decisions. Finally, as we now understand the ideal,in order to be fully democratic, a state would have toensure that all, or at any rate most, permanent adultresidents under its jurisdiction would possess the rightsof citizenship.

As we know, the democratic ideal that I have justdescribed is too demanding to be achieved in the ac-tual world of human society. To accomplish it as far asmay be possible under the imperfect conditions of thereal world, certain political institutions for governingthe state would be required. Moreover, since the eigh-teenth century these institutions have had to be suit-able for governing a state encompassing a large terri-tory, such as a country.

There is no need to describe here the basic politi-cal institutions of a modern democratic country; but it


06dahl.121_140 11/27/01 4:40 PM Page 136

should be obvious that just as in the ideal, so too in ac-tual practice, democratic government presupposes thatits citizens possess a body of fundamental rights, liber-ties, and opportunities. These include the rights to votein the election of officials in free and fair elections; torun for elective office; to free expression; to form andparticipate in independent political organizations; tohave access to independent sources of information;and to have rights to other freedoms and opportunitiesthat may be necessary for the effective operation ofthe political institutions of large-scale democracy.

Both as an ideal and as an actual set of political in-stitutions, democracy is necessarily, then, a system ofrights, liberties, and opportunities. These are requirednot merely by definition. They are required in orderfor a democratic system of government to exist in thereal world. If we consider these political rights, liber-ties, and opportunities as in some sense fundamental,then in theory and in practice, democracy does notconflict with liberty. On the contrary, democratic insti-tutions are necessary for the existence of some of ourmost fundamental rights and opportunities. If thesepolitical institutions, including the rights, liberties, andopportunities they embody, do not exist in a country,then to that extent the country is not democratic.When they disappear, as they did in Weimar Germany,Uruguay, and Chile, then democracy disappears; andwhen democracy disappears, as it did in these coun-tries, then so do these fundamental right, liberties, andopportunities. Likewise, when democracy reappeared


06dahl.121_140 11/27/01 4:40 PM Page 137

in these countries so, necessarily, did these fundamen-tal rights, liberties, and opportunities. The connection,then, is not in any sense accidental. It is inherent.

The links between political equality and democ-racy, on the one hand, and fundamental rights, liber-ties, and opportunities, on the other, run even deeper.If a country is to maintain its democratic institutionsthrough its inevitable crises, it will need a body ofnorms, beliefs, and habits that provide support for theinstitutions in good times and bad—a democratic cul-ture transmitted from one generation to the next. Buta democratic culture is unlikely to be sharply bounded.A democratic culture will support not only the funda-mental rights, liberties, and opportunities that demo-cratic institutions require. People who share a demo-cratic culture will, I think inevitably, also endorse andsupport an even greater sphere of rights, liberties, andopportunities. Surely the history of recent centuriesdemonstrates that it is precisely in democratic coun-tries that liberties thrive.

� � �

IF WE BELIEVE THAT ALL HUMAN BEINGS ARE CREATED

equal, that they are endowed with certain inalienablerights, that among these are life, liberty, and the pur-suit of happiness, that to secure these rights govern-ments are instituted among a people, deriving theirjust powers from the consent of the governed, then weare obliged to support the goal of political equality.


06dahl.121_140 11/27/01 4:40 PM Page 138

Political equality requires democratic political in-stitutions.

The supposed conflict between liberty and politi-cal equality is spurious, first, because an inherent partof democratic political institutions is a substantial bodyof fundamental rights, liberties, and opportunities; and,second, because a people committed to democracy andits political institutions will almost certainly expandthe sphere of fundamental rights, liberties, and oppor-tunities well beyond those strictly necessary for de-mocracy and political equality.

Among a people committed to democracy and po-litical equality, a constitution should serve those endsby helping to maintain political institutions that fosterpolitical equality among citizens and all the necessaryrights, liberties, and opportunities that are essential tothe existence of political equality and a democraticgovernment.


06dahl.121_140 11/27/01 4:40 PM Page 139

06dahl.121_140 11/27/01 4:40 PM Page 140

chapter 7

Some Reflections on the Prospects for a More Democratic Constitution

I N A 1987 SURVEY THAT REVEALED STRONG SUPPORT

among Americans for the Constitution on thewhole, the results of one question stand out. When

respondents were asked, “How good a job has [the sys-tem of government established by the Constitution]done in treating all people equally?” Fifty-one percentanswered that it had done a bad job, 8 percent ex-pressed no opinion, and a minority of 41 percent saidthat it had done a good job.1

If we want to enjoy a system of government thatperforms better in treating all people equally—at leastin their roles as democratic citizens—what might wedo? As I said at the beginning of Chapter 1, my aim inthese essays is not to offer a set of specific proposalsfor changes in the Constitution but rather to encour-

07dahl.141_158 11/27/01 4:41 PM Page 141

age a change in the way we think about our Constitu-tion. How then might we begin to think realisticallyabout changing it? What possibilities would actually beopen to a body of framers in the early years of thetwenty-first century? What limits on the range of pos-sibilities should they expect to confront?

The Limited Role of a Constitution

To begin with, tomorrow’s constitutional reformerswould be wise to recognize that no matter what a consti-tution prescribes on paper, it can achieve only a limitedrange of goals. For example, as I pointed out in Chapter5, no constitution can ensure democracy in a countrywhere the conditions favorable to democracy are ab-sent. To preserve and improve these favorable condi-tions would accomplish far more in achieving a moredemocratic order than any changes in the constitution.

The Framers of 1787 were well aware of their lim-its. One of the striking features of the constitutionthey wrote is its admirable brevity. Depending on thestyle in which it is printed, the written text runs fifteento twenty pages, plus another five to seven pages ofamendments. What enables the written Constitutionto achieve this brevity is its almost exclusive focus onjust three matters: structures, powers, and rights.

Most of the Constitution is devoted to the first twomatters, structures and the powers allocated to thesestructures. Provisions about the third matter, rights, are

142 p ro s p e c t s f o r a m o r e d e m o c r at i c co n st i t u t i o n

07dahl.141_158 11/27/01 4:41 PM Page 142

found mainly in the Bill of Rights and in later amend-ments. An important feature of these constitutionalrights is that they are guaranteed almost entirely byimposing constitutional limits on the government. TheConstitution tacitly assumes that citizens themselveswill somehow possess the opportunities and resourcesnecessary in order for them to act on their rights. I’llcome back to this assumption in a moment.

Constitutional Structures

Let me say a few words about structures. I am inclinedto believe that three, and possibly four, structural ele-ments in our constitutional system are not, realisticallyspeaking, open to change in the foreseeable future.

Federalism. One element is our federal system.Just as the Framers knew that they could not abolishthe states, constitutional reformers today should prob-ably assume that the states will remain—and in myjudgment they should remain—as fundamental unitsin a federal government that are endowed with signifi-cant powers. As has been true for two centuries, thequestion of how power is shared between the federalgovernment and the states will persist as a subject ofendless dispute. But I do not believe that constitu-tional framers today would or should attempt to dis-solve the existing states.

Presidentialism. A second limit on constitutionalreform, I believe, is our presidential system. We may

p ro s p e c t s f o r a m o r e d e m o c r at i c co n st i t u t i o n 143

07dahl.141_158 11/27/01 4:41 PM Page 143

be able to alter it a bit around the edges with amend-ments or changed practices, but the option of a parlia-mentary system is, I would guess, simply not attractiveto most Americans. We have so deeply invested our-selves in the mythical aspects of the presidency thatshort of some constitutional breakdown, which I nei-ther foresee nor, certainly, wish for, we won’t seriouslyconsider changing it. For better or worse, we Ameri-cans are stuck with a presidential system.

Inequality in representation. The other featurethat I fear constitutional reformers would be unablechange is the gross inequality in representation result-ing from the fixed allocation of two senators to eachstate without regard to population. Let me remind youagain that Section 3 of Article I reads: “The Senate ofthe United States shall be composed of two Senatorsfrom each State for six Years, and each Senator shallhave one Vote.” A constitutional amendment to changethis provision faces two formidable obstacles. The firstmakes it most unlikely that any constitutional amend-ment to change the composition of the Senate will beadopted. The second makes it virtually impossible.First, under Article V of the Constitution, you may re-call, amendments can be proposed only with a two-thirds vote in both houses or by a convention sup-ported by two-thirds of the state legislatures; and theycan be adopted only after ratification by the legisla-tures or conventions in three-fourths of the states. Ican’t help thinking that at least thirteen of the least


07dahl.141_158 11/27/01 4:41 PM Page 144

populated states would exercise their veto to preventthe adoption of any amendment that would reducetheir influence in the Senate. But in the highly im-probable event that such an amendment might some-how make its way past this formidable obstacle, thesecond barrier promises to be totally impregnable.“No State, without its Consent,” Article V of the Con-stitution concludes, “shall be deprived of its equal Suf-frage in the Senate.”

In effect, those fifteen words end all possibility ofamending the constitution in order to reduce the un-equal representation of citizens in the Senate. So weare destined, it seems, to be saddled indefinitely witha greater degree of unequal representation in theupper house than exists in any of the other establisheddemocracies.

If I am correct about these three fixed elements—federalism, presidentialism, and unequal representa-tion in the Senate—they in turn seem likely to imposeother limits on what today’s framers might accomplish.

The electoral college. For example, what can wedo about changing the electoral college? In Chapter 4 Ishowed how the inequality of representation in the Sen-ate reproduces itself, though somewhat weakened, inthe electoral college. It plays out yet again in diminish-ing the chances that the Constitution can be amendedto replace the electoral college with a system of popu-lar election. Consequently, I suggested, the relativedesirability of three possible solutions to the problem


07dahl.141_158 11/27/01 4:41 PM Page 145

of unequal representation in the electoral college is in-versely related to the probability of their enactment.

Consensual, majoritarian, or neither? I have al-ready expressed some sympathy toward a consensualsystem as an alternative to a majoritarian system. TheUnited States, I suggested, is neither one nor the other.It is a hybrid that just might have the vices of both andthe virtues of neither.

That our political leaders manage much of thetime to avoid complete deadlock and make the systemwork—more or less—is testimony, I think, to their ex-ceptional political skills, which tend to be vastly under-valued in the media and among ordinary citizens. Ihave also suggested that the necessary wheeling anddealing, the inevitable behind-the-scenes compro-mises, and the unavoidable gap between public rheto-ric and the mutual concessions among insiders, resultin a political system so opaque and so at odds withgeneral conceptions of public virtue that it weakensboth civic understanding and citizens’ confidence inour political institutions.

If we are unlikely to change either our presidentialsystem or our severely unequal representation in theSenate, then it will also be very difficult to bring aboutsome constitutional options that otherwise could be,and should be, seriously considered. In particular, Iam not entirely sure that we can redesign our presenthybrid so that it facilitates either greater consensus orstronger majoritarianism.


07dahl.141_158 11/27/01 4:41 PM Page 146

Neither alternative can exist without an appropri-ate political culture. Yet, a system designed to be con-sensual would harbor a serious danger if it were notstrongly implanted in a political culture that fostersagreement. Lacking the appropriate political culture, aconstitutional design intended for a consensual systemwould enable a minority to veto any changes from thestatus quo that threaten its privileges, as the Southernstates did before the Civil War. Or a regionally privi-leged minority could extort concessions from the ma-jority by threatening to use its veto, as the Southernstates did when they compelled the rest of the countryto abandon efforts to protect the civil rights of African-Americans after the Civil War.

I am inclined to believe that our political culture—unlike, say, that of Sweden, Switzerland, or Holland—would prevent a consensual design from realizing itspotential advantages.

The corresponding fear about a strictly majoritariansystem is that it might fail to provide the leaders of amajority with adequate incentives to seek greater con-sensus before they invoke their power as a majority.I’m not raising again here the issue of liberty versus po-litical equality that I discussed earlier. Even if the lead-ers of a majority were to maintain the fullest respectfor the democratic rights of minorities, they might seeslight reason to explore options that would achieve awider range of agreement and support than they needto push through a law or policy by majority vote.


07dahl.141_158 11/27/01 4:41 PM Page 147

The ideal solution, it seems to me, would be a po-litical system that provides strong incentives for politi-cal leaders to search for the broadest feasible agree-ment before adopting a law or policy and yet allowsthe decision to be made, if need be, by majority vote—always, of course, within the limits set by the need topreserve fundamental democratic rights. No majorityshould have the right, moral or constitutional, to fore-close decisions by future majorities.

Here again, I fear that this ideal solution may notbe open to us so long as severely unequal representa-tion in the Senate allows some geographical minoritiesto block decisions by representatives elected by a ma-jority of their fellow citizens. Because of this minorityveto, the search for consensus could easily turn intowhat might be harshly described as extortion and black-mail by a minority of Senators.

Although this barrier to majority rule may be im-possible to change, rules of the Senate that furthercompound the power of privileged minorities mightbe more amenable to change, as, in my view, they def-initely should be. For example, it is doubtful that ourfutile and counterproductive policy toward Cuba wouldhave remained so long in effect were it not for theability of a handful of U.S. Senators to extort from Con-gress and the president the policies they favor, in re-turn for their support on other issues.

So we arrive at this unhappy conclusion: The Fram-ers of 1787 appear to have limited today’s framers to asystem that is neither consensual nor majoritarian but


07dahl.141_158 11/27/01 4:41 PM Page 148

is a hybrid that possesses the vices of both and thevirtues of neither.

Hidden costs and uncertainties. The difficulty oftransforming a long-established political culture intoone appropriate to a new and different constitutionalstructure—one more consensual, say, or more majori-tarian—illustrates yet another problem. Major consti-tutional change involves large hidden costs and a greatuncertainty. The hidden costs arise because of theneed to abandon familiar habits, practices, beliefs, andunderstandings that exist among the political elitesand that are embedded in the popular culture as well.Creating an appropriate political culture may be al-most as far beyond the capacities of constitutionalframers today as it was for the Framers in 1787. Whatis more, even if our knowledge about the likely conse-quences of alternative constitutional structures is im-measurably better than that of the Framers, after twocenturies of experience with different democratic con-stitutions predicting the outcome of major changesremains fraught with considerable uncertainty.

Powers

Are the constitutional powers of the states, the federalgovernment, and the three main branches of the fed-eral government appropriate to our democratic needsand values today? An attempt to answer this dauntingquestion would so far exceed my limits here that I shallsimply call attention to its relevance and importance.


07dahl.141_158 11/27/01 4:41 PM Page 149

Rights

As with powers, the subject of rights is so vast that Ican do no more than sketch a view that seems to meuseful for appraising fundamental rights in the contextof a democratic country’s constitution.

It is a standard view in jurisprudence that rightsimply duties: in order for a right to be effectively exer-cised, government officials and others must assumethe duty of protecting the right against persons whoseek to violate it. A less-common assumption, but onepresent by implication, is that rights also imply oppor-tunities: your right to vote is meaningless if you don’tactually have an opportunity to vote. So too with free-dom of expression. What would a right to free speechmean to you if you didn’t have any opportunity tospeak freely?

Now I want to add a fourth element to our discus-sion. To rights, duties, and opportunities I would addresources.2 Suppose, for example, that voting boothswere placed by officials in locations far from thehomes of many citizens and were open for only onehour in mid-morning: most citizens would lack bothopportunities and resources necessary to cast a vote.They would be outraged, and so would you and I.

To clarify further what I have in mind when I referto resources, I’m going to invite you to contemplate afictional scenario.

Say that we are all citizens in a New England townwith a traditional town meeting. As usual, a modest

150

07dahl.141_158 11/27/01 4:41 PM Page 150

proportion of the citizens eligible to attend have actu-ally turned out, let’s say four or five hundred.

After calling the meeting to order, the moderatorannounces:

“We have established the following rules for thisevening’s discussion. After a motion has been properlymade and seconded, in order to ensure free speechunder rules fair to everyone here, each of you whowishes to do so will be allowed to speak on the motion.However, to enable as many as possible to speak, no onewill be allowed to speak for more than two minutes.”

Perfectly fair so far, you might say. But now ourmoderator goes on:

“After everyone who wishes to speak for two minuteshas had the floor, each and every one of you is free tospeak further, but under one condition. Each additionalminute will be auctioned off to the highest bidder.”

The ensuing uproar from the assembled citizenswould probably drive the moderator and the board ofselectman away from the town hall—and perhaps outof town.

Yet isn’t this in effect what the Supreme Court decided in the famous case of Buckley v. Valeo? In a seven-to-one vote, the court held that the FirstAmendment–guarantee of freedom of expression wasimpermissibly infringed by the limits placed by theFederal Election Campaign Act on the amounts thatcandidates for federal office or their supporters might


07dahl.141_158 11/27/01 4:41 PM Page 151

spend to promote their election.3 Well, we’ve had timeto see the appalling consequences.

What went wrong? The justices failed to view cam-paign expenditures and contributions in the context ofa democratic system that derives its legitimacy fromthe principles of political equality that I described ear-lier. In order to exercise the fundamental rights towhich citizens in a democratic order are entitled—to vote, speak, publish, protest, assemble, organize,among others—citizens must also possess the minimalresources that are necessary in order to take advantageof the opportunities and to exercise their rights.

The problem of matching resources to democraticrights admits of no easy answer. Nor can the problembe solved merely by constitutional prescription. Butsurely a constitution is deeply flawed if the highestcourt in the land can interpret it to impose an im-moveable barrier to the achievement of a satisfactorydegree of political equality among its citizens.

A Democratic Role for the Supreme Court

In earlier chapters I alluded to a problem that ourbest legal and constitutional scholars have disputed atlength yet remains with us still. In American constitu-tional circles it sometimes travels under the name of“the counter-majoritarian difficulty.”

I’ll put the difficulty this way. We cannot simulta-neously lodge the authority to make laws and policies


07dahl.141_158 11/27/01 4:41 PM Page 152

exclusively in the hands of elected officials who are, atleast in principle, accountable to citizens through elec-tions and at the same time give the judicial branch theauthority, in effect, to make crucial public policies.That dilemma presents us with a difficult choice. ManyAmericans will resist making it. But if we were ever toundertake a discussion about the adequacy of our con-stitution when we assess it against democratic stan-dards, this problem, which so far been discussedmainly among legal scholars, would have to be openedup to public debate and discussion.

There is, I believe, an important place in a demo-cratic country for a supreme court with the authorityto review the constitutionality of legislative and ad-ministrative enactments. For one thing, a federal sys-tem needs a high court empowered to decide if andwhen state authorities have exceeded their properbounds. But a supreme court should also have the au-thority to overturn federal laws and administrative de-crees that seriously impinge on any of the funda-mental rights that are necessary to the existence of ademocratic political system: rights to express one’sviews freely, to assemble, to vote, to form and to par-ticipate in political organizations, and so on.4

When the court acts within this sphere of funda-mental democratic rights, the legitimacy of its actionsand its place in the democratic system of governmentcan hardly be challenged. But the more it moves out-side this realm—a vast realm in itself—the more du-bious its authority becomes. For then it becomes an


07dahl.141_158 11/27/01 4:41 PM Page 153

unelected legislative body. In the guise of interpretingthe Constitution—or, even more questionable, divin-ing the obscure and often unknowable intentions ofthe Framers—the high court enacts important lawsand policies that are the proper province of elected officials.

Even within the realm of fundamental democraticrights the decisions of the court will arouse contro-versy. And controversy is all the more likely becauseour understanding of democratic rights will surelycontinue to evolve.

Is Significant Change Possible?

My reflections lead me to a measured pessimism aboutthe prospects for greater democratization of the Ameri-can Constitution. Changes described in this chapter thatwould be desirable from a democratic point of viewseem to me to have very little chance of coming aboutin the indefinite future. Although my judgments aboutlikelihoods are necessarily subjective, I believe thatmost others familiar with American political life wouldconcur.

The likelihood of reducing the extreme inequalityof representation in the Senate is virtually zero. Thechances of altering our constitutional system to makeit either more clearly consensual or more definitelymajoritarian are also quite low. The likelihood is verylow that the Supreme Court will refrain from legislat-


07dahl.141_158 11/27/01 4:41 PM Page 154

ing public policies, often highly partisan ones, and in-stead focus its power of judicial review strictly on theprotection of fundamental democratic rights and is-sues of federalism. The combination of chief executiveand monarchy in the American presidency is not likelyto change. Finally, the probability that democraticchanges in the electoral college will occur appear to beinversely related to their desirability, with the most de-sirable having the lowest probability of occurring.There is at least a modest chance that some statesmight require their electoral votes to be allocated inproportion to the popular votes. But a constitutionalamendment that makes the number of a state’s elec-tors proportionate to its population stands little chanceof adoption. And the inequality in representation inthe Senate makes a constitutional amendment provid-ing for direct popular election of the president virtu-ally impossible.

Those who regard the Constitution as a sacredicon will, no doubt, take comfort from this conclusion:my pessimism is the mirror image of their optimism.

Yet the historic, if fitful, American impulse towarddemocracy and political equality has not come to anend. And so, we face a challenge. Given the presentlimits I have described, how might we advance towarda fuller achievement of democratic processes, rights,liberties, opportunities, and resources?

Let me suggest two very general strategies.First, it is time—long-past time—to invigorate

and greatly widen the critical examination of the Con-


07dahl.141_158 11/27/01 4:41 PM Page 155

stitution and its shortcomings. Public discussion thatpenetrates beyond the Constitution as a national iconis virtually nonexistent. Even when in-depth analysisdoes occur—mainly among constitutional scholars inschools of law and departments of political science andhistory—the Constitution as a whole is rarely testedagainst democratic standards or against the perform-ance of constitutional systems in other advanced dem-ocratic countries.

I can envision the possibility—here a degree ofoptimism breaks through—of a gradually expandingdiscussion that begins in scholarly circles, moves out-ward to the media and intellectuals more generally,and after some years begins to engage a wider public.I cannot say what the outcome might be. But surely itwould heighten understanding of the relevance ofdemocratic ideas to the constitution of a democraticcountry, and specifically it would heighten understand-ing of the shortcomings of the existing constitutionviewed from that perspective and of the possibilities ofchange.

Meanwhile, however, we need a second strategy,one designed to achieve greater political equality withinthe limits of the present American Constitution. A majorobjective of such a strategy would be to reduce thevast inequalities in the existing distribution of politicalresources. The characteristics of the Constitution thatI have described in this book will, of course, stand asobstacles to the success of such a strategy, for they armthose who possess the greatest resources with strong


07dahl.141_158 11/27/01 4:41 PM Page 156

defenses—opportunities to veto changes—against allefforts to reduce their privileged positions.

I cannot foresee how successful either strategymay prove to be. But the belief of most Americansthat a democratic government, warts and all, is betterin the end than any feasible alternative to it is justifiedon fundamental principles of human equality that can-not be tightly bounded. Our understanding of the im-plications of those principles will therefore continue to evolve indefinitely. So, too, will the implications ofthose principles for our democratic political system,and its Constitution, under which we Americans freelychoose to live.


07dahl.141_158 11/27/01 4:41 PM Page 157

07dahl.141_158 11/27/01 4:41 PM Page 158

appendix a

On the Terms “Democracy” and “Republic”

T HE VIEW THAT THE FRAMERS INTENDED TO

create a republic, not a democracy, probablyhas its origins in comments by Madison in

Federalist No. 10. Although there as elsewhere he alsoused the expression “popular government” as a kind ofgeneric term, he distinguished further between “apure democracy, by which I mean a society consistingof a small number of persons, who assemble and ad-minister the government in person,” and a “republic,by which I mean a government in which the scheme ofrepresentation takes place.” “The two great points ofdifference between a democracy and a republic are:first, the delegation of the government, in the latter, to a small number of citizens elected by the rest; secondly, the greater number of citizens, and greater

08dahlAppA.159_162 11/27/01 4:41 PM Page 159

sphere of the country, over which the latter may be ex-tended.”1

Here Madison was making the common distinctionthat political scientists and others would later differen-tiate as “direct democracy” and “representative de-mocracy.” For it was as evident to the Framers as it isto us that given the size of a nation composed of thethirteen existing states, with more to come, “thepeople” could not possibly assemble directly to enactlaws, as they did at the time in New England townmeetings and had done two millennia earlier in Greece,where the term “democracy” was invented. It was per-fectly obvious to the Framers, then, that in such alarge country, a republican government would have tobe a representative government, where national lawswould be enacted by a representative legislative bodyconsisting of members chosen directly or indirectly bythe people.

Madison was probably also influenced by a longtradition of “republicanism” that in both theory andpractice leaned somewhat more toward aristocracy, lim-ited suffrage, concern for property rights, and fear ofthe populace than toward a broadly based popular gov-ernment more dependent on “the will of the people.”

It is also true, however, that during the eighteenthcentury the terms “democracy” and “republic” wereused rather interchangeably in both common andphilosophical usage.2 Madison, in fact, was well awareof the difficulty of defining “republic.” In FederalistNo. 39, he posed the question “What, then, are the

160 o n t h e t e r m s “ d e m o c r acy ” a n d “ r e p u b l i c ”

08dahlAppA.159_162 11/27/01 4:41 PM Page 160

distinctive characters (sic) of the republican form?” Inresponse he pointed to the enormous range of mean-ings given to the word “republic.” “Were an answer tothis question to be sought . . . in the application of theterm by political writers, to the constitutions of differ-ent states, no satisfactory one could ever be found.Holland, in which no particle of the supreme authorityis derived from the people, has passed almost univer-sally under the denomination of a republic. The sametitle has been bestowed on Venice, where absolutepower over the great body of the people is exercised,in the most absolute manner, by a small body of here-ditary nobles.”

In view of this ambiguity, Madison proposed that“we may define a republic to be . . . a governmentwhich derives all its powers directly or indirectly fromthe great body of the people, and is administered bypersons holding their offices during pleasure, or for alimited period, or during good behavior.”3 By defininga republic as a government which derives all its pow-ers “directly or indirectly from the great body of thepeople,” Madison now seems to be contradicting thedistinction he had drawn earlier in Federalist No. 10.We might read his struggle with definitions as a fur-ther illustration of the prevailing confusion over thetwo terms.

If further evidence were needed of the ambiguityof terminology, we could turn to a highly influentialwriter whose work was well known to Madison andmany of his contemporaries. In The Spirit of the Laws

o n t h e t e r m s “ d e m o c r acy ” a n d “ r e p u b l i c ” 161

08dahlAppA.159_162 11/27/01 4:41 PM Page 161

(1748) Montesquieu had distinguished three kinds ofgovernments: republican, monarchic, and despotic.Republican governments were of two kinds: “When, ina republic, the people as a body have the sovereignpower, it is a Democracy. When the sovereign power isin the hands of a part of the people, it is called an Aris-tocracy.”4 But Montesquieu also insisted that “It is inthe nature of a republic that it has only a small terri-tory: without that it could scarcely exist.”5

Although the Framers differed among themselvesas to how democratic they wanted their republic to be,6

for obvious reasons they were of one mind about theneed for a representative government. But as eventssoon showed, they could not fully determine just howdemocratic that representative government would be-come—under the leadership of, among others, JamesMadison.

162 o n t h e t e r m s “ d e m o c r acy ” a n d “ r e p u b l i c ”

08dahlAppA.159_162 11/27/01 4:41 PM Page 162

Appendix B

Tables and Figures

09dahlAppB.163_172 11/27/01 4:41 PM Page 163

164 ta b l e a n d f i g u r e s

TABLE 1. Countries Steadily Democratic Since at Least 1950

1. Austria2. Australia3. Belgium4. Canada5. Costa Rica6. Denmark7. Finland8. France9. Germany

10. Iceland11. Ireland12. Israel13. Italy14. Japan15. Luxembourg16. Netherlands17. New Zealand18. Norway19. Sweden20. Switzerland21. United Kingdom22. United States

Note: The countries are broken down as follows: European: Austria, Belgium,Denmark, Finland, France, Germany, Iceland, Ireland, Italy, Luxembourg,Netherlands, Norway, Sweden, Switzerland, United Kingdom (15); English-speaking: Australia, New Zealand, United Kingdom, United States (4); LatinAmerican: Costa Rica; other: Israel, Japan. Although India gained independ-ence in 1947, adopted a democratic constitution, and has, except for one inter-val, maintained its democratic institutions in the face of extraordinary chal-lenges of poverty and diversity, I have omitted it from the list for two reasons.First, continuity was interrupted from 1975 to 1977 when the prime minister,Indira Gandhi, staged a coup d’état, declared a state of emergency, suspendedcivil rights, and imprisoned thousands of opponents. Second, because India isone of the poorest countries in the world, comparisons with the wealthy coun-tries in Table 5 would make little sense.

09dahlAppB.163_172 11/27/01 4:41 PM Page 164

ta b l e s a n d f i g u r e s 165

TABLE 2. How the U.S. Constitutional-PoliticalSystem Compares with the Other 21 Older Democracies

Among the other 21 countries,those that are similar

in this respect

(n) Countries

Federalism

1. Strong federalism 5 Australia, Canada,Germany, Switzerland,Belgium (since 1993)

2. Strongly bicameral 3 All federal: Australia, legislature Germany, Switzerland

3. Significant unequal 4 All federal: Australia, representation in the Canada, Germany, upper house Switzerland

Non-Federal Features

4. Strong judicial review of 2 Canada, Germanynational legislation

5. Electoral system: plurality 2 Britain, Canadaelections in single member districts (FPTP)

6. Strong two-party system: 3 Australia2, New third parties weak1 Zealand,3 Costa Rica

7. Presidential system: single Nonepopularly elected chiefexecutive with important constitutional powers

1Until the 1997 election, held under the new PR system, when the two majorparties received only 61% of the vote. The rest was split almost entirely amongthree minor parties.2Where the total vote cast for third party candidates is ordinarily under 10% ofthe total. 3Until 2000, counting the Liberal National (earlier Liberal-Country) coalitionas one party.

Characteristics ofxxxthe U.S. systemxxx

09dahlAppB.163_172 11/27/01 4:41 PM Page 165


TABLE 3. The Electoral Systems of the AdvancedDemocracies

Plurality in SingleMember

PR: PR: DistrictsList System Variants (FPTP) Variant

1. Austria 1. Australia 1. Canada France(AV)

2. Belgium 2. Germany 2. United (MMP) Kingdom (2 rounds)

3. Costa Rica 3. Ireland 3. United (STV) States

4. Denmark 4. Italy (MMP)

5. Finland 5. Japan (semi PR)

6. Iceland 6. New Zealand (MMP)

7. Israel8. Luxembourg9. Netherlands

10. Norway11. Sweden12. Switzerland

Source: Andrew Reynolds and Ben Reilly, The International IDEA Handbook ofElectoral System Design (1997)

Note: AV = Alternative Vote; MMP = Multimember Proportional; STV = SingleTransferable Vote.

09dahlAppB.163_172 11/27/01 4:41 PM Page 166


TABLE 4. Proportionality vs. Majoritarianism: 20 Democratic Countries

Predominantly PredominantlyProportional Mixed Majoritarian

Austria Ireland AustraliaBelgium Japan CanadaDenmark Spain FranceFinland USA GreeceGermany New Zealand (to 1993)Netherlands United KingdomNew Zealand (1993)NorwaySwedenSwitzerland

Source: Powell, Elections as Instruments of Democracy (2000), p. 41. I havechanged his table to show the adoption of proportionality in New Zealand in1993.

Note: The list is slightly different from Table 1. It includes Spain and Greece,democratized since 1950, and it does not include Costa Rica, Iceland, Israel,or Luxembourg.

09dahlAppB.163_172 11/27/01 4:41 PM Page 167


TABLE 5. Measures of U.S. Performance amongDemocratic Countries

Percentageof CountriesPerforming

U.S. No. of Better than Variable Rank Countries U.S.1

U.S. in BEST third

Economic 5th best (tied) of 18 24growth, 1980–95

U.S. in MIDDLE third

Women’s cabinet 8th most of 22 33representa-tion, 1993–95

Popular support 10th most of 22 43for executive,1945–962

Budget deficits, 8th highest (tied) 16 471970–95 of

Unemployment, 8th highest (tied) 18 591971–953 of

Family policy, 12th best of 18 651976–82

Inflation (CPI), 12th highest of 18 651970–95

09dahlAppB.163_172 11/27/01 4:41 PM Page 168


U.S. in WORST third

Women’s 18th lowest 22 81parliamentary (tied) ofrepresentation, 1971–95

Rich-poor ratio, 4th greatest of 18 821981–93

Energy 19th lowest of 22 86efficiency, 1990–94

Welfare state 17th lowest of 18 94index, 1980

Social 17th lowest of 18 94expenditures, 1992

Voter turnout, 21st lowest of 22 951971–96

Incarceration 1st highest of 18 100rate, 1992–95

Foreign aid, 19th lowest of 19 1001992–95

Source: I am indebted to Arend Lijphart for his permission to use the data sethe developed for Patterns of Democracy (New Haven: Yale University Press,1999) where the variables and their sources are described. I also owe thanks toJennifer Smith for her contributions to the construction of the table. 1The percentages are those of countries scoring better than the United States,and not tied. Variables for which the higher the ranking, the worse a country’sperformance are in italics.2“The average percentage of the voters who gave their votes to the party orparties that formed the cabinet, or in presidential systems, the percentage ofvoters who voted for the winning presidential candidate, weighted by the timeeach cabinet or president was in office.” Lijphart 1999, p. 290.3Unstandardized.

09dahlAppB.163_172 11/27/01 4:41 PM Page 169

FIG

UR

E1.

Pres

iden

tial E

lect

ions

Won

with

Les

s T

han

50%

of t

he P

opul

ar V

ote

09dahlAppB.163_172 11/27/01 4:41 PM Page 170

FIG

UR

E2.

Une

qual

Rep

rese

ntat

ion

in th

e E

lect

oral

Col

lege

Residents per elector (thousands)

Sta

tes

09dahlAppB.163_172 11/27/01 4:41 PM Page 171

09dahlAppB.163_172 11/27/01 4:41 PM Page 172

Notes

CHAPTER 1. Introduction

1. Although in three states—Delaware, New Jersey, andGeorgia—the vote was unanimous, in the rest it was divided,sometimes closely after sharp debate. For example, in Massa-chusetts the delegates split 187 to 168; in New Hampshire, 57to 46; and in Virginia, the state from which several of the prin-cipal authors of the Constitution came, the supporters of theConstitution won by a single vote: 80 to 79.

2. In the ten states where the Convention vote was notunanimous, a total of 1540 delegates voted on the Constitu-tion, 964 for, and 576 against.

CHAPTER 2. What the Framers Couldn’t Know1. I quote here from the Journal of the Convention, in

Max Farrand, ed., The Records of the Federal Convention of1787, 3 vols. (New Haven: Yale University Press, 1966), 1: 1.The three volumes edited by Farrand were reprinted in 1987with a fourth volume, a Supplement edited by James H. Hutson

10dahlNotes.173_190 11/27/01 4:42 PM Page 173

(New Haven: Yale University Press, 1987). My references tothe records of the Convention are drawn from these four vol-umes and are cited hereafter as Records.

2. William Miller, The Business of May Next: James Madi-son and the Founding (Charlottesville: University Press of Vir-ginia, 1992), 41ff. Lance Banning writes that “Madison hadcome to Philadelphia the best prepared of all who gathered forthe Federal Convention. . . . He first prepared elaborate re-search notes on the histories and structures of other ancientand modern confederacies.” 115.

3. Records, 1: 425.4. Madison took extensive notes during the Convention,

which he later edited and also collated with the very briefJournal of the Convention, published in 1819. His notes werepublished posthumously in 1840. These form a part of the se-ries described in note 1, above. I have maintained Madison’spunctuation and spelling.

5. His speech on June 18 as reported by Madison is inRecords, 1: 282ff. Hamilton said that “he had no scruple in de-claring . . . that the British Govt. was the best in the world. . . .As to the Executive, it seemed to be admitted that no good onecould be established on Republican principle. . . . The Englishmodel was the only good one on this subject.” 288, 299.

6. In order to serve as a check on the popular chamber ofthe national legislature, the second chamber “must have greatpersonal property, it must have the aristocratic spirit; it mustlove to lord it thro’ pride. . . . The aristocratic body, should beas independent & firm as the democratic. . . . To make it inde-pendent, it should be for life.” Ibid., 1: 512. With his usual ad-miration for the British system, in his first speech to the Con-vention, Hamilton opined that the “House of Lords is a mostnoble institution. Having nothing to hope for by a change, andsufficient interest by means of their property, in being faithfulto the National interest, they form a permanent barrier agst.every pernicious innovation.” 1: 288 (June 18).

7. A few delegates favored the idea of somehow gettingrid of the states and consolidating power in the national gov-ernment. George Read of Delaware “disliked the idea of guar-antying territory. It abetted the idea of distinct States wch.

174 n ot e s to pag e s 7 – 1 2


would be a perpetual source of discord. There can be [no]cure for this evil but in doing away with States altogether anduniting them all into [one] great Society.” Records, 1: 202(June 11). He had made a similar proposal a few days earlier,on June 6. 136–37. In his maiden address cited above, Hamil-ton proposed that “the Governor or president of each stateshall be appointed by the General Government and shall havea negative upon the laws about to be passed in the State ofwhich he is Governor or President” 293.

8. Records, 1: 466.9. Records, 1: 492–93. I have put these remarks in the

first person. In Madison’s published notes they are recorded inthe third person.

10. The advocates of equal representation in the Housewere defeated on June 29 by a vote of six states for and fourstates against, with one state (Maryland) divided. Their pro-posal for equality in the Senate was stalemated by a tie vote onJuly 2 (five to five, with Georgia divided) and finally carried onJuly 7 with six voting yes, three voting no, and two states(Massachusetts and Georgia) divided. Records, 1: 549.

11. Article I, Section 9. For an excellent account of theonly full public debate over the slavery issue, see Joseph J.Ellis, Founding Brothers: The Revolutionary Generation (NewYork: Alfred A. Knopf, 2000), 81–119. The debate took placein the House of Representatives in March 1790 in response topetitions from Quakers in New York and Philadelphia “callingfor the federal government to put an immediate end to the Af-rican slave trade.” (81).

12. Article IV, Section 2.13. Article I, Sections 2, 3.14. For a magisterial study of the evolution of American cit-

izenship, see Rogers Smith, Civic Ideals: Conflicting Visions ofCitizenship in U.S. History (New Haven: Yale University Press,1997). On the constitution’s omission of citizenship for women,Native Americans, and African Americans, see 130–34.

15. Article I, Section 3.16. By the same electorate as that for “the most numerous

branch of the state legislature.” (Article I, Section 2).17. Records, 2: 83.

n ot e s to pag e s 1 4 – 19 175


18. For evidence that the Supreme Court sometimes playssuch a role, see my “Decision-Making in a Democracy: TheSupreme Court as a National Policy-Maker,” Journal of PublicLaw 6, no. 2, 279–95.

19. It is only fair to point out that given the political oppo-sition to any increase in federal powers, the Framers may wellhave gone as far as they could go. Their major opponents, theAnti-Federalists, who saw the constitution as a threat to popu-lar government at the state level, objected that the powers ofCongress to regulate interstate commerce were excessive.Richard L. Perry, ed., The Sources of Our Liberties: Documen-tary Origins of Individual Liberties in the United States Con-stitution and Bill of Rights (New York: American Bar Associa-tion, 1959), 240.

20. For an account of the development in the Americancolonies of ideas and practices concerning popular govern-ment, see Edmund S. Morgan, Inventing the People: The Riseof Popular Sovereignty in England and America (New York:W. W. Norton, 1988), esp. Chs. 8 and 9, 174 –233.

21. Gordon S. Wood, The Radicalism of the American Rev-olution (New York: Alfred A. Knopf, 1992), 230.

22. The numbers are a matter of uncertainty. In somecolonies the suffrage may have grown more restricted duringthe colonial period. “What is also unclear is just how manypeople could and did vote. This issue is a source of controversyamong historians, some of whom conclude that colonial Amer-ica was a land of middle class democracy in which 80 or 80percent of all adult white males were enfranchised, whileothers depict a far more oligarchic and exclusive politicalorder. In fact, enfranchisement varied greatly by location.There certainly were communities, particularly newly settledcommunities where land was inexpensive, in which 70 or 80percent of all white men were enfranchised. Yet there werealso locales . . . where the percentages were far lower, closer to40 or 50 percent. Levels of enfranchisement seem to havebeen higher in New England and in the South (especially Vir-ginia and the Carolinas) than they were in the mid-Atlanticcolonies (especially New York, Pennsylvania, and Maryland;not surprisingly, they also tended to be higher in newer settle-

176 n ot e s to pag e s 19 – 2 2


ments than in more developed areas. On the whole, the fran-chise was far more widespread than it was in England; yet asthe revolution approached, the rate of property ownership wasfalling, and the proportion of adult white males who were eli-gible to vote was probably less than 60 percent.” AlexanderKeyssar, The Right to Vote: The Contested History of Democ-racy in the United States (New York: Basic Books, 2000), 7.

23. On the constitution’s exclusion of women, NativeAmericans, and African Americans, see Keyssar, 130–34.

24. Democracy in America, trans. Henry Reeve (New York:Schocken, 1961), 1: lxvii.

25. Although Jefferson and his followers often referred totheir political group as “Republican,” their party seems to havetaken the name “Democratic Republican” as early as 1796, andit retained the name through the election of 1828. In 1820,Monroe ran as a Democratic Republican and Adams as an In-dependent Democratic Republican. In 1824 all four candi-dates—Adams, Jackson, Crawford, and Clay—ran as factionsof the Democratic-Republican party. In 1828, Jackson ran as aDemocratic Republican, Adams as a National-Republican. In1832, Jackson ran as the candidate of the Democratic Partyand Clay as a National-Republican. Congressional Quarterly,Presidential Elections Since 1789, 2nd ed. (Washington, D.C.:Congressional Quarterly, 1979), 19–27.

26. Joyce Appleby, Inheriting the Revolution: The FirstGeneration of Americans (Cambridge, Mass.: Harvard Univer-sity Press, 2000), 65.

27. “When land offices opened on the frontier, land salessoared. In 1800 some 67,000 acres passed into private hands;497,939 acres did so in 1801. By 1815 annual sales hit one anda half million dollars, more than doubling four years later.”Appleby, Inheriting the Revolution, 64. As Gordon Wood re-marks in his review, “Tens of thousands of ordinary folk pulledup stakes in the East and moved westward, occupying moreterritory in a single generation than had been occupied in the150 years of colonial history.” “Early American Get-up-and-Go,” New York Review, June 29, 2000, 50.

28. For a rather unsympathetic portrait, see Stanley Elkinsand Eric McKitrick, The Age of Federalism: The Early Ameri-

n ot e s to pag e s 2 2 – 2 6 177


can Republicanism, 1788–1900 (New York: Oxford UniversityPress, 1993), 706ff.

29. Dubious as one might be—and I am profoundly doubt-ful—about the contemporary relevance of the Second Amend-ment securing “the right of the people to keep and bear Arms,”I have no doubt that contemporaries saw it as important tomaintaining their liberty from a potentially dangerous centralgovernment.

30. Democracy and the Amendments to the Constitution(Lexington, Mass.: Lexington Books, 1978), 166.

31. The Federalist (New York: Modern Library, n.d.), 53ff.32. The often murky maneuvers preceding the compromise

are described in Bernard A. Weisberger, America Afire: Jeffer-son, Adams, and the Revolutionary Election of 1800 (New York:William Morrow, 2000), 258 –77.

33. Elkins and McKitrick, 263 et seq.34. Quoted, ibid., 267.35. Records, 3: 452–55. Italics added. Spelling and punctu-

ation as in the original.36. The Forging of American Federalism: Selected Writings

of James Madison, Saul K. Padover, ed. (New York: HarperTorchbooks, 1953), letter to Thomas Ritchie, 1825, 46.

37. Marvin Meyers, ed., The Mind of the Founder: Sourcesof the Political Thought of James Madison (New York: Bobbs-Merrill, 1973), 520.

38. Ibid., 523, 525, 530.

CHAPTER 3. The Constitution as a Model: An American Illusion

1. In a 1997 survey, 34% strongly agreed and 33% some-what agreed with the statement “The U.S. Constitution is usedas a model by many countries.” Only 18% somewhat orstrongly disagreed. (Nationwide telephone survey of 1,000adult U.S. Citizens conducted for the National ConstitutionCenter, September, 1997.) To the statement “I am proud ofthe U.S. Constitution,” 71% strongly agreed and 18% some-what agreed. In 1999, 85% said the Constitution was a majorreason for America’s success in the twentieth century. (Survey

178 n ot e s to pag e s 2 7 – 4 3


of 1,546 adults for the Pew Research Center by the PrincetonSurvey Research Associates.)

2. Although India gained independence in 1947, adopted ademocratic constitution, and has, except for one interval, main-tained its democratic institutions in the face of extraordinarychallenges of poverty and diversity, I have omitted it from thelist for two reasons. First, continuity was interrupted from 1975to 1977 when the prime minister, Indira Gandhi, staged a coupd’etat, declared a state of emergency, suspended civil rights,and imprisoned thousands of opponents. Second, becauseIndia is one of the poorest countries in the world, comparisonswith the wealthy democratic countries would make little sense.

3. For a summary of the constitutional differences amongtwenty-two older democracies, see Appendix B, Table 2.

4. Plus six half-cantons.5. Robert Hazell and David Sinclair, “The British Consti-

tution: Labour’s Constitutional Revolution,” Annual Review ofPolitical Science, 3 (Palo Alto: Annual Reviews, 2000), 379–400, 393.

6. For example, in the Prussian elections of 1858, 4.8% ofthe inhabitants were entitled to one-third of the seats, 13.4%to another third, and 81.8% to the remaining third. Thusmembers of the wealthiest third in effect possessed 17 times asmany votes as members of the bottom third. Bernhard Vogeland Rainer-Olaf Schultze, “Deutschland,” in Die Wahl DerParlamente, Dolf Sternberger and Bernard Vogel, eds. (Berlin:Walter De Gruyter, 1969), 189– 411, Tabelle A 4, p. 348.

7. Lest you think me biased against Nevada, the RockyMountain states, or small states in general: I have the greatestaffection for Alaska, where I grew up in the days when it wasstill a territory, and for the Rocky Mountain states, where I liketo spend some time every summer. And at just over 3 millionpeople, Connecticut gives me a wholly undeserved voting ad-vantage of nine to one over my sons in California.

8. Alfred Stepan, “Toward a New Comparative Analysis ofDemocracy and Federalism: Demos Constraining and DemosEnabling Federations,” paper for the meeting of the Interna-tional Political Science Association, Seoul, Aug. 17–22, 1997.

9. For a comprehensive description, analysis, and critique

n ot e s to pag e s 4 3 – 50 179


of unequal representation in the Senate, see Francis E. Leeand Bruce I. Oppenheimer, Sizing Up the Senate: The Un-equal Consequences of Unequal Representation (Chicago: Uni-versity Chicago Press, 1999).

10. Stepan, supra n. 8.11. More precisely, a governmental unit of a “State” de-

fined as a territorial system with a government that success-fully upholds a claim to the exclusive regulation of the legiti-mate use of physical force in enforcing its rules within a giventerritorial area.

12. For Mason, see Records, 1: 483; for Madison, see 447– 48.

13. Barry R. Weingast, “Political Stability and Civil War: In-stitutions, Commitment, and American Democracy,” in RobertH. Bates, Avner Greif, Margaret Levi, Jean-Laurent Rosenthal,and Barry R. Weingast, Analytic Narratives (Princeton: Prince-ton University Press, 1988), 148–93, 166, and Table 4.3, 168.

14. Article II, Section 4 provides: “The times, places, andmanner of holding elections for Senators and Representatives,shall be prescribed in each State by the Legislature thereof;but the Congress may at any time by Law make or alter suchregulations, except as to the place of choosing Senators.” Ar-ticle II, Section 1 provides: “Each state shall appoint, in suchmanner as the legislature therefore may direct, a number ofElectors.”

15. Robert Hazell and David Sinclair, “The British Consti-tution: Labour’s Constitutional Revolution,” in Nelson W.Polsby, ed., Annual Review of Political Science, vol. 3 (PaloAlto: Annual Reviews, 2000), 379– 400, 382–85, 391.

16. Lani Guanier, “No Two Seats: The Elusive Quest forPolitical Equality,” Virginia Law Review 77 (1991).

17. Maurice Duverger, Political Parties: Their Organiza-tion and Activity in the Modern State (New York: John Wiley,1954), 217.

18. In an appraisal of Duverger’s propositions in 1958, JohnGrumm observed that “it may be more accurate to concludethat proportional representation is a result rather than a causeof the party system in a given country.” “Theories of Electoral

180 n ot e s to pag e s 50 – 6 2


Systems,” Midwest Journal of Political Science 2 (1958): 357–76, 375.

19. Arend Lijphart, Patterns of Democracy, GovernmentForms and Performance in Thirty-Six Countries (New Haven:Yale University Press, 1999) uses ten variables to distinguish“majoritarian” from “consensus” democracies. Table 14.1, p. 245.G. Bingham Powell, Elections as Instruments of Democracy,Majoritarian and Proportional Visions (New Haven: Yale Uni-versity Press, 2000) uses the terms of his title: “majoritarian”and “proportional.” See pp. 20ff and the classification oftwenty democratic countries on p. 41.

20. Records, 1: 288, 299.21. Madison’s notes, published in 1840 after his death. See

above, note 4.22. The only delegate recorded by Madison as speaking fa-

vorably about the British monarchy was Hamilton. See note 3above. Ironically, the Federalist Papers defending the provi-sions of the Constitution on the executive—Nos. 67–77—were by Hamilton.

23. In what follows I have drawn freely from my PluralistDemocracy in the United States (Chicago: Rand McNally,1967), 85ff.

24. A convenient source for some of the relevant discus-sions is Richard J. Ellis, ed., Founding the American Presi-dency (Lanham, Md.: Rowman and Littlefield, 1999), Ch. 3,63–96.

25. For a critical view, see my “The Myth of the Presiden-tial Mandate,” Political Science Quarterly 105, no. 3 (Fall 1990):355–72.

CHAPTER 4. Electing the President

1. For an excellent and much fuller account of the elec-toral college than brevity permits me here, see Lawrence Lon-gley and Neal R. Peirce, The Electoral College Primer (NewHaven: Yale University Press, 1999). See also Robert M. Hard-away, The Electoral College and the Constitution: The Case forPreserving Federalism (Westport, CT: Praeger, 1994).

n ot e s to pag e s 64 – 7 6 181


2. Records, 2: 501.3. “James Wilson’s Final Summation and Rebuttal,” De-

cember 11, 1787, in The Debate on the Constitution, BernardBailyn, ed. , 2 vols, Vol. 1, 849.

4. Records, 2: 497.5. Records, 2: 522.6. The Federalist (New York: Modern Library, n.d.), 441.

Although the other members of the New York delegation hadwithdrawn from the Convention in June, Hamilton stayed on,though he rarely intervened in the debates and seems to havelittle influence on the outcome.

7. Ibid., 443.8. My interpretation of “anxious.”9. Records, 2: 500. A fear of “cabal” had been frequently

expressed throughout the earlier discussions.10. Article II provided that “The Person having the great-

est Number of [electoral] Votes shall be the President,” and“after the choice of the President, the Person having the great-est Number of Votes of the Electors shall be the Vice Presi-dent.”

11. “How could the framers have made such an elementaryand colossal blunder? No less puzzling, why did the opponentsof the Constitution, who were generally so eager to seize onthe document’s weaknesses, never once expose this weaknessin the mechanics of the electoral college? The short answer isthat neither supporters nor opponents of the Constitution an-ticipated the formation of organized national political parties.”Richard J. Ellis, ed., Founding the American Presidency (Lan-ham, Md.: Rowman and Littlefield, 1999), 114.

12. It also provided that if no candidate received a majorityof electoral votes, the House would choose among the topthree (not five as in the original article).

13. The classic account is by C. Vann Woodward, Reunionand Reaction: The Compromise of 1877 and the End of Recon-struction (Boston: Little, Brown: 1951).

14. Congressional Quarterly, Presidential Elections Since1789, 2nd ed. (Washington, D.C.: Congressional Quarterly,1979), 11.

15. Ellis, 118.

182 n ot e s to pag e s 7 6 – 86


16. The figures may change slightly when the House seatsare reapportioned to conform to the 2000 census figures.

17. Ellis, 118.18. Ibid., 119.19. For a more extended examination of possibilities for re-

form, see Lawrence D. Longley and Alan G. Braun, The Poli-tics of Electoral College Reform, foreword by U.S. SenatorBirch Bayh, 2nd ed. (New Haven, Yale University Press, 1975).

20. The need for a second election could be avoided bymeans of an electoral system various called Instant Run-off,Alternative Vote (AV), or Preferential Voting. “[U]nder AVelectors rank the candidates in the order of their choice bymarking ‘1’ for their favorite candidate, ‘2’ for their second-choice, ‘3’ for their third choice, and so on. . . . [A] candidatewho has won an absolute majority of votes (50% plus one) isimmediately elected. However, if no candidate has an absolutemajority, under AV the candidate with the lowest number ofpreferences is ‘eliminated’ from the count, and his or her . . .second preferences . . . are then assigned to the remainingcandidates in the order as marked on the ballot. This process isrepeated until one candidate has an absolute majority, and isdeclared duly elected.” Andrew Reynolds and Ben Reilly, TheInternational IDEA Handbook of Electoral System Design(Stockholm: International IDEA, 1997), 38. This system isused in Australia for electing members of parliament in single-member districts. Since 1922 an analogous system—the SingleTransferable Vote (STV)—has been used in the Republic ofIreland for the election of members of parliament. However,unlike the presidential elections in the U.S. and parliamentaryelections in Australia, in Ireland the members of parliamentare elected in districts returning three, four, or five members.The STV system produces a high degree of proportionality be-tween the size of a party’s votes and the number of its M.P.s.(Ibid., 85ff ).

21. A Gallup survey in 1968 asked respondents, “Wouldyou approve or disapprove of an amendment to the Constitu-tion which would do away with the Electoral College and basethe election of the President on the total vote throughout thenation?” Eighty-one per cent approved, 12 percent disap-

n ot e s to pag e s 86 – 87 183


proved, and seven percent had no opinion. Longley and Braun,154. In a 1992 survey, respondents were presented with thestatement that “If Perot runs, there is a chance that no presi-dential candidate will get enough electoral votes to win. If thathappens, the Constitution gives the House of Representativesthe power to decide who will be the next President. Do youthink that is a fair way to choose the President, or should theConstitution be changed.” Only 31% felt it was a fair way,while 61% answered that the Constitution should be changed.CBS News-New York Times national telephone survey of1,346 adults, July, 1992. Another survey in 1992 indicated thatAmericans were far from agreed on how their Representativein Congress should vote if none of the candidates received anElectoral College majority. The respondents split their choicesamong the options they were offered. Their Representativeshould vote for the candidate:

With the most popular votes nationwide 29%Who carries your state 16%Who carries your congressional district 14%Who would make the best president 33%Don’t know, no answer 7%

Gallup national telephone survey of 1,006 adults, August 1995.22. Longley and Braun, 154.23. Shlomo Slonin, “The Electoral College at Philadelphia:

The Evolution of an Ad Hoc Congress for the Selection of aPresident,” Journal of American History 73 (June 1986): 35–58, cited in Ellis, 110.

24. Longley and Braun, 169.

CHAPTER 5. How Well Does the Constitutional System Perform?

1. I have drawn feely here from my “Thinking About De-mocratic Constitutions: Conclusions from Democratic Ex-perience,” in Ian Shapiro and Russell Hardin, eds., NOMOSXXXVIII, Political Order (New York: New York UniversityPress, 1996).

184 n ot e s to pag e s 87 – 9 4


2. The survival of basic democratic institutions should notbe confused with the stability or turnover of parliamentary“governments.” Parliamentary systems display a huge variationin the durability of cabinet coalitions—”governments”—rang-ing among our twenty two democratic countries from the rela-tively high turnover of cabinets in Italy to their much higherstability in Norway and Britain. Even in Italy, however, an out-going cabinet may return with essentially the same membersand parties represented in the government.

3. Mathew Soberg Shugart and John M. Carey, Presidentsand Assemblies: Constitutional Design and Electoral Dynamics(Cambridge: Cambridge University Press, 1992), 41.

4. Support for this view and also more skeptical views maybe found in Juan Linz and Arturo Valenzuela, eds., The Failureof Presidential Democracy: Comparative Perspectives, vol. 1(Baltimore: Johns Hopkins University Press, 1994).

5. Shugart and Carey, 42.6. Aili Piano and Arch Puddington, “The 2000 Freedom

House Survey,” Journal of Democracy 12 (January 2001): 87–92.

7. Freedom House, Press Freedom Survey: Press Free-dom World Wide (January 1, 1999).

8. Here I follow the usage in G. Bingham Powell, Jr.,Elections as Instruments of Democracy (New Haven: Yale Uni-versity Press, 2000).

9. An excellent analysis supported by extensive analysisand data is provided by Powell, Elections.

10. See ibid., Chs. 4 and 6.11. Ibid., 129, 130, 197.12. Arend Lijphart, Democracies, Patterns of Majoritarian

and Consensus Government in Twenty-One Countries (NewHaven: Yale University Press, 1984), and Patterns of Democ-racy, Government Forms and Performance in Thirty-Six Coun-tries (New Haven: Yale University Press, 1999).

13. Hans Daalder, “The Netherlands: Opposition in a Seg-mented Society,” in Robert A. Dahl, ed., Political Oppositionsin Western Democracies (New Haven: Yale University Press,1966), 188–236; Arend Lijphart, The Politics of Accommoda-

n ot e s to pag e s 96 – 108 185


tion: Pluralism and Democracy in the Netherlands, 2nd rev.ed. (Berkeley: University of California Press, 1975), 104ff.

14. Leif Leiwin, “Majoritarian and Consensus Democracy:The Swedish Experience,” Scandinavian Political Studies 21,no. 3 (1988): 195–206.

15. Christopher J. Anderson and Christine A. Guillory,“Political Institutions and Satisfaction with Democracy: ACross-National Analysis of Consensus and Majoritarian Sys-tems,” American Political Science Review 91, (March 199):66–81.

16. Ibid., fig. 4, p. 77.17. Ibid., 78.18. David R. Mayhew, Divided We Govern: Party Control,

Lawmaking, and Investigations, 1946–1990 (New Haven: YaleUniversity Press, 1991), 1.

19. Ibid., 76.20. John J. Coleman, “Unified Government, Divided Gov-

ernment, and Party Responsiveness,” American Political Sci-ence Review 93 (December 1999): 821–36.

21. Jeffrey K. Tulis, The Rhetorical Presidency (Princeton:Princeton University Press, 1987), 87ff. Gil Troy, “CandidatesTake the Stump, Then and Now,” letter, New York Times, Jan-uary 17, 1988.

22. I have examined these at greater length in “The Mythof the Presidential Mandate,” Political Science Quarterly 105(Fall 1990): 355–72.

23. Fred I. Greenstein, “The Benevolent Leader: Chil-dren’s Images of Political Authority,” American Political Sci-ence Review 54 (December 1960): 934 – 43. The views ofAmerican school children about the president differ fromthose of British and French school children about their chiefexecutives. “Children and Politics in Britain, France, and theUnited States: Six Examples,” Fred I. Greenstein and SidneyTarrow, Youth and Society 2 (1970): 111–28.

24. Edward Tufte and I undertook such an inquiry in Sizeand Democracy (Stanford: Stanford University Press, 1975)but regrettably the topic seems not to have attracted muchsubsequent investigation.

25. Dahl and Tufte, Size and Democracy, 95ff.

186 n ot e s to pag e s 10 9 – 1 18


26. Though on a Purchasing Power Parity Basis, the differ-ence diminishes to less than four times. U.S. Census Bureau,Statistical Abstract of the United States, The National DataBook, 1999 (Washington, D.C.: U.S. Government Printing Of-fice, 1999), 841, Table 1362.

27. For more details, see Table 5.1, Appendix.28. Lijphart (1999), supra, 301–2.29. Juan Linz and Alfred Stepan provide impressive evi-

dence that by providing more points at which privileged mi-norities can veto the enactment of federal polices, the Ameri-can federal system is the worst performer on social policiesamong all of the OECD countries. A brief preliminary versionof their findings is “Inequality Inducing and Inequality Reduc-ing Federalism: With Special Reference to the ‘Classic Out-lier’—the USA,” paper prepared for the XVIII World Con-gress of the International Political Science Association, August1–5, 2000, Quebec City, Canada.

CHAPTER 6. Why Not a More Democratic Constitution?

1. Michael Schudson, The Good Citizen, A History ofAmerican Civic Life (Cambridge, Mass.: Harvard UniversityPress, 1998), 202.

2. Constitutional Knowledge Survey, National Constitu-tional Center, September, 1997, question 2.

3. Gallup Organization, 1999.4. In the following sections I have drawn freely on my

essay “The Future of Political Equality,” in Keith Dowding,James Hughes, and Helen Margetts, eds., Challenges to Democ-racy (Hampshire, U.K.: Palgrave, 2001).

5. For a full account of failures to provide equal citizenshipamong Americans, see Rogers M. Smith’s masterly work, CivicIdeals: Conflicting Visions of Citizenship in U S. History (NewHaven: Yale University Press, 1997).

6. Probably because he was heavily in debt, he freed onlyfive slaves on his death. Annette Gordon-Reed, Thomas Jeffer-son and Sally Hemings: An American Controversy (Charlottes-ville: University of Virginia Press, 1997), 38. Although his rea-sons for freeing these five are unclear, all were related to his

n ot e s to pag e s 1 2 2 – 1 2 7 187


mistress, Sally Hemings, and two were probably sons by her.Although the issue of paternity is disputed, Gordon-Reed pro-vides strong circumstantial evidence. For her “Summary of theEvidence,” see 210ff and see also Appendix B, “The Memoirsof Madison Hemings,” 245ff. DNA tests provide additional cir-cumstantial, though not conclusive, evidence. See DinitiaSmith and Nicholas Wade, “DNA Test Finds Evidence of Jef-ferson Child by Slave,” New York Times, November 1, 1998.

7. New York: Schocken Books, 1961, Henry Reeve, trans.,vol. 1, p. lxxxi.

8. I have drawn these estimates from Adrian Karatnycky,“The 1999 Freedom House Survey: A Century of Progress,”Journal of Democracy 11, no. 1 (January 2000): 187–200;Robert A. Dahl, Democracy and Its Critics (New Haven: YaleUniversity Press, 1989): Table 17.2, p. 240; and Tatu Van-hanen, The Emergence of Democracy: A Comparative Study of119 States, 1850–1879 (Helsinki: Finnish Academy of Sci-ences and Letters, 1984), Table 22, p. 120.

9. As one example, he writes, that “among the untouch-ables of India there is persuasive evidence that the Hindu doc-trines that would legitimize caste domination are negated,reinterpreted, or ignored. Scheduled castes are much lesslikely than Brahmins to believe that the doctrine of karma ex-plains their present condition; instead they attribute their sta-tus to their poverty and to an original, mythical act of injus-tice.” Domination and the Arts of Resistance (New Haven: YaleUniversity Press, 1990), 117.

10. The late Joseph Hamburger showed that to secure theexpansion of the suffrage (and ultimately the passage of theReform Act of 1832), James Mill, though opposed to violenceas a means, deliberately set out to create a fear of revolutionamong members of the oligarchy. “Since Mill wished toachieve fundamental reforms without violence, it became nec-essary to devise means by which an oligarchy would be led togrant concessions out of self-interest. . . . [T]here were onlytwo alternatives: ‘[The people] can only obtain any consider-able amelioration in their government by resistance, by apply-ing physical force to their rulers, or at least, by threats so likely

188 n ot e s to pag e s 1 2 9 – 131


to be followed by performance, as may frighten their rulersinto compliance.’ Since the use of physical force was to beavoided, Mill built his hopes on the second alternative. . . .Mill was proposing that revolution be threatened. He assumedthat the threat would be sufficient and that it would not benecessary to carry it out.” James Mill and the Art of Revolution(New Haven: Yale University Press, 1963), 23–24.

11. I provide a fuller account in Democracy and Its Critics,84ff. There and elsewhere I have drawn on Stanley I. Benn,“Egalitarianism and the Equal Consideration of Interests,” inJ. R. Pennock and J. W. Chapman, Equality (Nomos IX) (NewYork: Atherton, 1967), 61–78.

12. This assumption is more fully developed in Democracyand Its Critics, 105ff, and restated in briefer form in OnDemocracy (New Haven: Yale University Press, 1998), 74ff.

13. For an excellent analysis, see Amartya Sen, InequalityRexamined (Cambridge, Mass.: Harvard University Press,1992). “Libertarians,” he writes, “must think it important thatpeople should have liberty. Given this, questions would imme-diately arise regarding who, how much, how distributed, howequal. Thus the issue of equality immediately arises as a sup-plement to the assertion of the importance of liberty.” (22)

14. These sentences are a close paraphrase of his state-ments in 1: 298, 304, and 2: 380–81.

CHAPTER 7. Some Reflections on the Prospects for a More Democratic Constitution

1. CBS News /New York Times telephone poll of 1,254adults, May 1987, question 53.

2. See also Amartya Sen, Inequality Reexamined, 36–37and passim. Ronald Dworkin, “What Is Equality? Part 2:Equality of Resources,” Philosophy and Public Affairs 10 (1981).

3. Buckley v. Valeo, 424 U.S. 1 (1976).4. For an extended argument consistent with my brief

comments, see John Hart Ely, Democracy and Distrust: ATheory of Judicial Review (Cambridge, Mass.: Harvard Uni-versity Press, 1980).

n ot e s to pag e s 131 – 15 3 189


APPENDIX A1. The Federalist (New York: Modern Library, n.d.), 59.2. Willi Paul Adams, The First American Constitutions:

Republican Ideology and the Making of State Constitutions inthe Revolutionary Era (Chapel Hill: University of North Car-olina Press, 1980), 106ff.

3. “The Federalist No. 39,” in The Federalist, op. cit., 242ff.4. Montesquieu, De l’Esprit des Lois, Tome I (Paris: Edi-

tions Garnier Fräres, 1961), Bk 2, Ch. 2, p. 12.5. Ibid., Bk. 8, Ch. 16, p. 131. One might expect the rea-

son for this conclusion to lie in the difficulty of assembling thepeople in a large territory. But in direct contradiction to Madi-son’s later argument in Federalist No.10 that the danger of fac-tionalism would be reduced by increasing the size of the polit-ical unit, Montesquieu contended that in a large republic thecommon good would suffer. “In a little (republic), the commongood is better felt, better known, and closer to each citizen.”

6. For further discussion, see my Pluralist Democracy inthe United States (Chicago: Rand McNally, 1967), 34ff.

190 n ot e s to pag e s 16 0 – 16 2


Absolute majority, 56Adams, John, 4–5, 78Adams, John Quincy, 69Adams, Samuel, 4–5Advanced democratic countries,

41–72, 91–119, 164–169African Americans: basic human

rights, 53; civil rights, 125, 147;discrimination, 28; voting rights,16, 28, 128, 130. See also Slavery;Suffrage

Alien and Sedition Acts, 26Amendments, 32–34, 155. See

also Electoral college; specificamendment

American Revolution, 95Anti-slavery measures, 53Argentina, 49, 50Aristocracy, 11–12Article I, Section 3, 144

Article II, 65Article V, 144–145Assumption of intrinsic equality,

131–132Australia, 46, 98Austria, 50

Belgium, 44–45, 98Bicameralism, 45–46Bill of Rights, 27, 30, 32, 51, 143Brazil, 49, 50Britain: bicameralism, 45–46; civil

liberties in, 98; constitution of,65; constitutional system, 12,70–71; “first-past-the-post” system, 58; franchise, 127–128;majoritarian system, 108–109;parliamentary system, 70–71;third party in, 103

Buckley v. Valeo, 151–152

Index

11dahlindex.191_198 11/27/01 4:42 PM Page 191

Bundesrat, 105Burke, Edmund, 47Burr, Aaron, 31, 78

California, 49–50, 81Campaign speeches, 112Canada, 58, 116Castes, reduction of discrimination

against, 130Chief executives, 62–72. See also

Head of state; President; Presi-dential systems

Chile, 135, 137Civil Rights Act of 1964, 128, 130Civil Rights Acts, 130Civil War, 95Cleveland, Grover, 69Committee on Detail, 66, 74Compromises, 12–13Confederal system, 12Congressional power, 19–20; judi-

cial review of actions, 54–55Connecticut, 27Connecticut Compromise, 18, 48Consensual systems, 103–109,

117–118, 146–149, 154Constitution: advanced democratic

countries, 41–72, 91–119,164–169; amendments, 26–28;brevity, 142; commitment toself-government and, 20; com-promises and, 12–13; consen-sual, majoritarian, or neither,146–149, 154; democraticchanges to, 26–31; emergentdemocratic beliefs, 20–26; favorable conditions, 94–96,142; federalism, 143; fundamen-tal questions and, 1–6; futuredemocratic trends, 141–157;hidden cost and uncertainties of

change, 149; historical develop-ments and, 9–10; inequality ofrepresentation in Senate,144–145, 154; judicial review,54–55; limited role of,142–143; as model, 41–72;models for, 9; as national icon,121–123, 155; political practicesand institutions, 29–31; possi-bility of significant change,154–157; practical limitations,9–10; presidential system,143–144; protections in, 51;public discussion of, 155–156;rights, 142–143, 150–152;shortcomings of, 15–20; statesand, 12; structures, 143–149;Supreme Court’s role, 152–155;survey results, 108–109, 141;undemocratic elements in,15–20, 38–39; utility and legiti-macy of, 39. See also Bill ofRights; Constitutional systems;Political equality; specificamendment

Constitutional Convention, 1–8; delegates, 1–2, 4–5, 24;Founding Fathers and, 4–5;preparation for, 4–5; votingcoalitions, 15

Constitutional systems: accounta-bility, 101–102, 115; advanceddemocratic nations, 41–72,91–119, 164–169; British, 12,70–71; consensual systems,103–109, 117–118; defined, 41;democratic effectiveness,116–119, 168–169; democraticfairness, 99–103; diversity and,116; divided government,110–111; electoral arrange-

192 i n d e x


ments, 99–103; electoral systems, 35–36, 55–61, 166; encouraging consensus,103–109; evaluating perform-ance, 91–119; favorable condi-tions, 94–96, 99, 142; federal or unitary, 43–45; historical development, 10; hybrid systems,59, 62, 110–115, 117–119,146–149, 167; maintainingdemocratic stability, 93–97; majoritarian systems, 100–103,117–118, 167; mixed system,62–63; Netherlands, 104–105,106, 107, 108–109; parliamen-tary system, 63, 65; party systems, 61–62; performanceof, 91–119; powers in, 149;presidential system, 62–72, 96,111–115, 143–144; propor-tional vs. majoritarian, 167; proportionality, 100–109; protecting fundamental rights,97–99; relative affluence and, 116; size and, 116; state authority, 51; Sweden, 106–107;Switzerland, 105–106. See alsoConstitution

Constitutionality of laws, 18–19Continental Congress, 124Costa Rica, 43, 62, 98, 116Cuba, 148

Declaration of Independence,4–5, 21, 22–23, 123, 124–125

Declaration of Rights, Virginia, 27,32

Delegates, 1–2, 4–5, 24Democracies, 12; advanced demo-

cratic countries, 41–72,91–119, 164–169; federal

system, 12; judicial review, 55;oldest, 41. See also Constitution;Constitutional systems

Democracy: breakdown of, 95–96,134–135; citizen survey of satis-faction, 108–109; creation of, 6;meaning of, 10; ordinary citizensand, 25; rights, liberties, and opportunities, 136–138; unfold-ing of ideas and institutions, 10;versus republic, 159–162

Democratic Party, 24, 111Democratic-Republican Party,

24, 33Democratic republics, 5–6, 10,

22–26Democratization, result of, 10Denmark, 45, 71, 116Direct popular election: of presi-

dent, 86–88, 155; of senators, 28Discrimination, 28; Civil Rights

Acts, 128, 130; in India, 130. Seealso African Americans

Duverger’s Law, 61

Elections: direct election of presi-dent, 86–88; direct popular,86–88, 155; of 1800, 37, 68,77–78; of 1876, 79–80; elec-toral arrangements, 99–103;Framer’s design for, 16–17; ofpresident, 16–17, 73–89; ofsenators, 28. See also Electoralcollege; Presidential elections

Electoral college, 73–89; alteringor abolishing, 83–88, 179;changes in, 86–89, 145–146,155; choosing electors, 82–83,86–87, 155; constitutionalamendment, 86–88, 155; creation of, 73–76; democratic

i n d e x 193


Electoral college (continued)changes in, 155; direct electionof president, 86–88, 155; districtsystem, 82, 87; election of 1800,31, 68, 77–78; election of 1876,79–80; election of 2000, 31, 68,73, 79–81; entitlement to extraprotection, 84–85; failure of,77–79; Framer’s design, 16; future of, 86–89; improbableoutcome, 30–31; inherent defects, 79–82; lacking majorityof popular votes, 80; losingthough preferred by majority,81; majority of popular votes,80; need for extra protection,85–86; popular vote vs. elec-toral vote, 79–81; reasons for,76–77; reform of, 87–88,145–146; remedial defect,82–83; Senate and, 87–88;small states, 84–86; swingstates, 83; unequal representa-tion of voters, 81–82, 171; winner-take-all-system, 82–83;winning with minority of popu-lar votes, 80–81, 170

Electoral systems, 35–36, 55–61,166

Equal representation in Senate,13–15, 17–18. See alsoUnequal representation

Equal Rights Amendment, 28

Factions, 29–30, 33Federal courts, 18–19Federal Election Campaign Act,

151–152Federal systems, 12, 43–50, 49,

51, 54. See also Constitutionalsystems

Federalism, 33, 44–45, 143Federalist No. 10, 29, 33, 34, 159,

161Federalist No. 39, 160–161Federalist No. 68, 76Federalist Papers, 63–64Federalists and Federalist Party,

24–26, 29, 30Fifteenth Amendment, 28Finland, 63First Amendment, 26“First-past-the-post” system,

56–62, 100, 166. See also Politi-cal parties; Two-party systems

Founding Fathers, 4–5Fourteenth Amendment, 28Framers: awareness of limitations,

142; defined, 4–5; FoundingFathers distinguished, 4–5; limits of opportunities, 11–15;practical limitations on, 9–10;what they couldn’t know, 7–39. See also ConstitutionalConvention

France, 26, 58, 63, 71, 98Franchise. See Elections; Suffrage;

VotingFreedom House, 98French Revolution, 95Fugitive Slave laws, 16Fundamental questions, 1–6,

50–54Fundamental rights, 51, 97–99,

136–138, 153–154, 155

Georgia, 27, 81Germany, 46, 98, 137Gore, Al, 81Government: limits on, 143;

majority rule, 36–37; popularrule, 24–25; proto-republican

194 i n d e x


phase, 21–22; republican, 11;requirements of self govern-ment, 21–22. See also Constitu-tional systems

Greece, 160Gross National Product (GNP), 116

Hamilton, Alexander, 11, 13–15,64, 76, 77

Hayes, Rutherford, 80Head of state, 71, 112. See also

President; Presidential systemsHouse of Commons, 59, 128House of Lords, 17, 45–46, 128House of Representatives, majori-

ties in, 110Hybrid systems, 59, 62, 110–115,

117–119, 146–149, 167

Iceland, 116Incarceration, 117, 169Income inequality, 92Income taxes, 28Independent Commission on the

Voting System, 59India, 130Inequality of representation in

Senate, 144–145, 154. See alsoSenate

Israel, 43, 98, 104Italy, 98

Jackson, Andrew, 24, 69, 112Japan, 43, 71Jay, John, 64, 78Jefferson, Thomas: creation of

political party by, 24; democracyand, 10; election of 1800, 31, 78;foreseeing democratic republic,25; as Founding Father, 4–5;opposed to Federalist policies,

32–33; representative of popu-lar will, 69; role in two-partypolitics, 29–30; slavery and,125; supporter of democratic republic, 23–25

Judicial legislation, 19, 20, 153–155Judicial power, 18–19Judicial review of national legisla-

tion, 54–55

Länder, 43Landowners, 25, 35–36Legislation, judicial review of,

54–55Lijphart, Arend, 103–104,

117–118Lincoln, Abraham, 69

Madison, James, 5, 7–8; choosingchief executive, 17, 66; creationof political party by, 24; demo-cratic revolution and, 31–37;drafting amendments, 27; earlydemocrat, 10; Federalist Papersand, 33, 64; intention to createrepublic versus democracy, 5–6,159–162; judicial power and,19; life of, 31–37; minority interest protection and, 51; opposed to Federalist policies,32–33; political development of,31–37; principle architect ofConstitution, 5; representativeof popular will and, 69; role intwo-party politics, 29–30; Senate representation and,13–15; separation of powersand, 65; small state protectionand, 52; supporter of demo-cratic republic, 23–24; unequalrepresentation and, 84

i n d e x 195


Majoritarian systems, 62, 100–109,117–118, 146–149, 154, 167

Majority: absolute majority, 56; inHouse of Representatives, 110;relative majority, 56, 100; threemajorities, 110. See also Majorityrule

Majority rule, 36–37, 50–54, 101.See also Majoritarian systems

Mandate, presidential, 69–70, 113Mason, George, 19, 27, 32Massachusetts, 82Mayhew, David, 110–111Mill, John Stuart, 60Minorities: balancing rights and

interests in representation,50–54; eliminating discrimina-tion against, 28. See also AfricanAmericans

Minority parties, 100Monarchies, 11, 65, 70–71Monroe, James, 69Morris, Gouverneur, 11, 15, 17, 77Multiparty systems, 57–59, 100,

102. See also Constitutional systems

Native Americans, 16, 128Nebraska, 45Netherlands, 98, 104–105, 106,

107, 108–109New England town meetings,

150–151, 160New Hampshire, 1–2New York, 49–50New Zealand, 127Nineteenth Amendment, 28Norway, 45, 71, 98, 116

Parliamentary Commission, 46Parliamentary systems, 63, 65–68,

70–71, 96

Partisan politics, 78–79Party systems, 61–62, 167. See

also Constitutional systems;Multiparty systems; Politicalparties; Two-party systems

Peaceful democratic revolution,8–10, 20

Plurality, 56, 100Political equality, 123–139;

greater, 128–130; importantdemocratic value, 4; justifiablegoal, 130–132; realistic goal,123–128; strategy for future,156–157; threaten liberty,132–138

Political parties, 29–30, 33–34.See also Constitutional systems;Multiparty systems; Two-partysystems

Poll tax, 28Popular rule, 24–25President: American hybrid system

and, 111–115; campaignspeeches, 112; change in role of,155; creation of, 66–68; electionof, 16–17, 73–89; failure ofFramers’ design, 68–72;Framers’ design for, 64–68; judicial review of actions,54–55; mandate, 69–70, 113;myth of the presidential mandate, 69–70; role of,111–115. See also Electoral college; Presidential elections;Presidential systems

Presidential elections: of 1800, 31,68, 77; of 1876, 79–80; of 2000,31, 68, 73, 79–81

Presidential systems, 62–72, 96,143–144. See also Electoral college; President

Prime ministers, 63, 70–71

196 i n d e x


Proportional systems, 57–59,100–109, 166, 167. See alsoConstitutional systems; Multi-party systems; Political parties

Proto-republican phase, 21–22Prussia, 48Pseudo-democratization, 70

Reconstruction, 53Regulatory powers, 19Relative majority, 56, 100Republican Party, 29Republican phase, 22Republicanism, 160Republics: aristocratic, 5–6; citi-

zen body of, 25; creation of, 5;democratic, 5–6, 10, 22–26;form of government, 11, 20;proto-republican phase, 21–22;republican phase, 22; Roman, 8;selecting chief executive,64–65; Third Republic inFrance, 71; U.S. transition todemocratic republic, 22–26;Venetian, 8; versus democracy,159–162

Revolution, peaceful democratic,8–10, 20

Rhode Island, 1–2Rights: fundamental, 51, 97–99,

136–138, 153–155; minority interests, 50–54

Roosevelt, Franklin D., 69Roosevelt, Theodore, 69Russia, 41, 50

Scotland, 59Second Continental Congress, 124Self-government requirements,

21–22Senate: choosing of senators, 17;

constitutional amendment and,

154; direct elections of senators,28; election of senators, 28;electoral college and, 87–88;equal representation in, 13–15,17–18; inequality of representa-tion in, 144–145, 154; majoritiesin, 110; representation in,13–14; unequal representation,48–50, 144–145, 148

Separation of powers, 65Seventeenth Amendment, 28Sixteenth Amendment, 28Slavery, 13, 15–16, 27–28, 53,

124–125Small states, 52–54, 84–86South Carolina, 82Southern veto, 53Soviet Union, 41Spain, 71States, 43–44; judicial review of

actions, 54; role of, 12; small,52–54, 84–86. See also Federalsystems

Suffrage, 16, 34–36; AfricanAmericans, 28; Britain,127–128; New Zealand, 127;unequal representation and,47–50; universal, 127; women,16, 28, 127, 128. See also Elec-tions; Voting

Supreme Court, 55, 152–155Sweden, 12, 45, 71, 116Swing states, 83Switzerland, 44, 46, 50, 98,

105–106, 116

Taxation, 19, 28, 48Third parties, 83, 100, 103Third Republic (France), 71Thirteenth Amendment, 28Tocqueville, Alexis de, 4, 23,

126–127, 133–135

i n d e x 197


Tories, 30Town meetings, 150–151, 160Twelfth Amendment, 31, 78Twenty-Fourth Amendment, 28Twenty-Sixth Amendment, 28Two-party systems, 29, 56–58,

61–62, 100–103, 110. See alsoMultiparty systems

Unequal representation: balancingrights and interests, 50–54; second chamber, 46–54; Senate,48–50, 144–145, 148; slavery,53. See also Equal representa-tion in Senate

Unitary systems, 12, 43–45United States: bicameralism, 46;

comparison with advanced democratic countries, 41–72,91–119; democratic effective-ness, 168–169; diversity, 116;“first-past-the-post” system,58–59; fundamental rights, 98;hybrid system, 59, 62, 110–115,117–119, 146–149, 167; incar-ceration in, 117; majoritariansystem, 62, 167; monarchy and,71; performance of constitu-

tional system, 91–119; presi-dential system, 62–72; ranking,117, 168–169; size, 116; unequal representation, 48–49

Uruguay, 135, 137

Veto, 69; judicial, 19; Southern, 53Vice president, election of, 31, 78Virginia, 27, 32Virginia Plan, 18–19, 32, 68Virtual representation, 47Vote of confidence, 70Voting, 28, 150; African Americans,

16, 28, 128, 130; age, 28; India,130; poll tax, 28; property taxclasses, 48. See also Elections;Suffrage

Wales, 59Whigs, 30White males: land owners, 25;

voters, 79Wilson, James, 13–15, 52, 54,

73–74, 75–76, 84Wilson, Woodrow, 69, 112Women: discrimination, 28;

suffrage, 16, 28, 127, 128Wyoming, 50, 81

198 i n d e x


Documents

The Castle Lectures in Ethics, Politics, and Economics