Voter knowledge and constitutional change: assessing the New Deal experience.

• Author: Somin, Ilya

INTRODUCTION

In recent years the world has seen a massive wave of constitutional change, most notably in Eastern Europe, Latin America, and South Africa. As this Article goes to press, the United States and its allies are beginning the process of creating a new constitution for Iraq. Here at home, there is an ongoing debate over how best to structure change in our own constitutional system.

Yet our understanding of the processes of constitutional change remains in many ways inadequate. (1) A particularly poorly understood issue is the role of voter knowledge. This Article represents the first effort to determine empirically whether voter knowledge increases during periods of constitutional change, thereby enabling voters to impose heightened constraints on political elites. (2) The answer to this question is important, not only for the empirical study of constitutional change, but also for the ongoing normative debate over how such change should be structured. As discussed more fully below, the problem of voter knowledge has crucial implications for the longstanding debate between those constitutional theorists who claim that Article V of the U.S. Constitution should be the sole legal means of constitutional change and those--now led by Bruce Ackerman and Akhil Amar--who contest this proposition. (3)

In all democratic nations that undergo constitutional change, voters are given a role in constitutional development. For both normative and empirical reasons, it is obviously important to understand how they play that role. Numerous political philosophers have argued that voter control of government is intrinsically valuable. (4) Even many who do not share this view believe that voter control is instrumentally valuable as a check on the power of political elites. (5) Robert Dalai, perhaps the most influential analyst of democracy among modern scholars, asserts that "a key characteristic of a democracy is the continuing responsiveness of the government to the preferences of its citizens." (6)

An essential element of voter participation in constitutional change, as in "normal politics," (7) is that the electorate be adequately informed. (8) Without adequate knowledge, voters cannot monitor and control the actions of their elected representatives. Voter knowledge is perhaps even more important in constitutional decision making than in normal politics. Almost by definition, a change in constitutional structure is difficult to reverse. On the other hand, an ordinary policy failure brought on by voter ignorance can potentially be rectified by voting out the government that sponsored the policy in the next election--a process epitomized by the concept of "retrospective voting." (9) In constitutional politics, by contrast, it is much more important that voters "get it right" the first time around, as another opportunity may not arise for many years, if at all.

Unfortunately, decades of voter-knowledge research has shown that knowledge levels are usually shockingly low. (10) Most citizens lack even basic political information, (11) and close to one-third are so completely ignorant that one leading scholar categorizes them as "know-nothings." (12) Thus, it is essential to know whether this dismal pattern holds true in periods when fundamental constitutional change is on the political agenda.

The issue of voter knowledge and its relationship to constitutional change has implications for two other important strands of scholarly literature: theories of the growth of government and normative theories of constitutional change. Many theories of government growth point to the ability of governments to exploit periods of crisis--especially periods of crisis massive enough to engender constitutional change--to expand their powers beyond what is necessary to resolve the crisis at hand. (13) What these writers fail to explain is why voters allow political leaders to expand their power to such an extent, sometimes even to expand it in ways that do not address the crisis at all. Voter ignorance may be a crucial part of the story, in that measures taken for other reasons may be packaged to ill-informed voters as crisis management strategies.

In recent years, a number of prominent legal scholars, notably Bruce Ackerman and Akhil Amar, have advocated supplementing the cumbersome amendment procedures of Article V of the Constitution with procedures that allow amendment through majoritarian voting processes. (14) Article V requires constitutional amendments to run the forbidding gauntlet of gaining the support of two-thirds majorities in both houses of Congress, followed by ratification by three-quarters of the state legislatures. (15) Throughout American history, critics of Article V have sought to replace or supplement it with a more majoritarian alternative. (16)

Perhaps the most important defense of such proposals is the claim that they represent the considered will of We the People, to cite the titles of Ackerman's books, (17) and the informed Consent of the Governed, to cite the title of one of Amar's influential articles. (18) Ackerman's and Amar's criticisms of Article V are but the latest in a long line of similar attacks dating back to the Anti-Federalists. For example, Patrick Henry argued at the Virginia ratifying convention that Article V's obstruction of the will of the majority would lead to the destruction of American liberty "forever." (19) Henry condemned Article V for enabling "a contemptible minority" to "prevent the good of the majority." (20) This longstanding argument for non-Article V constitutional revision would be strengthened by evidence showing that voters in times of constitutional change actually have sufficient knowledge to make meaningful decisions. Conversely, it would be weakened if majoritarian voting processes during periods of constitutional change are primarily conduits for voter ignorance.

Given the critical role of voter knowledge in the process of constitutional change, it is perhaps surprising that neither the literature on constitutional transitions, nor the equally impressive literature on voter knowledge contains a study of the role of voter knowledge during periods of constitutional upheaval. (21) There is, likewise, no study that attempts to analyze the extent to which voters are able to control the constitutional change process. Neither the advocates of non-Article V constitutional change nor their critics who defend Article V's claim to be the exclusive legitimate mode of constitutional change have even attempted to investigate this important issue. (22)

In this Article, these questions are addressed by examining the behavior of both voters and elites during the New Deal era, the most important and wide-ranging period of constitutional change in American history since the Civil War. This work is the first part of a broader research project investigating the relationship between voter knowledge and constitutional change.

In Part I of the Article, I define the concept of constitutional change used in my analysis and also explain the reasons why voter knowledge in periods of constitutional change should be analyzed separately from voter knowledge in periods of "normal" politics. Part II lays out the opposing hypotheses of the heightened attention and rational ignorance theories. The former predicts that voter knowledge should increase during periods of constitutional change because voters are more likely to pay attention to the unusually important political issues that arise during these times. The latter, by contrast, predicts that voter knowledge should remain relatively constant because the reluctance of voters to invest in the costly process of acquiring substantial political knowledge is a rational result of the insignificance of the individual vote. Part III shows that the New Deal era was an especially important period of constitutional change that provides a critical testing ground for the opposing theories.

In Part IV, I examine survey evidence of voter knowledge during the New Deal period. This evidence strongly suggests that there was no significant increase in voter information levels. Also in Part IV, I briefly address some survey data relevant to the question of elite constraint.

Part V looks at voter knowledge from the perspective of elite political leaders. It examines three critical cases--the National Industrial Recovery Act, the National Labor Relations Act, and the battle over President Roosevelt's plan to "pack" the Supreme Court--to determine if key political leaders perceived the electorate as unusually well informed or whether, by contrast, they saw the constitutional moment as an opportunity for deceptive manipulation of voter ignorance. Sadly, the latter scenario is much more strongly supported by the evidence than the former.

None of the individual pieces of evidence presented in this Article--either quantitative or qualitative--is by itself definitive. Cumulatively, however, they provide strong support for the conclusion that there was little or no increase in voter knowledge or elite constraint during the twentieth century's most important period of constitutional change.

Part VI of the Article discusses some of the implications of this finding for normative theories of constitutional change. Although my results do not refute all possible arguments in favor of non-Article V constitutional change, they do identify serious shortcomings in some of the best-known such claims. Moreover, I argue that the problem of voter ignorance provides an unanticipated positive rationale for the much maligned supermajority requirements of Article V. This point is significant because even prominent scholars who do not fully endorse the arguments of Ackerman and Amar have been troubled by the absence of a positive justification for this critical element of the Constitution's amendment process. Sanford Levinson, for example, admits that he "can think of no good reasons to...