Following a narrow 49-51 defeat of a bill to extend abortion rights at the federal level, Senator Elizabeth Warren (D-MA) sounded an alarm that it was time to end the filibuster. The esteemed Senator was quoted as saying, “I believe in democracy, and I don’t believe the minority should have the ability to block things that the majority wants to do. That’s not in the Constitution.” Senator Bernie Sanders quickly chimed in with his own critique of filibustering, linking it to the abortion rights issue.
Ironically, the bill failed because it could not garner simple majority support; it had nothing to do with the filibuster rule, which raises the threshold of support needed for a bill to pass. It might be too much to ask our elected legislators to understand the basic workings of government because they are too busy legislating after all!
Nonetheless, Warren’s statement reflects a broader effort to rid the Senate of this practice. For those unfamiliar, a filibuster is a technique used to prevent a bill from moving to a vote on the Senate floor by calling for continued debate on the issue. For the debate to be ended, 60 Senators must vote for cloture. Arguments against the filibuster appeal to a simple majoritarian view of democracy, rejecting the notion that some issues are so important as to require supermajority support.
But is such a simple perspective on democracy warranted (or “Warren-ted”)? Might economics, or more specifically public choice theory, have something to add to this conversation which would temper calls to eliminate the filibuster?
Public Choice: Choosing How to Choose
While not the be-all and end-all of democracy, voting is a critical component of making choices in representative governments. Public choice theory, a subfield of political economy, examines how we select our voting rules (or make choices more generally). Even before a vote on a specific policy is undertaken, there needs to be some agreement on what the voting procedure will be. In other words, public choice is the study of how we choose to choose.
James Buchanan and Gordon Tullock contributed to this literature in 1962 with their landmark work The Calculus of Consent. One of their central arguments was that the process of voting, irrespective of the issue, entails two important costs – decision costs and external costs – that must be considered when designing constitutions and electoral rules. Understanding these costs, and how they affect what voting rule is best for what issue is important for understanding why the filibuster might be a very important tool in representative democracy.
Decision Costs: Voting Isn’t Free and Easy!
Decision costs are the costs associated with collectively organizing enough individuals to support a vote. The more individuals needed for passage of a policy, the more costly the process will be in terms of time and resources expended, ceteris paribus.
To understand this, consider two extremes. First is the dictatorial rule wherein one person gets to decide the policy irrespective of what any other person wants. This is a relatively low-cost decision rule, as the dictator only needs to consult himself.
At the other extreme, a unanimity rule imposes very high decision costs on the group because everybody in the group must be in agreement for the policy to pass. Given that some people may have different preferences, negotiation amongst the members of the group must occur to arrive at full consensus. This takes time and energy.
Additionally, given that unanimity requires everyone to agree, any one individual could hold out support until granted some concession. This increases the costs of unanimity as every potential holdout will need to be appeased somehow. Unanimity moves slowly.
Not surprisingly, the larger the relevant voting group is, the more costly voting becomes. A group of three friends choosing a restaurant will probably not require much effort, particularly if their preferences are relatively similar (which may be the reason they are friends). However, a group of 100 Senators who represent different geographic constituencies with varying preferences will have a much harder time coming to agreement.
As the number of voting decision-makers grows, the costs naturally rise. Between the extremes of solo dictatorship and unanimity, decision costs increase along some curve (see Figure 1). The slope of this curve can be determined by variables other than size, including the diversity of the group, issue saliency to each member, and communication technology.
External Costs: Voting Imposes Costs on Others.
The second set of voting costs are the external costs placed upon individuals that result from the decision made (see Figure 2). In other words, voting imposes negative externalities on some individuals. This is important to remember, as there is a general notion, often taught in high school civics classes, that voting helps divine the “will of the people.” But not everybody benefits from a vote, particularly those who voted against the proposal at hand, which should seem obvious. Under simple majority rule, 51 percent of the population can vote themselves a benefit and impose the cost on the other 49 percent. The “will of the people” only truly exists if there is unanimous agreement, a rare event indeed.
Buchanan and Tullock pointed out that external costs tend to decrease with the size of the population needed to pass a vote. A dictator can obviously impose high costs on the rest of the population. Unanimity, on the other hand, minimizes external costs as everybody in the unanimous coalition will want to minimize the negative aspects of the policy imposed upon themselves (see Figure 2). Again, between these two extremes is a curve determined by a variety of factors including the overall costs of the policy being voted upon and the saliency of the issue to losing minorities.
The Optimal Voting Threshold
One of the implications of this analysis is that there are different voting thresholds for different issues being considered. There is no “one-size-fits-all” voting rule for every policy that comes before a group. For Buchanan and Tullock, the optimal voting threshold (k) is determined where the decision and external cost curves intersect, as that will represent the minimum combined decision and external costs (see Figure 3a). This threshold could be a plurality, simple majority, or a super-majority.
This optimal voting threshold can vary by issue depending on the slopes of the decision and external cost curves. For a policy that isn’t all that costly – e.g., declaring National Pancake Day – the threshold for a vote can be set below a majority since it imposes few costs on anybody (see Figure 3b). Granted, there likely will be some waffle partisans who are upset, but in reality the saliency of this issue is quite low. (Ironically, symbolic votes such as this are often used to teach young students about the ease of voting, making it appear much more important as a method of social choice than it should be.)
On the other hand, a tax policy that raises middle class taxes significantly, or a constitutional amendment restructuring the nature of government and civil rights, will undoubtedly impact a large number of people, and likely have high external costs. Not surprisingly, such hefty decisions should require a supermajority, as seen in Figure 3c. If a policy is going to harm a significant portion of the population, the legislature better have strong support for that policy.
Variations in the decision cost curve (not shown) may also have an impact on the size of the optimal voting threshold, though decision costs are not as likely to vary as frequently as external costs.
As it is difficult to determine the actual external costs of a policy a priori, legislatures often block different issue areas into a small number of categories reflecting the expected external costs. Issues of symbolic importance with little social cost usually can pass with a plurality rule (k < 50 percent). High-impact policies such as tax rules or constitutional changes generally require a supermajority (k > 50 percent +1), often in the range of a two-thirds or three-quarters threshold. Other policies in between usually default to a simple majoritarian rule.
Having to change the voting threshold for every issue would be amazingly time-consuming, adding to decision-making costs and require a priori knowledge of the potential external costs. Because of the practical difficulty in calculating the optimal voting threshold, the majority rule is often the default voting procedure, which makes it appear to be the “most democratic” rule. Senator Warren would agree.
Enter the Filibuster
If one considers Buchanan and Tullock’s discussion of decision and external costs, the Senate filibuster reveals itself as a rather ingenious mechanism to alter the level of the optimal voting threshold via increasing decision-making costs. The filibuster, in essence, allows any group of Senators to shift a simple majority vote to the status of supermajority by having the determination to speak on the issue for endless hours until a cloture vote (k = 60 percent) is taken to move the bill to a floor vote.
Any member of the Senate who felt the external costs of a bill were too high relative to a simple majority vote could shift the threshold up by calling for extended debate. Thus, even if most bills are slotted into a simple majority vote category for expediency’s sake, there is at least a procedural option to push that threshold upwards if conditions warrant.
Consider Figure 4. If a bill that is slotted into a simple majority voting rule is believed to impose very high external costs on a significant number of citizens, a group of Senators can filibuster so as the bill cannot be voted on until there is at least 60 percent support to end debate. Although it is difficult to precisely determine the external and decision costs, the filibuster allows Senators to push the voting threshold to what is probably the more optimal point (above a simple majority).
Can the filibuster rule be abused? Most certainly, and the practice itself has had a checkered past. But changes to the filibuster over time have helped to balance the need to shift voting thresholds for important legislation while tempering the possibility of abuse by ensuring that filibustering Senators have to pay a price for their desire to debate.
Informal, and sometimes theatrical rules have been employed historically to force Senators to speak continuously in order to keep the filibuster alive. This has led to rather humorous situations wherein Ted Cruz read a children’s book on the Senate floor and other politicians needed to get creative when nature called. (See here for a list of the longest filibusters.) In recent years, there have been threats of a “nuclear option” that would (and have in some instances) limit the use of the filibuster. All of these tactics are meant to keep the use of the filibuster costly so that it isn’t invoked recklessly.
As for Senator Warren’s concern that the filibuster is not in the Constitution, it should be noted that both Senators and Representatives are given great latitude in determining how to design and pass legislation. The Founding Fathers had significant foresight (perhaps accidental) to allow for flexibility within the legislative process, and the filibuster is just one example of what emerged.
If Senator Warren was really concerned about legislating based upon only what was allowed in the Constitution, she would need to reconsider the formal rules and informal norms surrounding the committee system, including the use of seniority and party affiliation for determining appointments. Those “not-in-the-Constitution” rules play a major role in determining what comes up for a vote and when. Oh, and we won’t even mention omnibus bills, logrolling, and all the feeding that goes on at the pork barrel.
Implications: Flexibility and Smallness Are Beautiful
Representative democracy invariably requires voting. While simple majority votes seem to be the “fairest” process, Buchanan and Tullock pointed out that there is no one single voting procedure that is best for all issues. As the external costs imposed upon losing minorities can vary, so too should the voting threshold under which those policies are decided. Alas, a different voting threshold for every issue creates too many transaction costs, but an important implication from Buchanan and Tullock’s reasoning is that there should be some degree of flexibility in voting procedures. The filibuster builds in some of that flexibility by acting as a “trigger warning” for issues that may be more salient than what a simple majority would recommend.
As noted above, the filibuster, albeit very useful, is not a perfect tool. And we have seen efforts over time to tinker with the procedure so as to create varying thresholds or impose significant costs upon Senators, so it isn’t used with reckless abandon. That is a good thing, because democracy is always about experimentation. And some of this experimentation occurs in state legislatures around the US, with varying rules needed to break a filibuster.
Eliminating the filibuster entirely would not be wise, as it would remove an important means of adjusting voting thresholds to the proper saliency of the issue at hand. This is important for all political parties given that constituencies and policy issues change over time; while one political group may not benefit from the filibuster on a current issue, they may find it a very valuable protection in the future. Warren be warned: You may not like the consequences of your campaign.
Finally, a less obvious implication of Buchanan and Tullock’s voting model is that small is beautiful. Buchanan went on to champion the notion of federalism, devolving decision-making to the lowest level possible. By keeping the relevant voting constituency small, it is more possible to reach unanimous consent because communication between the affected parties is easier to achieve (i.e., low decision costs).
More importantly, voting among a smaller constituency allows individuals to more directly “feel the pain” (costs) of their decisions. This engages the virtue of prudence. It is easy for a distant body of legislators to pass off diffuse costs to people who are not directly benefiting from a policy. Those diffuse costs can add up, as we see with the ever-expanding federal budget. On the other hand, when individuals more directly feel the costs of their own decisions, they will tend to think more carefully about those decisions. Prudent deliberation and voting can only be beneficial to a democratic polity.
To that end, if Senator Warren really cares about democracy, she would do well by leaving the filibuster in place and allowing more legislative decisions to devolve to the state, county, and municipal level of government where “the people” can really rule themselves.