Minnesota v. White

Statements by Justices Scalia, O'Connor, and Ginsburg in the case of Minnesota v. White (2002) show how deeply even the critics of judicial elections assume elections to be democratically sound. Minnesota v. White raised the question of whether Minnesota could limit the speech of judicial candidates by barring candidates from making promises and articulating “views on disputed legal or political issues.”Footnote ¹¹ Minnesota justified these constraints according to the state's interest in maintaining an impartial judiciary. It was thought that an election process in which judges committed themselves to certain courses of action or debated issues likely to come before the Court would undermine the capacity of selected judges to decide according to the facts of the particular case. In White, the Court ruled that these speech codes are unconstitutional.

Justices Scalia, O'Connor, and Ginsburg articulated three different approaches to the problem of judicial speech codes. Scalia argued that judicial speech was protected under the First Amendment. He reinforced his conclusion with an argument about the meaning and function of elections. Scalia argued that speech codes make it impossible for voters to distinguish the candidates from one another on the basis of their political differences. But the point of an election is precisely to allow voters to come to judgments about which candidate can better represent the public on the issues he is likely to engage. So Scalia argued that, in addition to violating the First Amendment, speech codes also undermine the entire point of elections.

Justice O'Connor agreed with Scalia on First Amendment grounds, but she differed with him on the value of the candidates' speech for the judicial selection process. Taking a strong stand on the value of judicial impartiality, O'Connor urged Minnesota to abolish judicial elections. For O'Connor, it is “the very practice of electing judges” that undermines impartiality, and so the problem of partiality was one that Minnesota had “brought upon itself.” O'Connor's insistence on the value of impartiality over and against the value of democratic process amounted to an implicit acceptance of the equation Scalia drew between elections and democracy.

Justice Ruth Bader Ginsburg, by contrast, affirmed the value of democratic representation at stake in the elections, while at the same time resisting Scalia's argument that judicial elections are just like other elections. In this, her response was highly original. Ginsburg sought to “differentiate” between elections for political offices and elections “designed to select those whose office it is to administer justice without respect to persons.” She agreed that the representative role of candidates in legislative and executive elections required them to be “free to inform the electorate of their positions on specific issues.” But she argued that judges are not representatives in the same way as these other candidates. They are rather representatives in expounding the law. As Ginsburg phrased it, “they do not sit as representatives of particular persons, communities, or parties; they serve no faction or constituency.” That function creates reasons for a special structuring of judicial campaigns. She hence argued against resolving the “tension between the ideal character of the judicial office and the real world of electoral politics” by “forcing States to choose one pole or the other.” Rather, out of respect for the “exercise of [the] sovereign prerogatives” of the people, states should be free to promote both judicial integrity (through speech codes) and democratic representation (through judicial elections).

Ginsburg is the only commentator in the judicial elections debate who recognizes the full contours of the delicate institutional problem associated with judicial elections. She treats judges as political representatives, not simply as mouthpieces for the law. Yet while judges are indeed political representatives, they are representatives on matters of legal meaning. This is a unique kind of representational demand. Although Ginsburg was right to attend to the multiple demands on the judicial office, her opinion is wrong to suggest that judicial elections are an adequate way of relating these multiple demands. The problem with judicial elections is not only that they might lead to speech codes; it is that they are poorly-suited to achieve democratic representation on matters of legal meaning, as well. Judicial elections risk sacrificing impartiality, but without the promise of a gain for democracy.

Political Judgment on the Law

Some scholars argue that judges cannot be representatives of the people because their function is only to expound the law.Footnote ¹² Others treat judges as democratic representatives as a matter of fact.Footnote ¹³ Still other work leaves it open whether or how judges could be conceived as democratic representatives, emphasizing the importance of advocacy, inclusiveness, or democratic judgment for achieving “representativity” (in Urbinati's memorable phrasing) among public officials without examining the very different meanings these terms bear when applied to judges as opposed to legislators.Footnote ¹⁴ The scholarly movement towards achieving a general account of representational ethics has worked at odds with the need to achieve an account of what is distinctive in a judges' representational task and how that should bear on the configuration of the judicial office.Footnote ¹⁵ This silence on the question of what it means, exactly, for judges to serve both as political representatives, and as representatives on matters of legal meaning, makes the originality of Ginsburg's approach in White easy to miss. This silence also reinforces the cleavage that understands legal impartiality as the absence of politics, and politics as the absence of impartiality.

The appeal that election proponents make to the value of democratic representation is reasonable. Democratic decisions can be made on matters of legal interpretation, just as on any other complex political question. Those who believe that citizens are intrinsically incapable of making such decisions should consider the democratic credentials of constitutions themselves. If the people are worthy of making decisions about their constitutive political and legal institutions, we should be suspicious of claims that the people are not capable of judging what kinds of interpretive methodologies would best give effect to the promises of the document they themselves have authorized.

Conscientious judges face a vast array of decision-making approaches. In matters of constitutional jurisprudence, originalist, natural law, and representation-reinforcing approaches vie with theories of textualism, structuralism, and fundamental values jurisprudence to create a complex array of principled possibilities for judicial interpretation. Jurists also differ in principled ways over questions like the breadth of the equal protection guarantee, or what kind of behavior is encompassed under speech. Judges whose primary responsibilities are in statutory interpretation or in common law development similarly face a vast array of sources of principled guidance. Should judges consult legislative history, or only the text of statutes? What is the proper role of precedent? Legal theorists are divided over these questions. The idea that such disputes should be decided democratically is no stranger than thinking that the resolutions to such difficulties should be achieved by recourse to the judge's own conscience. As long as the method that the judge uses is principled, fair, neutral to the parties, guided by the use of authoritative legal materials, and reasonable, for political purposes it is appropriate to say that the judge has achieved adherence to the legal values, regardless of the origins of the method she uses or the commitments she has arrived at. The question of which of these many reasonable approaches should be institutionally empowered is a properly political question which, in principle, there is no reason to keep out of the domain of democratic decision-making. Attitudinalism, by emphasizing the significance of political judgments for how judges exercise their discretion, has reinforced democracy's claim here.

But what is the nature of the question that the public should decide? Importantly, the question that it is appropriate to submit to democratic decision-making is a question about how to best construct and elaborate the law. This is a distinctive question, different from the question that voters appropriately ask themselves when they select legislative candidates. The distinctiveness of the question arises from the constitutive role that legal impartiality plays for the judicial office. Legal impartiality structures the way that it is appropriate for judges' political commitments to be expressed. Whereas it is appropriate for legislators to advance the interests of their own political parties, trade votes, or make promises regarding their treatment of particular cases, this behavior is not appropriate for judges. Hence, scholars who argue that a “judge depends on the will of the sovereign (the law) but should not depend on the opinion of the sovereign” have failed to recognize that the opinion of the people can theoretically shape the contours of the discretion that judges face without circumscribing the judges' right relationship to the law.Footnote ¹⁶ This can only happen, however, if the question the people are answering is a question about interpretive approach rather than particular outcomes.

While some scholars resist conceiving of judges as democratic representatives, others, looking at the realist literature, believe that judges are representatives in exactly the same way that legislators are. These scholars would not agree that political ideology is differently relevant for judges than it is for legislators. Although not even the most hard-core realist defends judicial vote-trading, too few scholars appreciate the distinctiveness of the judicial role that that constraint implies, or the implications of this distinctiveness for judges' representational task.

Legal realism emphasizes the basic congruity between the task of the legislator and that of the judge. In both contexts, official behavior is determined by attitudes, strategy, and political judgment. Realists use Bush v. Gore as a strong demonstration of the failure of legal rules to bind—as Segal and Spaeth argue, a “prime example” of attitudinal decision-making.Footnote ¹⁷ Even scholars who resist many core conclusions of attitudinalism still offer Bush v. Gore as an example of partisan decision-making oriented towards advancing the fortunes of each justice's favored party.Footnote ¹⁸ Given this strong consensus, the case is a good place to excavate how legal elements of judicial decision-making can distinguish judicial decisions from legislative ones, even as both judicial and legislative decisions are properly understood to be “political.”

First, a word on legal formalism. Legalism is sometimes understood to mean either that legal rules can fully determine legal outcomes or that there is a “right answer” to each legal dispute. This conception of legalism equates it with the “mechanical jurisprudence” so despised by legal realists. Although realists were right to resist the idea that law could determine all outcomes without reference to contestable political judgments, their characterization of the plastic meaning of law went too far. We can instead understand legalism in terms of the capacity of rules to generate coherent legal reasons, not unique case outcomes. This less demanding test makes it more apparent how judicial discretion can be consistent with legal decision-making.Footnote ¹⁹ It also allows us to invoke a minimalist conception of legal formalism to block nonsensical claims—for example, that precedent is entirely meaningless. Legal rules may generate decisive reasons in some clear cases; in hard cases, legalism's accomplishment rests in the development of coherent reasons over a series of cases based on authoritative legal materials. Let us call this a minimal conception of legal formalism. Minimal legal formalism is what I mean when I discuss legalism, and it is what structures the distinctiveness of the judges' representational task.Footnote ²⁰

Consider Bush v. Gore. In that case, conservative jurists transparently adopted a liberal reading of equal protection to advance the Republican candidate, and liberal jurists adopted a conservative reading of equal protection to advance the Democrat. Bush v. Gore has two especially notable features. First, the decision was based, in part, on a new rule about equal protection. That rule stated that “having once granted the right to vote on equal terms, the state may not, by later arbitrary and disparate treatment, value one person's vote over that of another.” Whether we understand the rule broadly or narrowly, the announcement of this rule was a political choice with political consequence. It can be positioned on an ideological scale in comparison with other possible rules on equal protection. Other rules governing the relationship between equal protection and voting are imaginable; the adoption of those rules would have led to a different outcome.

Bush v. Gore articulated a rule for equal protection that was blatantly political. It was political first in the sense that it did not follow as a matter of obvious fact from the text of the Constitution itself or from the Court's own precedent. Furthermore, it was political in the sense that the justices' motives were transparently oriented towards a political result. Finally, it was political in that each of the justices, by reversing his or her own prior legal commitments, displayed remarkable failures of integrity. Do these political features make the case not law? Not necessarily. Many badly-reasoned or ill-motivated cases are still authoritative law. Many cases in which justices reverse their own previously articulated standards, when confronted by the implications of those rules in new circumstances, still have the force of law. Significant political commitment is consistent with the generation of law.

The element of Bush v. Gore that could make the case “not law” lies in its second notable feature, the limiting language of the majority opinion. That opinion stated that the Court's consideration was “limited to the present circumstances,” ambiguously opening the possibility that the opinion was intended to apply only to that single case. This would mean that the rule announced in Bush cannot resolve the disposition of any other case. We need no recourse to metaphysical abstractions to demonstrate that, according to such an interpretation, the case is not “legal.” That interpretation would turn Bush v. Gore into decision-making by a single decree; a literal failure to generate reasons on whose basis future cases can be decided; a literal failure to generate binding law.

What will determine whether Bush v. Gore is or is not law is whether its core ruling (however narrow that ruling may prove to be) will be extended to other cases. Since 2000, new cases have emerged under the standards articulated in Bush v. Gore. One judge has already argued that, because of Bush v. Gore's limiting language, circuit courts should not consider themselves bound by its reasons—in other words, that Bush v. Gore literally is not law.Footnote ²¹

No matter which way the Court decides these new cases, its decision will be political. The Court will be deciding between competing and contested conceptions of equal protection. But if it treats Bush v. Gore as authoritative precedent even when the ACLU takes up its mantle, the Court will be acting not only to advance a political ideology of equal protection, but it will also be acting with legal impartiality. This impartiality is necessary if we are to claim that Bush v. Gore was law at all. This impartiality is hence a constitutive feature of law. In other words, impartiality, expressed through the consistent application of even a particular political ideology, is literally constitutive of the legal function as such.

In fact, the reason that legalism is consistent with many (although not all) political ideologies is that naming a political ideology as ideology already implies a measure of consistency in application. Political ideology is not randomness, nor is it nepotism or favoritism. The very concept of “ideology” implies a measure of structured consistency. Depending on the nature of the structured consistency, some political ideologies are perfectly consistent with legal impartiality.Footnote ²² Constitutional scholars know that some longstanding regime orders have generated a series of judicial decisions which are both ideologically saturated and legally defensible.Footnote ²³ But not all ideologies are legally defensible. Bush v. Gore shows us that political ideologies that revolve solely around advancing the interest of one party, whatever those interests consist of, can never be consistent with legalism because of the minimal role of legal formalism—the minimal role of impartiality—in making judicial decisions legal, instead of decision by fiat.

The significance of impartiality in constituting legalism has profound implications for the nature of public choice in selecting judges. While the open-ended possibilities of judicial political ideology mean that democratic selection of judges is highly appropriate, the significance of impartiality in constituting an ideology as legal ideology means that, when selecting judges, the nature of the question that citizens answer should be appropriately informed by the value of legal impartiality. I name this the standard of “democratic processualism.” It refers to the match between the form of reasons that are elicited in a particular institutional environment and the form of reasons that are appropriate for the particular choice being made. I will return later to discuss the concept of democratic processualism in greater detail.

Elections and Democratic Representation

I have noted that the relationship between political ideology and legalism makes it reasonable to believe that judges should be selected through democratically responsive procedures. It is reasonable to believe that citizens should have a say in questions such as the proper construction of equal protection. Given the success of scholars in using partisanship to predict judicial decisions, should this democratic voice be exercised through an electoral mechanism?

Elections are only one of a variety of democratic procedures, a set that also includes lotteries, juries, appointments, and plebiscites. All of these procedures have serious deficiencies. It is critical work for theorists of institutional design to specify the particular democratic deficiencies of these various procedures. Such specification allows for more thoughtfulness in constitutional design, and hopefully, a better fit between procedures and the nature of the political problems those procedures are meant to address.

Although elections are one paradigmatic tool of democratic authorization, troubling deficiencies of bare elections include low turnout and low levels of voter interest;Footnote ²⁴ the “minimalism” of voter cognition, which is characterized by low levels of knowledge, low levels of attention, and little mastery of political concepts; and the incoherence and temporal instability of public attitudes.Footnote ²⁵ These problems sometimes make it seem surprising that elections could provide any portrait of the people's will at all.Footnote ²⁶ It is these concerns that drive some normative constitutional theorists towards the defense of non-electoral governance mechanisms, such as independent judiciaries, on democratic grounds.Footnote ²⁷

But not all elections are equally authoritative (or non-authoritative). Some elections provide clearer indications of what the people want than others—a more reliable, although always partial, portrait of the people's will. These elections are ones characterized by high levels of voter involvement and an unusually clear political option between competing agendas, a clarity that makes the informational burdens that voters face less significant and that makes it easier for voters to register a discernable intent.Footnote ²⁸ In other words, bare electoral environments can be rendered more authoritative by the provision of motivational and deliberative goods that help to close the gap between the preferences of the people and the content of policies that are actually enacted. These goods are mostly supplied by parties. And it is because of the critical role that parties play in supplying motivational goods—rallies, party conventions, bright colors and waving flags—that nonpartisan and noncontestatory judicial elections are subject to an array of devastating democratic criticisms.Footnote ²⁹

The strongest democratic defense for judicial elections is for partisan judicial elections.Footnote ³⁰ But partisan judicial elections face democratic problems of their own. Although partisan cleavages often give voters the cues they need to make accurate choices, partisan cleavages are not structured—now or in the foreseeable future—according to the minimal requirements of legal formalism. Hence partisan cleavages fail to provide citizens in a low-information environment with the right kind of information they need to avoid mistakes. (Here, “mistake” means that the voter chooses otherwise than she would if she had full information). Understanding this failure requires taking a step back to explore the complex linkages that a political cue must make to enable voter choice.

The Deliberative Contributions of Political Parties

Political cues are critical for forming the deliberative environment within which voters receive the information necessary to connect their own judgments to the choices the election offers. But partisan cues in judicial elections fail to provide two critical elements of such an appropriate deliberative environment.

The first element of an appropriately deliberative political environment is the creation of appropriate structured political cleavages that enable voters to exercise meaningful political choice. Voters are not able to express meaningful choice on their own. The ballot offers few options, and voters cannot elaborate to communicate why they made their choice. Before the election, voters are absolutely reliant upon others to construct meaningful political messages that they can either rebut or endorse; after the election, voters are again reliant on others to interpret the meaning of what they have done. Parties, through campaigns, create the messages that allow for such interpretation.Footnote ³¹

Parties and candidates that fail to construct messages that are meaningful to voters are frequently portrayed as failing to offer an “option.” The criticism is revealing. As long as the election is contested, this criticism cannot be literally true, since the bare existence of an opposing candidate means that voters have an option of some kind. But when a voter chooses a candidate, what has he selected? Has he endorsed her promise to block certain judicial nominees? Or perhaps her support of a certain defense policy? Or has he simply approved her capacities in managing a budget?

The criticism that parties have failed to construct a message only makes sense as a criticism when we understand the profundity of voter's reliance on others to connect political responses into positions that can be communicated through a vote. Voters cannot provide their “own” messages. Voters with idiosyncratic interpretations of politics are simply unable to communicate that meaning, as voters, at all. Once voters have responded to the messages of others, they are again absolutely reliant on others to interpret their responses—interpretive work that is much easier, as voters intuit, when the campaigns have done a good job of offering well-organized agendas in the first place. If elites fail to properly interpret the cues that voters have sent them, the resulting communication breakdown can be experienced as alienating. The breakdown is that of democratic responsiveness. Politicians have failed to “get the message,” and voters, complaining now as citizens, find themselves unable to exercise the political choice that is rightfully theirs. Such citizens must take refuge in other forms of political action, forms that allow for the expression of complex messages. If possible, they must transform the agendas that are offered by elites into ones that they, as voters, can use to express the content of their judgments in a form that is publicly authoritative.

None of this activity can be captured by comparing the political positions that emerge from this process within an ideological scale. Whether or not partisan debate around an issue actually revolves around the cleavages that people care about cannot be resolved by reference to scholarship on where officials are positioned on such scales. In academic work, partisan cues are labels pointing to a relational positioning in an outer political reality. We can measure the strength of partisanship as a predictor and thereby measure the accuracy of partisanship as a cue. But in a campaign environment, partisan cues are active agents in organizing politics. Their effectiveness as cues is not a signal to be evaluated but is rather an accomplishment to be maintained in the face of centrifugal forces.

Because of this significant organizing work of partisan cues, party efforts to achieve “a common focus, to make one question and one cleavage paramount in voters' minds,” are critical.Footnote ³² This dynamic may be crucial for ensuring responsible governance, as well; Keith Poole and Howard Rosenthal note that in government, it is parties who “help to map complex issues … into a low-dimensional space.”Footnote ³³ Nancy Rosenblum labels partisan positions as comprising “the principal terms of justification in electoral political life.”Footnote ³⁴ They are indeed among some of the only terms of justification that can allow large numbers of people acting as voters to express significant choice together. In fact, the failure of a political system to structure voter choices represents a democratic breakdown no matter where it occurs. Disengagement from Vietnam was almost impossible while the parties failed to distinguish themselves on the issue. Term limits have overwhelming support from voters, but the parties will not pick the issue up, and as a result, voter choice in this dimension is disabled. When parties refuse to distinguish themselves on an issue that matters to voters, voters are unable to exercise political choice on that issue. If our politics regularly failed to achieve such focus, it would be appropriate to mount a democratic criticism of our entire political system, no matter how high the turnout or how fairly votes were counted. And if a certain issue is systematically likely to be subsumed by other choices, as I will show happens within partisan judicial elections, this is demonstration that voters will systematically be unable to make a choice along that dimension—that voter choice is impaired.

Beyond the creation of a structured cleavage, the second critical element of a deliberative environment is accurate information—that the cues political elites offer point accurately. Voters with little information about the candidates—that is to say, voters in non-ideal contexts—have difficulty registering the preferences they would have were they more fully informed. This problem creates the possibility of “mistakes”—a gap between what the voter expresses and what her actual preference would be if she had more information. Cues—endorsements, poll results, visual symbolism, and partisanship—can be partial compensatory mechanisms for this problem. All cues are, of course, shortcuts, and as such may be unable to completely close the gap between actual preferences and registered choice.Footnote ³⁵ But certain cues are surprisingly effective at allowing voters to come to judgment as if those voters were using very detailed political information.Footnote ³⁶ Partisan cues are normally especially effective—not surprising given the incredible work that political elites put into maintaining the meaning of a partisan cue.Footnote ³⁷

Yet cues can fail to point effectively because of faulty linkages between cue and content. An information shortcut relies on conceptual linkages between the signal (the cue) and political content. If those linkages are biased, information will be linked together in faulty ways, possibly creating a systemic distortion in the link between voter preferences and actual electoral choice.Footnote ³⁸ For example, a voter who uses race as a cue for competence may register votes that are systematically self-defeating from the perspective of that voter's own political values.

Because of the possibility of self-defeating cues, some research suggests that the use of heuristics is most valuable, not for the politically naïve, but for the politically sophisticated. This is because political sophistication helps voters understand the structure of the political world. Such knowledge, generally speaking, allows them to use cues to make accurate inferences—but can equally work against them “if the world (i.e., a political campaign) is not structured in the typical manner.”Footnote ³⁹ This may happen when the voter is trying to decode which candidate is closer to their preferences when one candidate takes a far more complex range of ideological views than the other. In this situation, a heuristic actually makes sophisticated voters more likely to make mistakes. The cue is only helpful if it points in the right direction; but if voters are quite unsophisticated, or if they are sophisticated but face political choices that are quite complex, it becomes increasingly unlikely that the cue will point true. Partisan cues in judicial elections face this problem. They point to a complex political environment and so face special barriers in carrying forward a popular political judgment.

The Democratic Failure of Partisan Judicial Elections

Although political scientists have firmly established the value of these deliberative goods for non-ideal democratic politics, nobody has taken full account of the significance of their findings for the design of judicial selection mechanisms. The motivational and deliberative goods that parties supply render legislative electoral decisions more democratic than they would otherwise be. But linkages between the cleavages of party cues and the relevant cleavages of legal interpretation are unlikely to be robust and effective. This means that when parties structure the judicial campaign environment, their efforts to engage voters do not lead to the production of the deliberative goods that the public needs. In this context, partisan structuring of the campaign environment distorts, rather than enhances, the efforts of voters to register their political preferences.

The first, most straightforward problem is that the cues voters use to make reliable low-information judgments in other contexts may simply be less reliable here. Cues in judicial elections may not fulfill the first criteria of democratic utility. For example, while past positions may be an important source of information for voters deciding on a congressional candidate, what weight should voters give to briefs that lawyers submit on behalf of clients? The groups that judges may belong to are also often esoteric. What does it mean to belong to the Federalist Society?

Voters who use partisan identification as their cue may believe they are using a relatively high-quality cue. In general, partisanship cues are highly reliable and allow voters to register choices that are interpretable by the broader political community. But structuring political judgments on matters of legal interpretation according to the logic of partisan cleavages is, in some cases, impossible; in other cases, it renders those judgments self-defeating.

Sometimes partisanship is not a valuable cue simply because the parties do not have positions on the issues the judges face. For the first case, consider whether due process imposes limits on the extent of punitive damages a court can assign. It is unclear what the positions of the political parties on this issue would be; and the Supreme Court justices' opinions here cannot be arranged on a left-right continuum (while O'Connor and Stevens have voted to affirm the idea that due process limits the punitive damages juries can award, Ginsburg, together with Rehnquist, resists that conclusion).Footnote ⁴⁰ For this issue, partisanship is not a valuable cue.

For other cases, parties may have positions on the issues the judges face, but those positions do not aggregate according to the structured cleavages of minimal legal formalism. Remember that the existence of minimal legal formalism means that judgments on the meaning of the law aggregate together in distinctive ways. A justice committed to a broad interpretation of Congress's commerce power might be committed to allowing Congress to regulate intrastate marijuana sales as well as passing environmental protection. Justice Stevens apparently find himself in just such a position, voting in Gonzalez v. Raich to uphold congressional regulations of marijuana even though, as an electoral voter, he would favor de-criminalization. The complexity of the question is reflected in the votes of Gonzalez, which are ideologically splintered, with Stevens joined by Scalia, and facing dissent by O'Connor, Thomas, and Rehnquist. The template of this reasoning does not follow the ordinary template of how political issues are bundled together, where “liberals” would want Congress to fashion environmental protection laws, but without being able to regulate marijuana sales; and “conservatives” would want the opposite. The problem is not that there is no “liberal” or “conservative” position. Attitudinalist research clearly shows that these issues do manifest liberal and conservative dimensions. The problem is that the politically liberal position in the legislature does not match the politically liberal interpretive position that is constituted by reference to legalism. The consistency that is characteristic of Steven's political/legal position is not consistent with the structure of the partisan political register as communicated through campaigns.Footnote ⁴¹

This point is especially meaningful when we consider the significance of communicative transparency for partisan success. Political elites who want success must simplify the political environment so that relatively unsophisticated voters can still make accurate judgments about how to translate their preferences into votes. I name this the “imperative of communicative transparency”—the imperative that the behavior of political elites be transparent to voters. This political imperative—that decision-making happen within a context that is fairly easily communicable to voters in a low-information environment—structures a great deal of legislative strategy. A lot of work goes into making sure that partisanship can be a cue that easily communicates in a low-information environment. For example, Poole and Rosenthal argue that vote trading will happen within a constrained political space: they write that “although a Kerry-Helms trade is possible, such a trade would be highly unexpected and therefore highly publicized. Even if the trade were beneficial to constituents, the constituents might not process the relatively complex information correctly. If constituents are mainly sensitive to consistent voting patterns along the predictive dimension, Kerry and Helms may find such trades ill-advised.”Footnote ⁴² Although Poole and Rosenthal do not name it as such, the logic they are pointing towards is that of the imperative of communicative transparency. The desire to communicate with constituents in transparent ways (such that voters “process the relatively complex information correctly”) leads legislators to avoid vote trading in ways that do not fit the cognitive map that political parties have established. Poole and Rosenthal suggest that this imperative may lie behind the consistent phenomenon of attitudinal unipolarity in the legislature: “if legislators, perhaps as a result of being concerned about establishing a reputation for consistency, seek to sustain a pattern of unidimensional voting, vote trading may allow observations of roll call votes to appear as if they were preferences mapped onto an underlying dimension even when true preferences have a far more complex pattern.”Footnote ⁴³

Poole and Rosenthal suggest that one point of this unipolarity is to foster communicative transparency—to collapse complexity so that voters can make accurate judgments based on cues. But this unipolarity cannot be expected to incorporate the complexity of the principled consistency of minimal legal formalism. Legal impartiality may lead to results which appear chaotic when mapped onto the political space of legislators, violating the imperative of communicative transparency.

Let us examine the consequences of the tension between impartiality and communicative transparency by playing out a specific example: free speech and abortion. A liberal, according to attitudinalists, would support the unrestricted right to choose abortion. On matters of free speech, a liberal would support “those alleging deprivation of First Amendment freedoms.”Footnote ⁴⁴ This is not only the attitudinal conception of a liberal; it is also the popular conception—liberals favor free speech and abortion rights. Supporting free speech and abortion rights is hence communicatively transparent (communicatively transparent only as a result of the work of partisan elites in bundling those two issues together). When the Supreme Court allows for restrictions on protests outside abortion clinics (and thereby creates the legal basis for gated, concealed pens, surrounded by armed guards, at political party conventions ironically designated as ‘free speech zones’), has the Court rendered a decision that is “liberal” or “conservative”?Footnote ⁴⁵ This question, crucial to locating a politician within a sensible political map, does not begin to make sense of the question posed by the dilemma the judges face.

On this question of abortion and free speech, different judges will draw the line at different places. Their choices are political in that there are many reasonable positions—many responsible positions—for the line to be drawn. Yet a commitment to legalism means that any line drawn must be a principled one. A commitment to legalism means the judge must achieve impartiality, or minimal legal formalism. For a judge to act out her “liberalism” in a manner that is communicatively transparent according to the broader political register—by allowing for restrictions on abortion protestors, but not on political convention protestors—would be for her to fail to make law.

If judges decided to allow the imperative of communicative transparency to trump the imperative of legal impartiality, they would fail to make law. Bush v. Gore is communicatively transparent in that the Republican appointees voted to advance the career of a Republican candidate. If the justices apply the demanding standards of equal protection from Bush v. Gore even in cases where the ACLU stands to benefit, the communicative transparency of the case will be destroyed just as surely as if Helms and Kerry were to join forces. But the judges would still be advancing their own “policy preference” for a certain equal protection rule. However, if the justices fail to impartially advance the equal protection policy they have generated in Bush, they not only will fail to advance their own “policy preference”; they will also have failed to make law. Note that in this case, impartiality is consistent both with the advancement of a particular political ideology, and with the creation of law. It is the maintenance of partisan communicative transparency that is inconsistent both with advancing a political ideology and with creating law. This cleavage is nothing like “law versus politics.” It is rather a matter of both law and politics being threatened by the imperative of communicative transparency, which is the first imperative of partisan cues.

The reason that the parties' communicative/attitudinal unipolarity cannot be expected to incorporate the complexity of minimal legal formalism is not because it would be impossible for it to do so. It is rather because maintaining the principled consistency of minimal legal formalism cannot be expected to be a high priority for parties, given the centrifugal forces that threaten the coherence of partisan cues at all. Remember that the transparency of a party cue is an accomplishment. It requires work to sustain.

Yet, if party cues fail to track the structured cleavages of legalism, then the cues will also fail to correspond to the proper question, which is a question about which approaches to the law should be democratically empowered. Because political parties' dominant concern is to support their own capacity to win elections, and because the large political register is only barely concerned with legal interpretive issues, there are good reasons to believe that the packaging that is a democratically necessary element of a campaign risks systematically misleading voters about the dimensions of choice that they face when they vote for a judicial candidate. This problem has been approached before from the perspective of concern for the legal values; many scholars have written about the possible bad legal consequences of judicial elections. But the mismatch poses an underappreciated political and democratic problem as well; voters who want to select judges on the basis of appropriate legal/political positions will find themselves without the cues they need to do so.

How can it be that partisan cues would be misleading for voters when they are so revealing for scholars of judicial choice? The answer, as we shall see, is in the context within which partisan cues appear in each case. The ideology scores that we can give judges position them relative to each other; but they do not position them within a political space that is common to legislators or others operating within the same partisan structure. Partisan elections work, on the other hand, to portray judges—and hence actively position them—within this common political space.Footnote ⁴⁶ This argument is emphatically not about the intrinsic qualities of law coming into conflict with the intrinsic elements of politics. The point is about contingent political reality. It is that the structure of choice necessary for voters to communicate decisions on matters of legal meaning (one constituted by reference to the minimal requirements of legal impartiality) corresponds poorly to the structure of choice useful for communicating in an environment structured by parties (a structure constituted by reference to the imperative for communicative transparency). There are good reasons to believe that, given the political context voters already operate in, judicial elections confront voters' efforts to register a sensible choice with enormous difficulties. Voters in such an environment are unable to register a relevant choice, just as they are unable to register a choice when the parties fail to differentiate themselves on Vietnam or term limits. In both cases, the partisan failure to appropriately structure campaign cleavages leaves voters unable to exercise their political agency, as voters, on the questions relevant to that election.Footnote ⁴⁷

I emphasize the contingency of this argument. It is logically conceivable that major party divisions could occur along jurisprudential lines. Such a cleavage would render operative democratic choice on matters of legal interpretation, and in so doing, it would address the concern that voters be able to connect their judgments on the law to an electoral decision.

Democratic Processualism

My argument departs from received ways of thinking about the democratic authority of elections. Deliberative democracy theory supports my argument because it insists that legitimate outcomes must be sensitive to the right kinds of reasons. But what is problematic about judicial elections also shines light on an undertheorized element of deliberative democracy itself.

A classical cleavage in democratic theory concerns the divide between procedure and substance.Footnote ⁴⁸ Proceduralism contends that democratic authority emerges from the collective authorization of the people through a set of fair procedures. Hence proceduralists emphasize the significance of fair procedures, equal opportunity to affect the democratic process, and of turnout.Footnote ⁴⁹ Substantive conceptions of democratic authority also claim that the content of decisions is fair ground for democratic criticism. A law that impairs the right of citizens to free speech is a law that violates core democratic ideals so profoundly that it is subject to criticism on democratic grounds, even if it was fairly decided with full participation. Some use a substantive account of democracy to justify the institution of judicial review, which renders decisions through procedures that are non-accountable to the electorate.Footnote ⁵⁰ Others develop the substantive core of democracy to defend rights such as a right to privacy.Footnote ⁵¹

Some theorists of democracy are remarkable for their effort to overcome the cleavage between procedure and substance by showing how the core values of democracy contain both procedural and substantive elements and by arguing for democratic scrutiny both of procedures and of the content of political proposals.Footnote ⁵² Although different theorists formulate the demand differently, a core ideal of deliberative democracy is that public decisions and procedures be justifiable according to reasons that take each citizen as a full, equal, and free member of the political community. Reconciling the substantive and procedural expressions of citizens' free and equal status provides rich terrain for scholarship.

But significantly missing in the academic conversation about reason-giving is a developed sense of institutional context, with limited academic discussion over the value of parties from a deliberative perspective.Footnote ⁵³ Very little theory actually examines the institutional settings that elicit reason-giving, or the kinds of reason-giving characteristic of different settings.Footnote ⁵⁴ The most important debate that engages this question, that between Ronald Dworkin and Jeremy Waldron about the extent to which courts inspire public debate to “focus on” rights, has not labeled the criteria explicitly or made it useful for work beyond that debate.Footnote ⁵⁵ Furthermore, no work in deliberative democracy interrogates whether or not democracy might demand that the reasons citizens offer for their decisions be evaluated according to criteria that are themselves institutionally specific. Deliberative democrats have not adequately interrogated whether it is appropriate to ask, not only that public policies be justifiable according to certain conceptions of citizens' free and equal status, but also whether democracy demands that institutional settings be designed so as to maximize the likelihood that public decisions will be made with reference to the reasons that are the most critically relevant for those decisions.

Such a concern cannot be accommodated within the language of “procedure versus substance.” Nor can it be accommodated with the language of “justification according to a conception of free and equal citizenship.” The argument is not that partisan structuring of judicial elections simply fails to support democratic process. Quite the contrary: parties are the primary mechanism for diffusing political power beyond the narrow circles of backroom political elites. To the extent that elections are inclusive and egalitarian in the real world, it is because of parties. But neither is the argument that partisan judicial elections risk arriving at the wrong outcome, perhaps selecting the racist authoritarian judge over the egalitarian democrat judge, such that the voters' bad decision should be overridden through some alternate procedure. The democratic concern is more appropriately captured as an institutional concern about the relationship between formal procedures, informal processes, and deliberative rationality. The insight is that procedures, and the way procedures are handled in the settings of non-ideal politics, affect the way that people structure the reasons they rely upon to make decisions. Different procedures differentially affect this process.

Democratic processualism insists that public policies should be decided on the basis of reasons that are appropriately sensitive to the critical features of those policies. Hence if a democratic public were to make a decision about war without considering the question of national security, it would be appropriate to criticize that public on the grounds of democratic processualism. In this case, a democratic public's failure to make a decision about judicial selection that is sensitive to the particularities of the judicial function is also a failure of democratic processualism.

Let us imagine a reasonable, fair-minded, and democratically-minded voter. In the partisan case, this voter is faced with a choice between the interests of business versus individuals, employer versus employee, and defendants versus the state; this is largely how partisan conflict is structured today. Given such a choice, our voter might believe that, given the particular political circumstances she lives in, the best conception of democratic equality requires advancing the interests of employees more than employers, those accused of a crime more than the state, and consumers more than business. Our voter's judgment would be a pragmatic judgment about the policies most likely to foster ideal conditions of democracy in a particular political moment. A sophisticated judgment about the social requirements for real-world democratic well-being might reasonably lead to precisely these conclusions. If the voter has a defensible conception of the meaning of democracy, and if social and political conditions are such that democratic equality requires advancing the interests of labor, employees, accused criminals, and consumers, then a substantive democrat will have little reason to challenge either the foundations or the pragmatic conclusions of our fair-minded voter. One who votes on these grounds in a judicial election is enacting the procedural ideal by participating, and is enacting the substantive ideal by advancing the right kinds of judgments through her vote. The vote could not be criticized either on procedural or substantive grounds.

But on institutional grounds, there is much to criticize. The structure of the choice our voter faces in the first place is not a structure that corresponds to the key cleavages of legalism. Even if she makes the right choice from within that structure, the failure of her reasons to meet the cleavages of the judicial task means there is a failure of correspondence between her choice and the choices legitimately available within a legal paradigm. We can call this a democratic failure of an institutionalist variety, a failure of democratic processualism.

Legislative and Judicial “Policymaking”

Some will resist the idea that this distinctive legal dimension creates any more complexity than what all political registers must handle. Citizens always face complex political environments, and it is almost unheard of for political actors to advance all of their political goals at once. Trade-offs must be made. Is the complexity of the legal register different from the ordinary complexity of ordinary politics? The concern is especially pressing given the success of attitudinalist scholarship in deflating claims that the law is different from any other form of political ideology.

Because the argument against judicial elections is not about intrinsic qualities of law and politics, but is rather about the way institutional contexts support or discourage voter choice according to the dimensions that are most relevant for the judicial task, the core claims of attitudinalist scholarship are consistent with the argument offered here. The only claim this article rejects is an inference sometimes drawn from attitudinalist literature, but never actually advanced by Segal and Spaeth—that judicial decision-making is no different in kind from legislative decision-making.Footnote ⁵⁶ To resist this claim, we can excavate the difference between the structure of political judgments and that of legal judgments by comparing the discoveries that attitudinalists advance about judicial behavior with the discoveries of congressional scholars about congressional behavior. If, as unsophisticated attitudinalists might claim, judges resemble legislators in that both enact policy preferences, then the predictive value of the attitudinal model should be the same whether applied to legislators or judges. In fact, this is not so.

In their pathbreaking work on congressional attitudes, Poole and Rosenthal constructed scores of legislators' ideologies and predicted their votes on that basis. Empirical testing of the model generated two major findings: first, that almost all policy positions of legislators can be predicted from a single dimension of political cleavage (today represented by whether the legislator has a “liberal” or “conservative” stance on economic redistribution); and second, that simply from knowing a legislator's position on this unidimensional scale, a researcher can predict over 90 percent of the legislator's votes.Footnote ⁵⁷ In other words, although legislators, like judges, face a vast array of decision-making approaches (trustee versus delegate models, decisions based on committee hearings, floor deliberations, or constituent pressures, etc.), in practice those approaches seem far less significant for their actual voting behavior than the single question of where they stand on economic redistribution. The work on unipolarity shows that, regardless of his position on the meaning of the legislative function, almost everything a legislator does can be captured by his stance on economic redistribution.

Segal and Spaeth also create scores of justices' attitudes and attempt to predict their decisions on the basis of that score. Segal and Spaeth find no single dimension that emerges from this work; instead, they find that they must analyze justices' votes within distinctive issue areas (criminal procedure, civil liberties, etc.). There are many areas they simply cannot incorporate at all (federalism, unions, judicial power, federal taxation, and economic activity).Footnote ⁵⁸ For their strongest case, civil rights and civil liberties, Segal and Spaeth's model predicts just over 70 percent of a justice's votes in that domain only. Their revised work, with a reworked and more appropriate independent variable, expands the categories of cases not scrutinized; now only civil liberties and civil rights issues are examined. Most importantly, Segal and Spaeth make no claims to unidimensionality. A justice with a strong conservative position on search and seizure may or may not have a conservative position on federalism. This lack of unidimensionality may be a consequence of the Court's commitment to deciding each case on the merits: although legislators achieve unidimensionality through logrolls (among other devices), Lee Epstein, in an examination of justices' papers, found “absolutely no examples of logrolling” on the Supreme Court, although presumably logrolling could offer policy-motivated justices the same advantages as it offers policy-motivated legislators.Footnote ⁵⁹ This doesn't mean partisanship isn't a useful indicator for scholars of judicial behavior. But it means that partisanship works in far more complex ways for judges than it does for other officials. Campaign environments cannot be expected to reflect this complexity.

Even from the perspective of being motivated to entrench a political ideology, the approach of a jurist should differ from that of a legislator because it is the judges' path of reasoning, not the outcome, that creates law. Lower courts do not decide in a pro-life direction on the cases they encounter because the Supreme Court voted in a pro-life position in its last case. Their method is rather to apply the structure of reasoning laid out by the Court. On the other hand, it is the legislator's conclusion, not her path, that forms the law which binds executive and administrative officials. Skillful agents who are motivated to achieve desired political outcomes, then, should rationally approach their work differently if they are occupying a legislative versus judicial office. This institutional particularity destabilizes any imputed equation between congressional and judicial “policy-making.”Footnote ⁶⁰

Objections

But let us follow the skeptic further. Even if we excavate another set of systemic dimensions to judicial decision-making, one which attitudinalists have missed—the skeptic could still say that such an excavation is only question-begging. For the point of judicial elections is that judicial decision-making, whatever that decision-making consists of, should be guided by the judgments of voters. That Poole and Rosenthal have found that legislators operate according to one or two dimensions does not mean that such a collapsed decision space is democratically necessary. The presence of more dimensions of judicial, or legislative, decision-making means only that officials, and hence voters, face a complex decision environment. If legislators make decisions according to three, four, fourteen, or forty dimensions of concern, this means only that voters must make more complex judgments than otherwise—not that their choice is rendered meaningless. Why isn't it simply the case that judicial elections offer voters a complex choice, but not a meaningless one?

Theoretically, democratic choice does not require a low-dimensional decision space in electoral campaigns or anywhere else. Voters, in fact, should be applauded when they manage to accurately evaluate candidates on multiple dimensions at the same time. We should make efforts to equip all voters with the sophistication they need to carry out those analyses. The problem for judicial elections rests in the application of this theory to the real world. In a low-information world, multidimensional, complex issue spaces undermine the likelihood that voters will know what choice they are making, and that they will reliably be able to register the choice they intend. If 90 percent of a legislator's behavior can be predicted from his stance on economic redistribution, even a mildly sophisticated voter can ascertain a great deal about him in a very low-information environment. The simplified decision environment supports communicative transparency, which gives us reason to trust that voters know who and what they are selecting. Such a collapse may bode poorly for good governance, but it bodes well for the efficacy of information short-cuts. The argument here, then, is not that democratic choice requires low dimensionality. It is rather that low-dimensionality provides a way in the real world to enhance the capacity of voters to register the choice they intend.

Let us return to Bush v. Gore. Should voters re-elect Justice Scalia? If the only information they had about Scalia was that he voted to advance the political fortunes of the Republican Party, they would be voting on the basis of his “politics,” but insofar as he actually was exercising a judicial function in that decision, the content of the interpretive ideology he advanced was liberal. Voters who wished to vote on the basis of the political decision that was at stake in the law—namely, whether equal protection should be construed to bind election procedures in this way—may want to make a “political choice,” but they would find that partisan information about the judge would not give them the information necessary to advance their judgments about the law. It is as if the only information that voters had available to them about legislative candidates was a price schedule for the legislator's votes on various bills, with candidates separated into high-rollers and cheapskates. The problem is not only that this method of decision-making is corrupt; it is that elections structured around such information fail to offer voters the opportunity to make democratic decisions on the basis of political information that is relevant. Because it is not sensible to imagine that current partisan structuring of the political environment will create cleavages and hence cues that correspond to legal, as opposed to partisan divides, it is not reasonable to imagine that voters in judicial elections will acquire the informational short-cuts they need to register a politically sensible choice about matters of legal interpretation.

My argument is contingent on several features of the electoral environment. First, it is contingent on low levels of information in that environment and hence the overwhelming significance of accurate cues for structuring voter choice. In higher-information environments the significance of partisanship might be different. If voters do not need partisan cues to make good judgments, or if partisan cues are not effective at carrying relevant information for any office (as might be the case in times of very low polarization), then there is less reason to worry about partisanship structuring voter cognition in ways that undermine voters' capacity to advance legal judgments. (Of course, we might then wonder what other resources are available for voters.)

The argument is also contingent on the nature of the substantive issues that partisanship tracks. One can imagine that, if judicial elections became extremely widespread, parties might develop cues that corresponded to principled legal distinctions, and voters would be able to reliably use those cues to vote their actual preferences on matters of political/legal commitment. Poole and Rosenthal's findings on the value of unidimensionality for communicative transparency make such circumstances difficult to imagine. But there is nothing here to resist the conclusion that, were such cues developed, judicial elections could indeed deliver the opportunity for voters to make democratic decisions on matters of legal interpretation.Footnote ⁶¹

Finally, the argument is contingent on partisan cues being used as deliberative tools. We can imagine contexts where partisanship functions to motivate actors into a forum, but loses its capacity to structure the practices of reason-giving once that forum is mobilized. Perhaps in appropriately contestatory appointments proceedings, where the decision-makers have more subtle and complex information available to them, partisan contestation could enhance the democratic authority of the selection procedure by providing a motivating force for the construction of a more appropriately-structured deliberative environment. Such cases must be examined in their particularities.

I began by affirming the value of democratic judgment on matters of legal interpretation. Partisan judicial elections, I argue, undermine that possibility. A truly committed critic might ask whether it is really so important that democratic judgment on matters of legal interpretation be supported. Why not let the voters decide whether they wish to vote on matters of legal interpretation or other grounds, and incorporate their decision into their votes on candidates?

This kind of criticism is faulty because it assumes precisely what this article denies—that the election mechanism in a partisan campaign environment can really induce cleavages to emerge that will allow voters to systematically register their preferences on such complex matters. Institutional design in a democracy should seek to minimize, not maximize, the cognitive demands voters face. Other avenues, beyond voting, are open for citizens to express complex positions on the value of the institutions themselves. Asking that voters choose whether they want judges to decide on the basis of partisanship, political ideology, or legalism needlessly complicates an election environment. That question is one that is appropriately addressed at the level of institutional design.

Moreover, at the level of institutional design, I am not arguing that it is procedurally illegitimate for citizens to authorize a constitution that allows for partisan judicial elections. I rather suggest that if citizens are interested in eliciting the judgment of voters on matters of the judicial function, they should think twice before ratifying such a constitution. Constitutional designers are of course free to redesign institutions entirely, perhaps devising quite a different institution and naming it the “judiciary.” Yet insofar as that institution is making judgments that are likely to be systematically distorted by partisan cleavages then the effort to enable democratic judgment through partisan elections is unlikely to be successful.