The American system of representative government is broken and in need of repair. That is a common refrain in a time of extreme partisan polarization and government paralysis in the lawmaking process. Many elected officials do not respond to the preferences of the median voter, but instead adhere to the desires of their party's base because of a lack of inter-party competition for legislative seats. Moreover, the people cannot effectively hold elected officials accountable for failed government policy because of the multi-layered lawmaking structures and complex lawmaking procedures that make it difficult to trace policy decisions to a particular actor or party.
At the same time, there does not seem to be a more compelling alternative to the representative system of government. Popular democracy, which involves lawmaking through citizen-proposed initiatives or government-initiated referenda, is sometimes offered as an alternative that would fix the flaws of representative democracy. However, popular democracy has proved lacking because voter resources and competence to make laws is limited.
Ethan Leib and Christopher Elmendorf deserve great credit for thinking imaginatively about how to repair both representative government, or what they describe as party democracy, and popular democracy because the two are surely in need of it. In their article, Why Party Democrats Need Popular Democracy and Popular Democrats Need Parties, Leib and Elmendorf respond to two pathologies of modern representative government-sustained breakdowns in competition between the parties and what they refer to as occlusion of accountability. Both problems result in a lack of responsiveness by legislators to the median voter. To redress these pathologies, they offer several innovative solutions that involve the creation of complementarities between party democracy and popular democracy to better secure lawmaking that is both responsive to the median voter and accountable to the people.
Leib and Elmendorf's central prescription is that in states where state legislatures suffer from sustained competitive breakdown and use complex lawmaking procedures like a supermajority voting requirement, the majority party should be able to bypass those procedures and refer complex bills like a budget to the people. They argue that popular democracy would provide a better check on majority party action than supermajority voting rules because it would create strong incentives for the majority party to position itself and its policies to be more responsive to the median voter. They also suggest that use of popular democracy in this way would lead to greater popular understandings of what each party stands for and thus ameliorate problems of accountability occlusion.
Both of these results are salutary and would surely improve both party and popular democracy. However, as the proposal is currently constructed, I argue that it has the potential to do more harm than good for three principal reasons.
First, insofar as the majority party actually does what Leib and Elmendorf suggest it should do, which is to refer complex matters like state budgets to the people with the party label attached to it, voter competence will not necessarily be enhanced. Instead, voter decisions will often be dictated by party cues. As social science studies demonstrate, party cues are the most powerful shortcut that voters depend on when they do not have the time or resources to systematically evaluate complex matters. Through the introduction of party cues into ballot proposition elections, voters with limited resources could therefore be transformed into mere tools of the party dependent on the party's cues to make their decisions. In a state suffering from sustained competitive breakdown in which an overwhelming majority of voters align with one party over the other, the same party in government can take advantage of voter dependence on party cues to secure adoption of any complex proposal it refers to the people. Therefore, the result of the adoption of Leib and Elmendorf's proposal would be the transformation of popular democracy into a supplicant of party democracy.
Second, adoption of the Leib and Elmendorf proposal has the potential to undermine the processes of deliberation and compromise that are a critical and valued feature of our representative process. Deliberation and efforts to compromise surely have costs in terms of slowing down the legislative process, sometimes to the point of stalemate. However, Americans continue to favor bipartisan approaches to problem solving because it produces more reasoned decision making and involves the consideration of the interests of minorities. The Leib and Elmendorf proposal will create incentives to bypass procedures like the supermajority voting requirement that are intended to advance deliberation and consideration of minority interests. In contexts of sustained competitive breakdown where the party in government could be nearly assured that any complex proposition they send to the public would be passed, this opportunity to bypass supermajority voting requirements could actually lead to lawmaking that is more responsive to the median voter in the party rather than the median voter in the public.
Finally, adoption of the proposal would do little to redress the occlusion of accountability that arises from the inability of voters to trace laws and their consequences to particular lawmakers or a particular party. By adding another layer of decision makers to the mix-the people in the referendum process proposed by Leib and Elmendorf-it will in fact be more difficult for the public to trace laws and their consequences to particular lawmakers or a particular party.
In the following pages, I address each of these three criticisms in turn. I also suggest modifications to the proposal that would better serve the goal of repairing our democracy. As modified, Leib and Elmendorf's cutting-edge proposal could lead to lawmaking that is more responsive to the median voter.
COMPLEMENTARITIES AND THE ELUSIVE GAINS TO VOTER CONFIDENCE
My first concern with the Leib and Elmendorf proposal is that it is not clear it will do much to correct the chief defect of popular democracy-the lack of voter competence to evaluate ballot propositions. Voters deciding on ballot propositions, such as referenda and initiatives, often lack the time or resources to seek out and systematically evaluate the available information to ascertain whether the proposition accords with their interests. This problem becomes more pronounced as ballot propositions increase in complexity. To overcome this information deficit, voters use heuristics, or information shortcuts, to make decisions. These heuristics, in the context of ballot proposition elections, have come in the form of interest group or candidate endorsements, news media, and paid advertisements that help the voter to assess whether the proposition is aligned with her interests. Historically, party identification, which is the most relied upon heuristic in candidate elections, has not been systematically used in ballot proposition elections.
Leib and Elmendorf argue that the majority party should be able to bypass supermajority voting requirements and refer bills to the people for a vote. As part of their proposal, they suggest that when sending referenda to the people, the parties should be able to explicitly endorse them on the ballot itself. This, they suggest, would help overcome the information deficit that voters face in popular democracy, as it would provide the most powerful signal of how a referendum aligns with the interests of the individual.
While it is certainly true that Leib and Elmendorf's proposal will provide voters with an information shortcut that was previously unavailable, it is not clear that such a shortcut will actually enhance voter competence. What we know from social psychology is that the greater the complexities involved in a decision, the more individuals will rely on heuristics. And the more powerful the heuristic, the more likely individuals will rely on it. In addition to requiring voters to evaluate complex matters, like a budget, the Leib and Elmendorf proposal would introduce the most powerful political heuristic into popular democratic decision making-the heuristic of party identification.
Because party cues are such a powerful heuristic, the introduction of these cues to ballot proposition elections will likely lead voters to rely on them at the expense of other cues. Voters may do so even though other cues like interest group endorsements provide voters with a more accurate assessment of whether a referendum aligns with their interests. In addition, because complex matters like a budget require quite a bit of time and resources to evaluate, the introduction of party cues might decrease the incentives for voters to engage in a more systematic evaluation of the referendum. Neither reliance upon general and sometimes invalid cues nor the substitution of cues for systematic evaluation will necessarily produce gains to voter competence over the alternative of a referendum process that does not include party cues. To see the potential elusiveness of the gains to voter competence, one needs only look to the possible effects of the introduction of party cues on voter evaluation of party-sponsored state budget referenda.
Leib and Elmendorf suggest that the California Constitution could be restructured so that during an impasse in budget negotiations, the two parties could send their budget proposals to the voters through a referendum. However, if the California State Legislature had adopted Leib and Elmendorf's proposal this year, the result would have been one wallop of a referendum. The 2012-2013 budget Governor Jerry Brown has proposed is approximately 1500 pages and the budget summary alone is nearly 200 pages! These pages are full of complex spending tradeoffs involving many different government programs and agencies that are challenging for any full time legislator, let alone any busy individual, to absorb and evaluate. To make matters more difficult, under Leib and Elmendorf's proposal, voters would have had to choose from two budget proposals-one from the majority party and the other from the minority party.
What would this have produced? Voters would have likely voted on budgets principally on the basis of their personal party identification rather than on the basis of any sort of systematic evaluation of the budgets. Most voters would have simply not had the time or the resources to do otherwise. While some studies show that through the use of heuristics, limited information voters decide similarly to fully informed voters, other studies show that heuristics introduce biases and errors into decision making, resulting in votes that do not accurately reflect individual interests. Voters' inability to separate invalid from valid cues is what prevents them from making accurate decisions.
For example, the party label of Democrat attached to a budget is traditionally associated with higher taxes, particularly on the wealthy, and more spending on education, health care, and protection of labor unions and the environment. However, in some contexts of sustained competitive breakdown, the Democratic Party might pass a budget that involves more aggressive spending and higher taxes than the voter anticipated on the basis of the party label-so much so that the budget may accord less with the interests or preferences of the more moderate Democratic voter than the competing Republican budget. Yet because voters are not able to engage in any systematic evaluation of two 1500 page budget proposals, they might rely on the party cue thinking that when they are voting for a Democratic budget, they are voting for a budget that reflects their interests or preferences. And because of the difficulty of tracing outcomes to particular votes, voters may never know that they in fact voted against their interests or preferences.
Moreover, it is not clear that voters are made more competent by the addition of party labels to the ballot referendum. In fact, an argument exists that they are made less competent because party labels are often so broad and amorphous that they do not provide much information about how the policy aligns with voters' interests. Yet because most voters, even many of those voters registered as independent, continue to have a primordial allegiance to a party label, they will think these labels mean more than they do and weigh them more than they should. In the process, voters may ignore or undervalue perhaps more informative cues about how a referendum aligns with their interests, such as endorsements from organizations and interest groups.
The elusiveness of the gains to voter competence from Leib and Elmendorf's proposal when dealing with complex matters in contexts of sustained competitive breakdown thereby diminishes the need for the party in power to enact or refer bills that are responsive to the median voter. Instead, as I discuss in Part II, because of these limitations on voter competence, the proposal will enhance the capacity of the majority party to pass laws responsive to the party core by weakening the capacity of the minority party to secure concessions through deliberation and compromise.
COMPLEMENTARITIES AND ITS COST TO DELIBERATION AND COMPROMISE
My second concern with the Leib and Elmendorf proposal is that it would undermine two of the most prized features of our political process, deliberation and compromise between majority and minority parties. Recent public clamor for bipartisan lawmaking has been premised on the idea that the two parties should come together and hammer out their differences to make laws that are broadly acceptable. It is hoped that deliberation and compromise will lead to the passage of laws that serve the interest of voters in the middle of the political spectrum and that account for a broader array of interests than lawmaking that proceeds in the absence of inter-party engagement. A principal tool that has been used to secure inter-party deliberation and compromise has been the supermajority voting requirement. Leib and Elmendorf criticize this deliberation tool. They argue that it contributes to political stalemates by allowing the minority party to block passage of bills and to accountability occlusion by frustrating the ability of voters to trace particular policy decisions to a particular party. Therefore, allowing the majority party to bypass the supermajority voting requirement and refer bills directly to the people during times of sustained competitive breakdowns avoids these problems.
Whether Leib and Elmendorf's proposal would redress the problems of political stalemates and accountability occlusion is a point I will return to in Part III. For now, what I find particularly troubling about the proposal is that even assuming it would reduce gridlock and increase lawmaker accountability, it would do so at the cost of decreasing the incentives for the majority party to carefully weigh and consider the views of the minority party or to try to settle differences in the proposed legislation by mutual concession. The result would be the passage of laws that are often less responsive to the preferences of the median voter than those that are passed pursuant to the supermajority voting requirement.
To see why this would be the case, it is important to recognize the primary source of sustained competitive breakdown. In most states where there are sustained breakdowns in party competition such that one party consistently controls a majority of the seats in the legislature, the reason for this dominance is that the constituents have overwhelming allegiance to one party. For example, Democrats have dominated state legislatures in the Northeast and West over the last twenty years because most northeasterners and westerners are Democrats. Because of the mobility of population and the need to secure governor approval for districting in many states, it is more rarely the case that sustained one-party domination can be secured through manipulating legislative districts to give one party consistent control over both houses of state legislatures. Surely, Democratic domination of state legislatures in the Northeast and West for the last twenty years could have been enhanced by their ability to manipulate district lines to their advantage, but such manipulation is usually not the source of sustained Democratic advantage.
Since sustained competitive breakdown results from differences in the allegiance of constituents to a particular party, the majority party will have less incentive to deliberate or compromise with the minority party to secure passage of legislation without a supermajority requirement that forces them to do so. Under the Leib and Elmendorf proposal, the mere threat that a majority party would bypass supermajority procedural requirements and refer bills, particularly complex bills, directly to the people with the party's label attached would diminish the need for the majority party to make concessions on its legislative agenda.
Because of the importance of party cues, the majority party in a state that suffers from sustained competitive breakdown would be confident that if it sent a referendum to the voters-particularly a complex referendum like a budget-with its party label attached to it, there would be a strong likelihood that the voters would favor it over an alternative minority party-sponsored budget. The majority party would also be confident that the voters would likely favor its budget proposal irrespective of the substance of the two budgets and irrespective of whether the alternative minority party-sponsored budget better accorded with their interests or preferences. This would potentially give the majority party the incentive to push a budget that would do more to satisfy the median voter in their party than the median voter in the jurisdiction that the elected official represents in order to strengthen his or her standing against party primary challenges from more partisan candidates. In states dominated by one party, party primaries are often the only elections that really matter. The result would potentially be a more partisan budget.
To obtain these advantages, the majority party would not even have to use the supermajority bypass. Instead, a threat to use it would be enough to silence minority party opposition in states with sustained competitive breakdowns. Whether the supermajority bypass was to be used or threatened, the likely result would be diminished need for the majority party to deliberate and compromise with the minority on important matters like a budget. Legislation would perhaps be more efficiently passed, but it would also be potentially less responsive to the preferences of the median voter and definitely less bipartisan.
COMPLEMENTARITIES AND THE ELUSIVE GAINS TO ACCOUNTABILITY
What about the gains in terms of increased accountability? Wouldn't this increased accountability of the majority party ensure that it passes laws and refers bills responsive to the median voter? Leib and Elmendorf suggest that allowing the majority party to bypass supermajority voting requirements and refer bills that they sponsor to the people will enhance political accountability. To the extent that one party is responsible for a legislative agenda, the people will be better able to trace laws and their consequences to that party. This can be contrasted with a decision making process that includes a supermajority voting requirement. Under such a system, the compromises the majority party must often make with the minority party to attain supermajority support for a bill limits the ability of the voting public to ascertain the source of particular lawmaking solutions and from holding parties accountable for their actions. Leib and Elmendorf suggest that their proposal would clarify the lines of accountability such that the people would be better able to trace policy and its consequences to a particular party.
The principal argument of the Progressive-era architects of popular democracy was that it would provide a check on some of the excesses of representative democracy, including special interest domination and party bossism. Today, commentators think that the system of popular initiatives and referenda has been somewhat effective in providing such a check. However, there is no evidence that popular democracy has been a particularly effective tool for holding elected officials accountable for the policy consequences of laws. Instead, popular democracy likely undermines rather than enhances this form of accountability by stripping from elected officials the exclusive power to make laws and giving some of this lawmaking power to the people. Legislation made on the initiative of the people breaks the chain of traceability of decisions to elected officials. It is therefore unlikely that the people would hold officials accountable for the consequences resulting from popular lawmaking since they neither proposed nor passed the law.
To the extent that giving the majority party the option to bypass the supermajority requirement makes it more likely that the lawmaking function is delegated to the people, it will be less likely that the party will be held accountable for the consequences of its policy decisions. This will be the case even accounting for the differences between the form of lawmaking by referenda under the Leib and Elmendorf proposal and lawmaking by initiative. Under the Leib and Elmendorf proposal, parties would sponsor referenda that would be sent to the people when the majority party chose to bypass the supermajority voting requirement. Such sponsorship is presumed to give voters a sense of what each party stands for. But this presumption is dependent on a critical assumption-that voters will systematically evaluate the referenda and hold the party accountable for its content. Such systematic evaluation may make sense for more simple referenda that impose low information costs for voters to evaluate. However, as discussed in Part I, to the extent the voters are required to evaluate more complex referenda like a budget, voters will have to rely on information shortcuts about what the party stereotypically stands for that are not necessarily grounded in the specifics of the policy being sponsored.
Voters lacking a clear, systematic sense of what is being sponsored and what they are voting for will have a difficult time tracing particular policy consequences to party-sponsored referenda. This challenge of traceability is further exacerbated by the fact that the people are the ones ultimately responsible for passing the referendum. Since the people have the opportunity to make the final choice between the alternatives, they are less likely to trace responsibility for resulting outcomes to the party that sponsored the winning proposal in the referendum. The additional layer of lawmaking introduced by the referendum thus undermines the accountability of elected officials for the policy consequences of laws. It is therefore not clear that to the extent the majority party refers complex bills to the people, the concerns about accountability would be any less than they are under a lawmaking system that includes a mandatory supermajority voting requirement.
Of course, the majority party may be able to use the threat of a supermajority bypass to cow the minority party into agreeing not to block passage of partisan legislation. The majority party would therefore be able to both avoid referring bills to the people and pass more partisan laws. Under a system of more partisan lawmaking, voters would presumably be better positioned to hold the majority party accountable for its policy decisions. This, however, assumes that the party would want the people to hold it accountable for the decisions that it makes. If lawmakers were principally focused on re-election, as most studies of legislative behavior suggest they are, rather than the transparent advancement of a party platform, then the assumption would be unlikely to hold. To the extent that there is a sustained breakdown in competitiveness and lawmakers are principally concerned with re-election, it may be rational for legislators of the majority party to avoid accountability for difficult and potentially electorally costly decisions. The Leib and Elmendorf proposal provides majority party lawmakers with the very means for doing so. The majority party could simply delegate these potentially politically difficult decisions to the people through a party-sponsored referendum. Because of the dominant allegiance of voters to one party over the other in states suffering from sustained breakdowns in competitiveness, the people would likely just give a rubber stamp of approval to the referendum, thus allowing the majority party to liberate itself from responsibility for potentially traceable policy consequences. The accountability gains from party and popular democracy complementarity are therefore likely to be elusive.
THE CONDITIONS FOR GAINS FROM COMPLEMENTARITIES
To sum up briefly, providing the majority party with the option to bypass supermajority requirements and refer complex bills to the people with the party label attached in contexts of sustained competitive breakdown could, because of limitations on voter competence, exacerbate defects of party democracy. It would do so by creating incentives for more partisan lawmaking that is less accountable to the median voter. The one thing that the proposal would likely do successfully is redress stalemates in the political process. The threat of bypassing the supermajority voting requirement and referring bills directly to the people would likely force the minority party in states that suffer from sustained competitive breakdowns to capitulate or face the potential of more unfavorable legislation being referred to and approved by the people. These efficiencies, however, come at an enormous cost to deliberation, compromise, and lawmaking that accounts for the interests of the minority-features that have been central to state and federal lawmaking processes since the Founding. Despite my criticisms of the Leib and Elmendorf proposal, I think the idea remains promising if modified to account for these concerns. Under three conditions, the proposed reform could represent an effective means of incentivizing legislatures to pass laws more responsive to the median voter.
First, the proposal could be effective in states that do not suffer from sustained competitive breakdowns-that is, in states where people are relatively closely divided in their allegiance to parties or even in states where voters are only loosely allied to the parties, such as those with a high percentage of independent voters. In these states, the threat of bypassing the supermajority vote requirement and referring partisan legislation to the people has risks for the majority party. Despite the party label attached to the proposal, there remains a possibility, and perhaps a strong one, that if the minority party referendum were more responsive to the preferences of the median voter, the voters would approve it over the one sponsored by the majority party. Given this risk, the majority party would have greater incentives to deliberate and compromise with the minority party to pass a bill responsive to the median voter instead of bypassing or threatening to bypass the supermajority requirement. Only when the political parties fail to achieve compromise would the majority party want to bypass the supermajority requirement because the costs of doing nothing with respect to particular legislation, like a budget, are too high. When the majority party does refer a bill to the people, it will have a strong incentive to construct a referendum targeted toward obtaining the approval of the median voter in the public rather than the median voter in its party to secure passage in a competitive state.
Second, the Leib and Elmendorf proposal could be workable if it limited legislatures to sending simple, single-subject referenda to the people. Voters would have little opportunity to systematically evaluate 1500 page budget proposals and instead would likely have to depend almost exclusively on cues, particularly party cues, in making their choice. However, voters would have a greater opportunity to systematically evaluate a single subject referendum and more independently assess whether it accords with their interests. For example, rather than referring an entire state budget to the people, the majority party could refer discrete and disputed parts of the budget to the people with an explanation of the tradeoffs involved. Simplifying the referendum would not mean that voters would stop paying attention to party cues, but it would be less likely that voters would be overly dependent on them. And if voters were not overly dependent on party cues, then the risk to the majority party that the voters would approve the opposition's less partisan referendum grows. This would correspondingly increase the incentives of the majority party not to bypass the supermajority requirement and instead deliberate and compromise. Also, when the benefits outweigh the costs of bypass in a stalemate situation, the decreasing dependence of voters on party cues would also increase the incentives of the majority party to refer bills to the people that are targeted at the preferences of the median voter.
Finally, in order for the Leib and Elmendorf proposal to be effective with respect to complex legislation, the referendum must not include party labels. For the reasons described above, such party labels will not ameliorate the problem of accountability occlusion. And the allure of this heuristic is too strong and the potential for errors or biases in judgment is too great when such labels are included. In addition, by excluding party labels, the majority party once again bears greater risk for bypassing the supermajority requirement and referring partisan bills to the voters. Since the voters will have to rely on cues over which the party has less control, such as cues from the media and interest groups, the majority party in contexts of sustained competitive breakdown would have less assurance that their referendum would pass. Under this condition, the majority party would again have greater incentives to either compromise on bills in the political process or refer bills responsive to the preferences of the median voter.
To correct dysfunctions in our democratic process, we need more innovative and thoughtful proposals like those offered by Leib and Elmendorf. Many of the problems that I have described with respect to their proposal may not be applicable to the other proposals included in their paper. That assessment will have to await future evaluation. It is important that for whatever democratic design proposals offered, there is continuous vetting by scholars and policymakers so that a fuller consideration of costs and benefits can be ascertained. I hope for the opportunity to continue to take part in this conversation in the future.
.See, e.g., Richard H. Pildes, Why the Center Does Not Hold: The Causes of Hyperpolarized Democracy in America, 99 Calif. L. Rev. 273, 275-76 (2011) (suggesting that American democracy is characterized by extreme polarization and attributing it to deep historical and structural transformations initiated by the Voting Rights Act).
.See Chris Tausanovitch & Christopher Warshaw, How Does Electoral Competition Affect the Link Between Constituents and Roll Calls in Congress? 17 (July 13, 2011) (unpublished comment), available at http://www.stanford.edu/~cnt/Competition_Paper_July2011.pdf (finding that responsiveness of legislators to the median voter depends on the competitiveness of the district); see also 2012 Competitive House Race Chart, The Cook Political Report (March 23, 2012), available at http://cookpolitical.com/charts/house/competitive_2012-03-23_10-46-41.php (considering only 55 of the 435 congressional seats to be competitive for the 2012 election).
.See R. Douglas Arnold, The Logic of Congressional Action 119-20 (1990) (describing congressional use of lawmaking procedures to hide actions from the public or to make actions difficult to trace to particular legislative actors); James H. Kuklinski & Norman L. Hurley, On Hearing and Interpreting Political Messages: A Cautionary Tale of Citizen Cue-Taking, 56 J. Pol. 729, 732 (1994) ("Given the complexity of national policy making, it is the rare citizen who can determine whether or not a particular policy has worked.").
.See, e.g., Richard R. Lau & David P. Redlawsk, Advantages and Disadvantages of Cognitive Heuristics in Political Decision Making, 45 Am. J. Pol. Sci. 951, 953 (2001) (describing a candidate's party affiliation as perhaps "the most important political heuristic").
.See, e.g., James H. Kuklinski, Daniel S. Metlay & W. D. Kay, Citizen Knowledge and Choices on the Complex Issue of Nuclear Energy, 26 Am. J. Pol. Sci. 615, 621 (1982) (finding with respect to an initiative on the nuclear energy that the "electorate highly skewed toward the low end of the knowledgeable scale").
.See, e.g., Regina P. Branton, Examining Individual-Level Voting Behavior on State Ballot Propositions, 56 Pol. Res. Q. 367, 368 (2003) (noting the complexity of voter decision calculations in ballot elections in the absence of a party cues, in which voters are forced to rely on initiative campaigns, elite messages, and media cues as alternatives); Mark Forehand, John Gastil & Mark A. Smith, Endorsements as Voting Cues: Heuristic and Systematic Processing in Initiative Elections, 34 J. Applied Soc. Psychol. 2215, 2217 (2004) (finding that in the absence of cues about party affiliation voters tend to rely most on interest group endorsements).
.See, e.g., Jeffrey J. Mondak, Source Cues and Policy Approval: The Cognitive Dynamics of Public Support for the Reagan Agenda, 37 Am. J. Pol. Sci. 186, 188 (1993) ("Reliance on heuristic processing of source cues is most likely when the individual has high need for cognitive efficiency.").
.The substitution of heuristics for systematic evaluation of referenda is particularly disconcerting given the strong evidence that heuristics introduce bias into decision-making. See generally Daniel Kahneman, Paul Slovic & Amos Tversky, Judgment Under Uncertainty: Heuristics and Biases (1982).
.Leib & Elmendorf, supra note 5, at 105. Leib and Elmendorf also suggest convening a group of randomly selected citizens to resolve legislative budget stalemates. Id. at 104. The focus of this response is not on this proposed suggestion, although many of the arguments against the proposal to refer complex bills directly to the people are applicable to the proposal to refer complex bills to a select group of citizens.
.See, e.g., Larry M. Bartels, Uninformed Votes: Information Effects in Presidential Elections, 40 Am. J. Pol. Sci. 194, 217 (1996) (finding through an empirical analysis that uninformed voters who rely on information shortcuts "do significantly better than they would by chance, but significantly less well than they would with complete information"); Lau & Redlawsk, supra note 10, at 966 ("[H]eavy reliance on political heuristics actually made decision making less accurate among those low in political sophistication.").
.See, e.g., Laurel Harbridge & Neil Malhotra, Public Preferences for Bipartisanship in Congress 11-12 (Aug. 31, 2009) (unpublished paper), available at http://papers.ssrn.com/sol3/papers .cfm?abstract_id=1451225 (finding on the basis of original survey experiments that when Congress frames its activity as bipartisan, people show significantly more support for this institution of government).
.A supermajority voting requirement that has been subject to a great deal of debate is the filibuster used in the United States Senate. See Standing Rules of the Senate, 110th Congress, R. XXII, available at http://rules.senate.gov/public/index.cfm?p=RuleXXII. Proponents of the filibuster argue that it is an essential tool for the advancement of deliberation between the parties. See, e.g., Matthew J. Franck, In Defense of the Filibuster, Public Discourse (Jan. 29, 2010), http://www.thepublicdiscourse.com/2010/01/1129. While opponents criticize the filibuster as a tool used by the minority party in Congress to obstruct majoritarian lawmaking. See, e.g., Catherine Fisk & Erwin Chemerinsky, The Filibuster, 49 Stan. L. Rev. 181, 184 (1997) (arguing that the current incarnation of the filibuster no longer serves as a tool to ensure deliberation, but instead "is simply a minority veto" used to obstruct laws supported by the majority).
.This result would be consistent with the finding that "both Republicans and Democrats are responsive to the median at high levels of competition, but responsive to their co-partisans at low levels of competition." See Tausanovitch & Warshaw, supra note 2, at 15. Assuming, as rational choice theorists do, that one of the primary goals of elected actors is to be re-elected, then in non-competitive jurisdictions the principal threat to re-election is often at the primary stage. See, e.g., Richard F. Fenno, Congressmen in Committees 1 (1973) (describing the three goals of congresspersons as re-election, influence within the House, and good public policy). Similarly, elected actors concerned with primary challenges will likely be motivated to refer laws responsive to the party median rather than the median voter in the jurisdiction.
.This argument follows from the predictions of responsible party government theorists. See Austin Ranney, The Doctrine of Responsible Party Government: Its Origins and Present State 12 (1954) (arguing for a system of strong and cohesive parties with the power to make policies without intraparty compromise, a system of party government analogous to that which exists in the British Parliament that allows the people to trace policy consequences to a particular party and hold them accountable).
.See, e.g., John Matsusaka, For the Many or the Few: The Initiative, Public Policy, and American Democracy 3 (2008) (finding that the initiative process leads to the promotion of the preferences of the public over those of special interest groups).
.See Magleby, supra note 4, at 186 (explaining "the initiative serves as an easy way for the legislature to avoid deciding an issue, passing it along instead to the voters" and thereby "discourages legislative accountability").