Stanley Bach ‘Reaching Bicameral Legislative Agreement in Canberra and Washington*’

Papers on Parliament No. 53
June 2010

Whatever the virtues of bicameralism may be, there is no doubt that it can complicate the legislative process in any assembly, and especially in national assemblies that represent the diverse interests and preferences of complex societies. Under some democratic constitutions, a proposed new law cannot take effect with binding legal force until both halves of a bicameral parliament or legislature have approved it in precisely the same terms. This requirement for bicameral legislative agreement can cause delays, require difficult and sometimes acrimonious negotiations, and even prevent enactment of a bill that each house already has passed, albeit with somewhat different provisions.[1]

The potential difficulties of reaching bicameral agreements in any national (or sub-national) political assembly depend on at least five factors: constitutional powers, partisan control, party cohesiveness, procedural comparability, and legislative autonomy. Individually and collectively, these five factors shape and condition the legislative process, especially at that final stage at which the initial legislative decisions of the two houses must be reconciled.

Five factors affecting bicameral relations

Perhaps the most important of these factors is the first: the respective constitutional powers of the two houses regarding legislation. It certainly is the most durable in its consequences for national assemblies. Do the two houses enjoy roughly the same legislative powers? Are there constitutional arrangements governing the process for reaching bicameral legislative agreements that favour one house at the expense of the other? Typically, national constitutions assign considerably more legislative power, or give considerably more democratic legitimacy, to one house of a bicameral assembly.[2] And in such cases, one way in which constitutions can establish the dominance of one house over the other is by enabling it to impose its will in cases of legislative disagreements.

For example, a constitution may deny to one house the power to initiate or amend certain kinds of bills—most likely, essential financial legislation. It also may provide that, in cases of persistent bicameral legislative disagreement over the final terms of a bill, the preferences of one house are to prevail over the other.[3] Or it may achieve the same result somewhat more indirectly by submitting such legislative disagreements to majority votes of both houses sitting together, so that if each house is unified in defence of its position, the position of the larger house will prevail.

In most democratic regimes with bicameral national assemblies, one house clearly dominates the other. Australia and the United States are unusual in having national assemblies in which the two houses have relatively comparable powers, both houses are directly elected so they enjoy the democratic legitimacy without which they might be reluctant to exercise their powers (compared to Canada’s appointed Senate, for example), and neither house has the constitutional means to impose its will on the other in the regular course of business.[4]

The other four factors are more subject to change over time, just as they are more subject to influence by national assemblies and their members.

First, with regard to partisan control, is the distribution of party strength largely the same in the two houses? If the same party or coalition of parties controls a majority of seats in both houses, then, everything else being equal (which, of course, it rarely is), it should be easier for the two houses to reach agreement than it would be if there are different and opposing partisan majorities in the two houses, or if at least one of them is not controlled by a single party or stable coalition.[5]

Second, to what degree is the majority party or coalition in each house unified or disciplined? If it consists of a collection of uncomfortable bedfellows who have policy disagreements among themselves and who do not feel obliged to vote with their fellow party members, much less with members of any coalition partners in the assembly, then the two halves of the assembly may prefer significantly different versions of the same bill, even if they have the same ostensible majorities. This difficulty can arise in assemblies controlled by multi-party coalitions, but it also can occur in what formally are two-party systems if one or both parties is, itself, a coalition of diverse factions (whether or not they are organised and recognised as such) and those factions are not represented in similar proportions in the two houses.

Third, do the rules or standing orders or the procedural practices of each house enable the majority party or coalition to control its legislative outcomes?[6] Even if both houses have majorities of the same party or parties and even if the parties are unified or disciplined, it also matters if the legislative procedures of each house allow a simple majority of its members to control its decisions without undue delay. If so, they are more likely to reach similar or identical decisions than if the procedures of one house give the minority (or opposition parties) more leverage that it (or they) can use to extract concessions and compel compromises.

And fourth, does the same sense of legislative autonomy prevail in each house? Again, even if the two houses are controlled by the same cohesive party or coalition, it also matters if the legislative agendas of the two houses and the specific legislative proposals they each consider are decided elsewhere—namely, by the executive government. If so, the differences in the versions of legislation that each passes are likely to be less significant than if each house exercises more control over its legislative agenda and if each acts more autonomously in drafting the legislation that it then passes and sends to the other house for its concurrence.

This combination of endogenous and exogenous conditions helps both to shape and to explain how bicameral legislative agreements are reached, as a comparison of the Australian Parliament and the US Congress illustrates. The interplay among these five factors accounts largely for the major systemic difference between the final stages of the legislative process in these two assemblies.

Comparing Parliament and Congress

How do the Australian Parliament and the US Congress compare with respect to these factors? No brief summary can begin to do justice to such an encompassing question, but it can lay the groundwork for a comparison of how the two national assemblies try to reach legislative agreements.

With respect to constitutional powers affecting legislation, the Australian and US national assemblies are remarkably similar, notwithstanding the other differences in the two constitutional systems of which the assemblies are a part. This is no accident, of course. In designing the Parliament, the authors of Australia’s Commonwealth Constitution drew knowingly and deliberately on the American example, just as they drew on the British example for the parliamentary relationships between the executive government and the House of Representatives.

In Canberra as in Washington, most legislation can originate in either house. In both assemblies, most bills are passed first by the House of Representatives and then by the Senate, although this tendency is much more pronounced in the Australian Parliament than in the US Congress. Also in Canberra as in Washington, most taxing and spending bills are exceptions to the general rule: they can be introduced only in the House of Representatives. In Washington unlike Canberra, however, the US Senate is free to amend these bills once it receives them from the House of Representatives.

By contrast, section 53 of the Australian Constitution provides that the ‘Senate may not amend proposed laws imposing taxation, or proposed laws appropriating revenue or moneys for the ordinary annual services of the Government’, nor may it ‘amend any proposed law so as to increase any proposed charge or burden on the people’. Instead, the Senate may request that the House of Representatives make specific amendments that the Senate cannot make itself. No taxing or spending bill can become law until any Senate requests for amendments are disposed of to the Senate’s satisfaction, so it can be disputed whether this constitutional restriction on the Australian Senate’s powers is a matter of substance or primarily one of form and procedure.[7]

With legislative powers allocated so equally to the two chambers of both assemblies, the need to reach legislative agreements acceptable to both of them becomes an inevitable part of the legislative process. The US Constitution is entirely silent on this subject. By contrast, the Australian Constitution (in s.57) lays out a procedure for resolving legislative differences, but it is so difficult and time-consuming to use that it has been employed only once (in August 1974) since the Constitution came into force in 1901. This process requires the two houses to attempt, and fail, to reach an agreement on the final provisions of a bill, and to do so on three different occasions with a ‘double dissolution’ of both houses and an election of all members of both the House of Representatives and the Senate intervening between the second and third attempts. Then, if these requirements have been met, the executive government can request the Governor-General to convene a joint sitting of both houses at which decisions are made by a majority vote of all members of both houses voting together.

If such joint meetings were held regularly, they would enable the Australian House of Representatives to prevail over the Senate (if both houses were unified in support of their legislative positions) because there are twice as many members of the House of Representatives as there are senators. But precisely because these procedures are too impractical for regular use, and especially to reach agreement on legislation that requires prompt enactment, the constitutionally created ability of the House to prevail over the Senate has little practical effect. The government of the day has seemed happy to have one or more bills fail of passage twice so it can ask the Governor-General to declare a double dissolution when it wishes, but more for the purpose of capitalising on what may be a temporary political advantage or resolving a general political impasse than breaking a particular legislative deadlock.[8]

Consequently, the practical situation in Canberra is largely the same as it is in Washington: the actual procedures for reaching legislative agreement are to be found in the assemblies’ standing orders and in the procedures and practices that have grown up around them.

With respect to partisan control, the situation in Canberra and Washington also is more similar than it might appear at first glance. In Australia, the electorate is closely divided between its support of the Australian Labor Party (ALP) and the coalition of the Liberal and National parties (the Coalition). This durable fact of electoral life, combined with the different systems for electing members and senators, has, during recent decades, tended to produce a situation in which one party has a clear if not large majority in the House of Representatives, but lacks a similar majority in the Senate.[9]

Since 1981, and regardless of which party has been in power, the government has had a majority in the Senate only once, between mid-2005 and mid-2008. At all other times, the Senate has had a non-government majority. In 2009, for example, the ALP government held only 32 of the 76 Senate seats and the Coalition held 37.[10] For the government to win a Senate division against the Opposition, it needed the support of all the other seven senators (five Greens, one independent, and one representing the Family First Party).

In the United States, as in Australia, one house is much larger than the other—435 members of the US House of Representatives (hereafter representatives) compared with 100 senators—and senators serve for longer than representatives—six years for US and Australian senators compared with two years for US representatives and a maximum of three years for members of the Australian House of Representatives (hereafter members). But the two countries have adopted different electoral systems. In the US, representatives and senators are elected in the same way—by plurality elections in individual constituencies. In Australia, by contrast, one member is elected from each constituency whereas senators are elected by proportional representation on a statewide basis.[11]

Both Australians and Americans have become accustomed to a situation in which the party that holds a majority of seats in its House of Representatives does not also hold a majority of the seats in its Senate. However, this similarity masks an equally important difference. During Australia’s recent history, it has become expected that neither party will have the votes to control the Senate by itself. In the US, by contrast, it has not been unusual since the 1980s for the House of Representatives to have a majority of one party (Democrat or Republican) and for the Senate to have a majority of the other.[12] The modern US Congress recently has tended to oscillate between unified and divided control of its two houses.

In practice, however, it is unrealistic to say that either party ever actually controls both houses of the US Congress. The reason lies in Senate rules permitting filibusters that can be ended only by an affirmative vote of 60 of the 100 senators on a motion to invoke cloture. Because neither party in the Senate often has 60 votes of its own, a unified Senate minority party (whether it is the government or the Opposition party in parliamentary terms) usually can prevent the Senate from passing almost any bill.[13] If one party has a majority in the US House of Representatives but is in the minority in the Senate, the situation in Washington is not much different from what usually prevails in Canberra. And even if both houses in Washington have the same majority party, that party still must negotiate with the minority party in the US Senate to attract the votes of some of its Senate members because those votes usually are needed to forestall or end a filibuster.

With respect to party cohesiveness, the situations in the two assemblies again are more similar in practice than they once were. One of the defining characteristics of Australian national politics is the disciplined voting that characterises both Labor and Coalition members and senators. Since its first days in Parliament, Labor has required its members in each house to vote in accordance with the party’s positions, and in self-defence if for no other reason, voting cohesion within the Coalition parliamentary parties is just about as perfect. It is newsworthy when an Australian member or senator crosses the floor on a division. The only exceptions to this pattern tend to be on the handful of so-called ‘conscience votes’, on matters such as abortion and euthanasia, on which the parties allow their members to vote as they please (although they still tend to vote with their party colleagues).[14]

In the Australian House of Representatives, therefore, the government can expect to win every division, even if its majority is a slim one. And in the Senate, the government and Opposition each look to minor party and independent senators with whom to form majority coalitions. Rarely is there any serious prospect for either major party to attract one or more Senate votes from the other.[15] To be sure, there are policy disagreements over legislation within Labor and within the Coalition, but these differences typically are resolved in the party rooms before the bill at issue comes before the House or Senate for formal consideration and public debate.[16] When the two parties face each other and the Australian public, they stand united.[17]

In Washington, the conventional wisdom for decades was that party discipline in Congress was very weak. That was true, and remains true, in the sense that national party organisations have few means to impose discipline by penalising representatives and senators who oppose the party position, if one can be identified, on specific issues. What has changed in recent decades, however, has been the degree of voluntary party cohesiveness in congressional voting.

Into the 1970s, there was a distinctive cohort of so-called liberal Republicans in Congress, who were noticeably to the left, on a simple left–right continuum, of most of their fellow partisans in the US House and Senate. Even more important was the split among congressional Democrats between a majority of generally liberal representatives and senators and a sizeable minority of institutionally powerful Southern Democrats who, in policy terms, often had more in common with their Republican counterparts than with most congressional Democrats. The result was that, on most of the most important rollcall votes, a minority of Republicans typically would vote with a majority of Democrats, and a minority of Democrats would vote with most Republicans. If the Democrats held only a narrow majority in the House, the Republicans could prevail anyway if, as sometimes happened, there were more conservative Democrats to vote with the Republicans than there were liberal Republicans to vote with the Democrats.

Today, however, far fewer representatives and senators of one party vote with members of the other. The so-called liberal Republicans have almost entirely disappeared, and the once solidly Democratic South is now the bastion of the Republican Party in Congress. The Southern states now elect more Republicans than Democrats to both the House of Representatives and the Senate. As one US senator is reported to have put it, clearly if a bit too simply, ‘Today, most Democrats are far left; most Republicans are to the right; and there are very few in between’.[18] Party unity in US congressional voting is not nearly as perfect as it is in Canberra—if it were, it normally would be impossible to break a Senate filibuster—but voting patterns in the two assemblies now are much more similar than they once were, regardless of whether those patterns are attributable to enforceable discipline or voluntary cohesiveness.

Still, the difference between legislative voting patterns in Canberra and Washington, though reduced, remains significant. Once a bill becomes subject to votes in the Australian Parliament, any necessary intra-party adjustments probably will already have been made. The government knows that it can win in the House of Representatives, so it can devote most of its efforts to attracting the increment of non-government votes that it needs to prevail in the Senate as well. In Washington, majority party and committee leaders try to ensure that the bills they bring to the chamber will satisfy their fellow party members, but they must rely primarily on persuasion, compromises, and concessions to prevent defections on rollcall votes. Increasingly in recent years, the majority party’s leaders in the US House have relied heavily, and sometimes only, on their own party members to form voting majorities, as each party has become more homogenous and as the two parties have become more polarised. In the US Senate, on the other hand, the ever-present threat of a filibuster gives a bill’s majority party supporters a powerful incentive to ensure that it is acceptable to at least some minority party senators.

Therefore, with regard to the third factor, procedural comparability, there is an important difference between the US House of Representatives and the US Senate as well as between the US Congress and the Australian Parliament. In Canberra, the legislative processes in the two houses are quite similar. In both houses, the critical votes on second and third reading, as well as intervening votes on amendments and other questions, all are decided by simple majority votes, with almost all members and senators participating in divisions when they are called. So too in Washington, where the votes in the House of Representatives and the Senate on final passage of legislation (the effective equivalent of a vote on third reading of a bill in Canberra) also are decided by simple majorities.[19]

It has become commonplace in Washington, however, to explain that it will take 60 votes—a three-fifths majority—for the Senate to pass a certain bill. Such a statement really is a shorthand way of saying that, although the Senate can pass the bill by a simple majority—51 or more—it will require 60 votes to allow the Senate to reach the point at which it can vote to pass the bill (or not). One reason is the prospect of filibusters, discussed above, which can continue indefinitely unless terminated by a three-fifths vote in favour of a motion to impose cloture.[20] A second reason is a series of requirements and prohibitions that affect budget-related legislation and that the Senate can waive in individual circumstances, but only by a three-fifths vote.[21] These three-fifths waiver requirements were imposed to make it more difficult for the Senate to impose constraints on itself in principle but then to circumvent them easily whenever tempted to do so.

The effect of these three-fifths voting requirements has been to create a different procedural dynamic in the US Senate than prevails in the House of Representatives. In the House, it is necessary to construct only a minimal winning coalition—one more than half of the representatives present and voting. And if this majority can be found solely within the ranks of the majority party, there is no need and not much incentive (and, today, not much likelihood of success) for efforts to attract more than a smattering of votes from minority party members. In the Senate, by contrast, there is a powerful incentive to attract some minority party support for legislation because the majority party in the contemporary Congress rarely can supply 60 votes of its own. If a bill runs afoul of the congressional budget process, those 60 votes are needed to overcome the procedural barrier it imposes. But even absent such a problem, the same 60 votes are required more and more often to break filibusters or to prevent them from beginning in the first place. Consequently, the incentives for inter-party compromises on legislation are much stronger in the US Senate than in the House of Representatives.

With regard to the fifth factor, legislative autonomy, there are three obvious and important differences between the situations in Canberra and Washington.

First, in Canberra, almost all legislation originates with the executive government, and the government controls the legislative agenda in both the House of Representatives and the Senate. The standing orders of both houses are designed for this purpose. A government bill may be amended, even in important respects, in one or both chambers, but it remains the government’s bill. The essential question before the Parliament is how that government bill might be modified during the course of the legislative process. To be sure, there are private members’ and private senators’ bills, but they are relatively few in number and even fewer are enacted.

In Washington, by contrast, all bills are private members’ and senators’ bills, although they are not called that because there are no government bills as such. The president and his Cabinet secretaries (ministers) may send draft bills to Congress, but they must be taken up and introduced by representatives or senators before they can be considered. Even then, it is the rare bill of any importance, whether it has its origins in the executive branch or in the imagination of a member of Congress, that is not substantially revised or even radically rewritten before it even reaches either chamber for consideration. US congressional committees originate much (perhaps most) of the most important legislation, either by writing new bills or by approving amendments that replace the entire texts of bills referred to them. And while the House of Representatives is considering a bill on a subject of importance, the Senate committee with responsibility for the same subject may be drafting its own bill that addresses the same policy issue in an entirely different way. As we shall see, this has consequences for how the two houses later try to reach their legislative agreements.

Second, in Canberra, the Senate does not request specific amendments to the basic appropriation bills that fund ‘the ordinary annual services of the Government’, even though it is constitutionally free to do so. There probably are at least two connected reasons for this restraint. First, these bills reflect the government’s priorities for the coming financial year, and so they go to the very heart of the programs and priorities of the party that the Australian people elected. Senate requests for amendments to these bills might be thought to challenge the government’s ability to implement its campaign commitments. And second, for these reasons, any contemporary government is likely to reject any such requests for amendments with indignation and with accusations that the Senate is trying to delay or block ‘supply’, a partisan stratagem widely discredited by the 1975 imbroglio in which the non-government majority in the Senate declined to act on essential funding legislation and convinced the Governor-General to dismiss the Labor Party government of Prime Minister Whitlam, even though it still enjoyed majority support in the House of Representatives.

In Washington, on the other hand, the presentation in February of the president’s spending plans for the next fiscal year that will begin in October only initiates a process of intense and tortuous negotiation that usually is not completed these days until well after the putative deadline of 1 October. When President Obama took office in January 2009, one of his first legislative priorities, amidst the greatest economic crisis in 70 years, had to be to negotiate with Congress for legislation that would complete funding for the financial year that already had been in progress for four months. Such a situation would be just as inconceivable in Canberra as would the possibility of basic annual budgetary legislation sailing through the US Congress, unchallenged and unchanged.

And third, in Washington, the House of Representatives and the Senate see themselves as competitors for influence and pre-eminence. Since the very early decades of the American republic, there have been periods in which the House has been more visible and influential than the Senate, and other periods in which the reverse has been true. I have heard it said, presumably in jest of course, that large majorities in both houses would prefer a unicameral Congress, but they could never agree on which house to abolish. Certainly, neither house has thought it fitting to defer to the other for constitutional reasons, and each tends to react swiftly to any perceived intrusion by the other into its constitutional powers and prerogatives. The Senate has never doubted its legitimacy as a legislative body, even before senators began to be directly elected in 1914. Representatives leave the House to run for the Senate, and senators make much more credible presidential candidates than do representatives. If anything, senators consider the Senate to be the more prestigious and important of the two halves of Congress. What else would we expect from an institution that has not blushed to call itself ‘the world’s greatest deliberative body’?

One of the most influential of the early academic studies of the modern US Senate was Donald R. Matthews’ U.S. Senators and Their World.[22] In his best known chapter, on ‘The Folkways of the Senate’, Matthews wrote that ‘Senators are expected to believe that they belong to the greatest legislative and deliberative body in the world. They are expected to be a bit suspicious of the President and the bureaucrats and just a little disdainful of the House. They are expected to revere the Senate’s personnel, organization, and folkways and to champion them to the outside world’.[23] Such ‘institutional patriotism’ as Matthews called it, clearly has diminished since his book first appeared, but still the best way to bring senators together across party lines is to disparage the Senate’s powers or challenge its prerogatives.

I don’t think it can be said that the same situation prevails in Canberra. Historically, both major parties and the governments they have formed have been sceptical or critical of the Senate when it has hampered or blocked their legislation. For years the Labor Party called for the abolition of the Senate; one of its recent prime ministers contemptuously dismissed senators as ‘unrepresentative swill’. Recent Coalition governments also have been less than enthusiastic about having to confront important Senate amendments to their key legislation (until they lost control of government, of course, after which they discovered new-found wisdom and value in an assertive Senate). Ever since Federation, some Australians have viewed the Senate as an awkward encumbrance on what always was intended to be a responsible parliamentary government emulating the British model. Indeed, this attitude can be found expressed in writing on the subject by political scientists. L.F. Crisp, who undoubtedly was one of the most influential writers of the late 20th century on Australian government, wrote as recently as 1983 that: [24]

A Senate for whose election the Constitution and the laws combine to provide conditions likely to produce at least party stalemate in that Chamber, and even a party majority hostile to, and empowered to frustrate, the Government of the day, is in the 1970s not simply an anomaly but, ultimately, a threat to the essential status of responsible government in the eyes of citizens.

From this point of view, the Senate is very much the second—the secondary—house of Parliament, and it should behave accordingly. The government is formed primarily in the House of Representatives and is responsible only to the House of Representatives. While Crisp’s conclusion is debatable, what cannot be debated is that individual ambition usually takes Australian politicians from the Senate to the House of Representatives, which is the reverse of the pattern in the United States. As an institution, the Australian Senate does not have the self-esteem and sense of self-importance that characterises the US Senate. All US senators would agree that the Senate is at least as important, powerful and valuable as the House of Representatives, and many probably would argue that the Senate really is the pre-eminent legislative body in Washington. It is hard to imagine such contentions being heard from either of the major parties in Canberra.[25]

In short, when the Australian House of Representatives and Senate try to reach legislative agreements, they do not approach the task with the shared understanding that each has an equal right to prevail. The reason lies not in the Senate’s constitutional weakness but in the opinion of governments and most members of the House of Representatives that it is the House—which, after all, is the agent of the executive government—that has the greater legislative legitimacy and, therefore, the stronger claim to prevail.

Paths to legislative agreement

These factors provide a context that helps to explain how the Australian Parliament and the US Congress go about trying to reach bicameral legislative agreements. As noted earlier, the need to do so arises because, under both national constitutions, both the House of Representatives and the Senate must pass a bill and agree completely on its text before it is eligible to become law.

These requirements mean that a bill that originates in one house must also be passed by the other: i.e., the Senate must pass a bill that the House of Representatives already has passed and then sent to the Senate, or conversely. As also noted earlier, the great majority of Australian bills originate in the House of Representatives, and this has been the tendency in Washington as well, although to a lesser extent. If either House of Representatives passes a bill and transmits it to the Senate, and the Senate then passes it without change, bicameral agreement has been reached and the bill is ready to become law.[26] If, however, the Senate amends the bill from the House of Representatives before approving it—by passing it (in the US) or by agreeing that the bill be read a third time (in Australia)—the two houses then must agree on how to dispose of the Senate’s amendments.[27]

Simply put, there are two ways in which the two houses of a bicameral assembly can act on amendments to a bill in order to reach legislative agreement: the two houses can act on the amendments one at a time or they can act on them all at once. In both Canberra and Washington, the standing orders of both houses, as well as other enforceable procedures that have not been formally incorporated into their adopted rules, include what sometimes are very complicated procedures by which they can act on the amendments individually, considering and disposing of one amendment and then taking similar action on each of the other amendments in turn. In both assemblies, this is the default option: they follow this individual approach as they try to reach a bicameral agreement unless they act affirmatively to do it differently. And both the Australian Parliament and the US Congress can do it differently, through a collective approach that involves the use of a conference in Canberra or a conference committee in Washington.[28]

The difference between the individual and collective approaches can be more formal than real. The individual approach suggests that the two houses treat each amendment in isolation from the others, disposing of it on its own merits and without regard to how they already have disposed of the other amendments or how they expect to dispose of them. The collective approach suggests, on the other hand, that the two houses act on each amendment in the context of all the others, so that how they dispose of one amendment very well may be affected by the agreements they have made or expect to make regarding the others. In practice, however, the difference between the two approaches is not nearly so clear. Just because the individual approach calls for the two houses to dispose of one amendment before taking up the next one, this does not mean that the members of both houses have not already discussed all of them, formally or informally. Indeed, they may very well have reached understandings—understandings that members are loathe to violate even if they are not enforceable under the rules of either house—as to what action they intend to take with regard to each of the amendments.[29]

Such behaviours reduce the practical differences between the two individual and collective approaches to reaching legislative agreements. In addition, and as I shall discuss later on, there have been recent procedural developments in Washington that further erode whatever practical differences remain in Washington between the two approaches. The next section of this essay discusses the nature and advantages of conference committees. The section that follows explores why conferences are not used in Canberra, even though the standing orders of both houses allow for them. Then, in the section after that, I muddy the waters by discussing recent procedural developments in the US Congress that have had the effect of reducing the clarity of the distinction between the two approaches in the US legislative process.

Conference committees and conferences

The most striking and important difference in how the US Congress and the Australian Parliament reach bicameral agreements emerges from an inquiry into how they do, or do not, use the collective approach to resolving their legislative differences. Therefore, the remainder of this essay focuses primarily on how this approach is, or is not, used to complete the legislative process in Canberra and Washington.[30]

In Washington, the process begins when one house approves amendments to a bill that the other house already had passed. For ease of explanation, assume that the House of Representatives passes a bill and the Senate then amends that bill before passing it as well. The Senate next returns the bill to the House, together with the amendments that it has approved, and asks the House of Representatives to concur in those amendments. If the House does so, the two houses have reached agreement and the legislative process is complete. If, however, the House does not do so—if the House simply disagrees to the Senate’s amendments or if it approves its own amendments to the Senate amendments—then the bill and amendments are returned to the Senate, which is expected to take further action of its own. This process of exchanging amendments and positions can continue through several more rounds and become very complicated.

At any stage during this process, however, either house of the US Congress has the option of proposing that the two houses appoint members to a conference committee. This is a temporary joint committee established solely for the purpose of considering all the Senate’s amendments to the bill in question and presenting a single report to both houses that recommends how to dispose of each and all of them. The House of Representatives or the Senate then debates and votes on this report, either accepting it or rejecting it in its entirety (or, rarely, returning it to the conference committee for revision).[31] If one house agrees to the conference report, it is sent to the other house which also debates it and may either accept or reject it.[32]

A conference report in Washington is a kind of bicameral treaty. The negotiators from the House of Representatives and the Senate propose in their report a package settlement of all the differences in the versions of a bill that the two houses originally had passed. It is virtually certain that, during the conference negotiations, there are trade-offs in which, for example, the Senate’s negotiators may agree not to press for agreement to one of its amendments in exchange for an agreement by the House’s negotiators to accept a change that the Senate also proposed somewhere else in the bill. The conference committee’s report incorporates all the formal, legislative agreements that the negotiators (the ‘conferees’) reached with respect to all the Senate’s amendments to the House’s bill.[33] The conferees are likely to agree to accept some Senate amendments, to reject others, and to agree to alternatives to still others.

When their treaty then reaches the floor of the House and Senate, it may not be picked apart by amendments. If a majority of the members in either house simply cannot stomach one of the agreements that the conference committee reached, the only recourse those members may have is to reject the report in its entirety.[34] They then may propose that a new conference committee be created to negotiate a different treaty (or, as noted above, they may be able to return it to the original conference committee for the same purpose). Otherwise, the bill dies.[35]

There is no reference in the US Constitution to conference committees, but they are covered by the standing rules of both houses of Congress. These rules govern the creation of conference committees, the appointment of their members, their authority and meetings, the permissible content of their reports, and the procedures by which each house debates and votes on their reports. There is one subject, however, on which the rules of procedure of the House and Senate are silent: how the conferees are to conduct their meetings and reach their agreements.[36] The reason is that conference committees are negotiating forums, so both houses leave it to the conferees to decide for themselves how best to proceed, recognising that the most effective way to reach agreement is not going to be the same for all bills.

Enactment of most national laws in the US has not required the appointment of conference committees. During the eight years from 1999 through 2006, more than 1900 public laws were enacted in Washington.[37] Only about seven per cent of them went through a conference committee. Almost 80 per cent of the total were passed initially by both the House and the Senate with exactly the same wording, so there was no bicameral agreement that had to be reached. In about twelve per cent of the cases, the first house that passed bills simply accepted the amendments to those bills that it received from the other house. The residue—1.5 per cent—were readied for enactment by use of the individual approach to reaching agreement that I defined above.[38]

The usual practice of the US Congress has been to create a conference committee only when bicameral agreement could not be reached quickly and easily. More often than not, a conference committee has not been necessary. But the final versions of most of the most important and contentious laws that the modern Congress has passed were written in conference committees.[39]

Some of the advantages of conference committees are obvious. First, they concentrate responsibility in the hands of a small proportion of all representatives and senators, and it always is easier for small groups to negotiate successfully than the 435 members who comprise the House or even the 100 members of the Senate. Second, conferees can negotiate quietly and privately, notwithstanding the requirement that their formal meetings be open to the public. Third, conferees can conduct their business, whether in public or in private, without being constrained by all the formal procedures and rules of debate that govern the conduct of plenary sessions of the House and Senate.[40] Fourth, conferees can meet while the other members of both houses are meeting at the same time to conduct other business.

Finally, the use of conference committees tends to encourage enactment of bills by presenting both houses with what typically are described by their proponents as the best final versions that could be negotiated of bills that majorities in the House and Senate already have passed. Members are warned, and often with good reason, that if they defeat the conference report on a bill, the bill itself is very likely to die.[41] Faced with the choice between accepting the recommendations of the conferees, however imperfect those recommendations may be, and leaving the statutory status quo unchanged, the members who supported the bill when they considered it initially are likely to prefer the conference report. So at first blush and for all these reasons, it would seem surprising that the use of this collective approach to reaching bicameral legislative agreements is essentially unknown in Canberra.

It is true that the standing orders of both Australian houses provide for conferences, and govern, in very similar terms, how conferences are to be requested, arranged, and convened.[42] However, conferences are intended to be a last resort, and primarily an alternative to laying a bill aside (that is, allowing it to die for lack of bicameral agreement).[43] The standing orders of the House of Representatives mention the possibility of a conference as an option only at the last stages of the process of trying to reach bicameral agreement regarding a House bill that the Senate has amended. Similarly, the standing orders of the Senate raise the possibility of a conference on a Senate bill that the House has amended only after the individual approach to reaching agreement has not worked.[44] The Senate may request a conference after the House has amended a Senate bill, but only ‘when agreement cannot be achieved, by an exchange of messages, with respect to amendments to Senate bills’.[45]

In Washington, by contrast, the collective approach to reaching agreement (that is, recourse to a conference committee) has not been an alternative that arises when the individual approach has failed. Instead, for the US Congress, the collective approach has been essentially an alternative to the exchange of positions and proposals between the two houses.

In part because conferences in Canberra are essentially a last resort, when the difficulty of reaching bicameral agreement on a bill already has been demonstrated, and in part for the other reasons discussed below, conferences are not a regular and familiar element of the procedural repertoire of the Australian Parliament. In fact, there have been only two formal conferences since Federation.[46]

In 1930, the House requested a conference after the Senate had insisted on its amendments to the Commonwealth Conciliation and Arbitration Bill 1930.[47] Each house appointed five managers who agreed to propose that the House should agree to some of the Senate amendments, that it should not agree to others, and that the House should agree to still other Senate amendments with modifications. Both houses agreed to these recommendations. In the following year, a conference on the Northern Territory (Administration) Bill 1931 was arranged and held in the same way. But during the almost 80 years that followed, no other formal conferences have been held.[48] The House of Representatives’ own guide to its procedures explains that, ‘in practice the conference procedure is not used, and if it is recognised that further negotiation by message would be pointless it is usual for the House to order the bill to be ‘laid aside’—that is, abandoned and removed from the Notice Paper’.[49]

The absence of conferences in Canberra

It is typically harder to account for the absence of something than for its presence—harder to explain why there are no conferences in Canberra than why there are conferences in Washington. That said, there would seem to be at least six reasons—reasons that reflect the factors I discussed in the first section of this analysis—why the collective approach to reaching legislative agreements has been used regularly in Washington but essentially never in Canberra.[50]

When a conference committee meets to negotiate in Washington, the agreements it reaches must be supported by a majority of the representatives appointed to the committee as well as by a majority of the committee’s Senate members. It never makes sense to take a vote among all the conference committee members; instead, votes are taken within the House delegation and within the Senate delegation. For this reason, it has never mattered how many representatives and how many senators are appointed to the same conference committee. Three Senate members have the same voting power as 33 (or any other number of) representatives; unless at least two of the three senators agree with a majority of the 33 representatives, there is no agreement.

As this voting rule suggests, a conference committee, in principle at least, is a forum in which representatives of both parties negotiate with senators of both parties. The delegation from each house is supposed to be defending the version of the bill that its house previously had passed; the negotiation is supposed to be between the two houses, not between the two parties. Not surprisingly, the reality has been somewhat different and, as I shall discuss shortly, the increased partisan polarisation in Congress during recent years has affected how often conference committees are appointed and how they work.

Still, the principle remains that the two houses of the U.S. Congress are expected to be negotiating with each other in conference committees. When conferees reached an agreement and the two houses then debated the merits of the conferees’ report, the members (especially the leaders) of each conference delegation traditionally were anxious to argue that the conference agreement resembled their house’s original version of the bill at issue more closely than it resembled the version brought to the conference committee by the other house. Political scientists have attempted to answer the question, ‘Who wins in conference more often, the House of Representatives or the Senate?’[51] There are good reasons to think that this question has no answer, but the fact that the question is asked is evidence that conference committees traditionally have been understood to be a setting for a competition between the two houses.[52] Will the final version of a bill reflect the collective preferences of the Senate more closely than those of the House, or will the House be recognised as having prevailed in conference over the Senate?[53]

The nature of US congressional conference committees as forums for competition between the two houses reflect three of the five factors I discussed at the outset: first, the essentially comparable constitutional powers over legislation that the two houses enjoy; second, the autonomy of each house from external direction as it makes its legislative decisions; and third, the strong sense in each house that its legislative preferences should carry at least as much weight as those of the other.

As I argued above, the situations in Canberra and Washington are quite similar with respect to the first of these factors. Under each national constitution, the legislative powers of the two houses are much the same. The House of Representatives and the Senate in each capital have the same powers over legislation except for certain money bills, and whereas the US Senate can amend these bills while the Australian Senate cannot, a money bill cannot become law in Australia unless the House of Representatives satisfies all the Senate’s requests for amendments or the Senate decides not to press its requests.[54]

With regard to the other two factors, however, the situation in Canberra is quite different than in Washington. While the Australian House and Senate each could appoint some of its members to meet and negotiate with counterparts from the other house, all concerned always would understand that the House of Representatives would be acting as a reliable agent of the government. If the two houses were to agree to convene a conference, it would ostensibly be a setting in which the Senate would negotiate with the House, but in practice, the Senate would be negotiating with the government. The members participating in a conference could be expected only to reach agreements with the Senate that the government already had signalled its willingness to accept.

In Washington, even when the president’s party has majorities in both houses, the versions of any major bill that each house passes is likely to differ in some respects from the president’s preferences. Members of a conference committee who also are members of the president’s political party are very likely to consult with him and his subordinates, but they are not there simply to do the president’s bidding. In Canberra, on the other hand, once any necessary intra-party negotiations in the party room have taken place, the government and its majority in the House of Representatives almost always speak with one voice, and that is likely to be the voice of the prime minister or his agent.[55]

So one reason why the collective approach fits the legislative process in Washington better than it does in Canberra is that when there are legislative disagreements to be resolved in Canberra, the actual parties to negotiations usually are representatives of the government, on the one hand, and representatives of the non-government majority in the Senate, on the other. Negotiations in a conference between members and senators in Canberra would, to a large degree, mask the part played by the government and exaggerate the autonomous part played by the participating members. In Washington, by contrast, while the president may be an interested observer or even a de facto participant in conference negotiations, the House and Senate members of a conference committee are autonomous actors who have interests and preferences of their own and who are willing to press for their satisfaction.

A second reason is closely related to the first. The regular use of conferences in Canberra would require the House of Representatives or the government (or both) to accept the Senate as an equal partner in writing new legislation, and that is not likely to happen. Notwithstanding the very similar legislative powers that the two houses enjoy under Australia’s Constitution, the opinion of L.F. Crisp, quoted above, that a Senate with a non-government majority is ‘a threat to the essential status of responsible government’ may still be widely held among Australian politicians, and especially among politicians of the party that happens to be in government at any given time. A government is elected with a mandate to enact its legislative program (or so the argument goes) and, for this purpose, it can and should depend on a reliable majority in the House of Representatives. If and when a non-government majority attempts to block or even demand major changes in government legislation, it thereby challenges the essential logic of responsible government by which the people elect the House of Representatives, which chooses the government, which decides on its legislative program, which the Parliament enacts, and on the basis of which the people decide whether to re-elect the same House majority and government at the next election.

If this logic is taken as a starting point, then the Australian Senate, especially when it has a non-government majority, is not supposed to play a part equal to that of the House in deciding on the content of new laws, and the government certainly should not be expected to negotiate with the Senate’s non-government majority as if it had equal constitutional claims to legislative influence or equal political claims to popular support for its legislative preferences. In short, the government, acting through the House, should prevail in the legislative process as a matter of constitutional right. Needless to say, the situation is very different in Washington, where the US Senate has absolutely no doubt that it is an equal partner with the House in writing law and, indeed, to some senators, it is unquestionably the senior partner.

The flip side of this coin, so to speak, reveals a third and distinguishable reason why the US Congress utilises conference committees and the Australian Parliament does not. Not only do the Australian Government and its majority in the House of Representatives have no doubt about their right and mandate to enact their legislative program, the Senate does not always press its own claim to equal partnership in the legislative process. If the Senate were more assertive, it could compel the House and the government to treat with it as a co-equal half of the legislative branch, however reluctant and unhappy they might be to do so. As we shall see, there is evidence suggesting that the Senate often has been willing to give way when the House objects to its legislative decisions. If so, why should the House agree to establish legislative conferences that, almost by definition, give the Senate equal standing when the Senate has not consistently pressed its own claim to be the constitutional equal of the House?

In short, conferences or conference committees make eminently good sense as a negotiating forum between equals. So they are appropriate for the US Congress, in which the Senate has no doubt that it is the legislative equal of the House and, in the view of many, the seat of broader perspectives and sounder judgements. On the other hand, conferences are inappropriate for the Australian Parliament so long as the House of Representatives asserts its legislative primacy and the Senate does not consistently challenge this assertion.

Another pair of reasons for the differences between the Australian Parliament and the US Congress in their use of the collective approach to reaching legislative agreements is related to their internal organisation and the preliminary stages of their legislative processes. Simply put, these reasons are reflected in the fact that the standing orders of both houses in Canberra provide for appointing conferences while the standing rules of both houses in Washington provide for creating conference committees.

Committees are at the heart of the legislative process in both the US House of Representatives and the US Senate. As long ago as the 1880s, Woodrow Wilson, later to be better known as a US president than as a political scientist, famously wrote that ‘it is not far from the truth to say that Congress in session is Congress on public exhibition, whilst Congress in its committee-rooms is Congress at work’.[56] This was an exaggeration then as it is now, but Wilson’s emphasis on the importance of congressional committees was not misplaced.

Most bills introduced in the US House of Representatives are referred to a committee (or sometimes more than one committee) for it to study and then to recommend whether the House should pass each bill.[57] Almost invariably, the committee’s written report on any significant bill includes one or more amendments that the committee proposes that the House make in it. Much the same situation prevails in the US Senate. In Canberra, on the other hand, while the Australian Senate is rightly proud of its committee system, the fact remains that, between 1990 and 2007, roughly 70 per cent of all bills introduced in the Senate or received from the House of Representatives were not referred to a committee.[58] Furthermore, amendments to a bill recommended by a US congressional committee receive priority consideration when the House or Senate takes up that bill in plenary session. Amendments recommended by an Australian Senate committee enjoy no such preferential status. So it may be said that, in the Senate’s consideration of all the legislation it passes, committees play a valuable but not a pivotal part. And in the Australian House of Representatives, committees have little legislative role at all.

Not only do committees in Washington review most bills before the House or Senate considers them in plenary sessions, the texts of the bills that the House or Senate passes frequently originate in the committees of the US Congress. House and Senate committees routinely conduct public hearings to hear testimony on the merits and contents of a bill referred to one of them, and the committee then typically conducts one or more ‘markup’ meetings at which committee members propose, debate, and vote on amendments to the bill, amendments that the committee proposes to recommend to its parent body (that is, either the Senate or the House of Representatives). And these amendments frequently are sweeping in nature. In Washington, every committee with legislative responsibilities regularly makes written reports on bills that the House or Senate had referred to it with recommendations for amendments that propose to change those bills beyond recognition.

Committee members are expected to be, and usually are, the legislative experts and specialists in the US Congress. No wonder, then, that when a bill is brought before the House or the Senate for its consideration, the debate is almost always initiated by members of the committee that had studied and reported on it. The chairman and the senior member of the minority party lead the debate for their respective parties, explain the bill and why it should be passed or defeated, and take the lead in proposing amendments and in reacting to amendments proposed by other members.[59] The importance that the House and Senate attach to the recommendations of their committees is reflected in the fact, mentioned above, that the first amendments that either house debates when considering a bill are any amendments that the committee which reported the bill had recommended. Almost always, the House or Senate debate on a bill focuses on what its committee has recommended, not on the bill as it was originally introduced, even if it had been suggested by the president or some other executive branch official. As an old saying goes, the president proposes, but the Congress disposes.

After each house has passed its own version of a bill and the need arises to reach agreement on how to reconcile these differing versions, it is only natural that the House and Senate should again turn to the members of its standing committees to take the lead in this process. These, after all, are the members who first studied the bill, who evaluated it and proposed amendments to it, who led the House or Senate debates on it, and who, by virtue of their expertise and specialised knowledge that often derives from having spent many years serving on the same committee, are assumed to know more about each bill than almost any other Representative or senator.

For many years, the membership of a conference committee on a bill would be drawn exclusively from the members of the House and Senate standing committees who had nurtured, shaped, and guided the bill to passage in their house. In fact, the delegation appointed by each house to a conference committee typically consisted of three members: the two most senior members of the majority party on the committee (including the chairman) and the committee’s most senior minority party member. That practice has given way to larger conference committees, a trend that reflects a dispersal of power that has taken place within Congress since the 1970s. Yet most conference committees continue to be dominated by members of the House and Senate committees who first had worked on the same bill in their separate committees and chambers, and who now meet together—not only as representatives of their house but as its policy experts—to reach a final agreement on the content of a bill as it approaches enactment.

To a considerable extent, therefore, (though less so than in earlier times) the use of conference committees in Washington is a natural extension of the deference that representatives and senators give to the knowledge and power that reside within their standing committees. Since it is the members of those committees who typically write the texts of the bills that the two houses debate in plenary sessions, who better to work on those bills once again and to propose final legislative agreements on them to the Senate and the House?

The situation in Canberra, of course, is quite different. Committees of the House of Representatives focus on non-legislative inquiries, or at least on inquiries that do not focus on the strengths and weaknesses—and the details—of individual bills that the government has proposed. In the Senate, and notwithstanding the much greater legislative role of Senate committees, those committees may consider only the bills that the Senate has voted to refer to them. The committees’ members do not necessarily dominate the debate when one of their bills comes before the Senate, a committee’s recommendations for changing the bill typically are not drafted as specific amendments, and any specific amendments a committee may have recommended to a bill do not receive priority attention when the Senate entertains amendments to the bill. Also, there is not nearly the same durability and stability in Senate committee memberships that there are in congressional committees in Washington, where representatives and senators sometimes chaired the same committees for decades until both houses began experimenting in the 1990s with term limits for committee chairmen.

Thus, it seems likely that the differences in the centrality of committees to the legislative process in Canberra and Washington are another significant reason why the collective approach to reaching legislative agreements has been relied on so much more in one capital than in the other.

There also is a fifth, procedural, reason why conferences are less common in Canberra than conference committees are in Washington: conferences in the Australian Parliament would be much less likely to enable the two houses to reach bicameral agreement on legislation.

As I discussed earlier, the report of a congressional conference committee is a non-amendable package. Most representatives and senators are inclined to accept the argument that the only alternative to approving a conference report is to witness the death of a bill that majorities in both houses already have voted to pass. And if both houses agree to a conference report, the effect is to complete the legislative process because all legislative disagreements with respect to that bill have been resolved.

In Canberra, by contrast, the report of a conference, were one to be created today, would present only a set of recommendations that are subject to further legislative action, including amendment by the House or the Senate. ‘The adoption of a report of a conference does not necessarily bind the Senate to the proposals of the conference, which, with reference to amendments in a bill, come up for consideration in committee of the whole’[60]—and it is during consideration of bills in committee of the whole that senators may offer amendments to them. So there would be no compelling reason to assume that a conference on legislation in Canberra actually would lead to a settlement of legislative disagreements.

Finally, there is another, more technical, reason why the collective approach is more appropriate in Washington than it is in Canberra, and this reason too derives from two procedural consequences of the work and influence of US congressional committees.

First, when a US House or Senate committee has completed its public hearings on a bill (or several bills on the same subject) and then convenes in the first of its markup meetings, the legislative text that the committee members debate and propose to amend frequently is not the text of the bill (or any of the bills) that had been introduced and referred to it. Instead, the committee may debate and amend an entirely different text that has originated within the committee. For example, the committee chairman may direct the committee staff to write a new draft of a bill that reflects a different approach to the subject, an approach that is preferred by the chairman and the other committee members of the majority party. Then, ultimately, the committee votes to send that new draft back to the House or Senate, either as a new bill or as an amendment that proposes to replace the entire text of a bill that had been referred to the committee.[61] In either case, it is the committee’s version of the bill, not the president’s draft or any other bill on the subject, which is taken up for consideration in plenary session.

Second, and reflecting the sense of legislative autonomy that is well-rooted in both the US Senate and the House of Representatives, neither house feels any compulsion to wait until the other has passed a bill before beginning its own legislative work on the same subject. It is not at all unusual for legislative hearings on a subject to take place simultaneously, or close to it, in both houses. The fact that a House committee has reported a bill on some subject usually does not discourage the corresponding Senate committee from developing its own bill on that subject (or conversely). Even if the House has passed its bill and sent it to the Senate, that is no bar to the Senate committee continuing to mark up and report its own bill which may take a fundamentally different approach to the subject. Finally, if either house then faces a choice between taking up, in plenary session, a bill already passed by the other house or the version of a bill on the same subject that originated in one of its own committees, it almost invariably chooses to devote its time and attention to the text written by some of its own members in one of its own committees.[62]

Imagine, for example, that the US House of Representatives has passed a bill, H.R. 1, to reform the health care delivery system, and has sent that bill to the Senate for its consideration. But imagine also, that the appropriate Senate committee already has reported, or is about to report, its own bill, S. 2, for the same purpose as H.R. 1.[63] Under these circumstances, the Senate is most likely to consider, debate, and amend S. 2 instead of H.R. 1. But if the Senate does so, and then passes S. 2 and sends it to the House, the two houses will have failed to pass the same bill, which they must do before they can begin any formal process to reach agreement on just what its provisions should be. After completing its work on S. 2, therefore, the Senate is likely to take up H.R. 1 (the House-passed bill) and amend it by replacing every substantive word it contains with the corresponding text of S. 2 (as the Senate may have amended it). In other words, the Senate substitutes the text of its own bill—I shall refer to it here as a full-text substitute amendment—for the text of the bill it had received from the House of Representatives.

This may seem complicated and technical (and it is), but it also is important because it means that when the two houses begin to reach agreement on all the substantive differences between the House and Senate versions of their health care delivery reform bill, all of these differences are embodied in what is formally only one Senate full-text substitute amendment to the House bill. A comparison of the text of H.R. 1 (as passed by the House) and the Senate substitute amendment to it (the amended text of S. 2, which now is the Senate’s amendment to H.R. 1) may reveal countless numbers of specific differences, but each of these differences is not a separate amendment; they all are part of one single Senate substitute amendment. Indeed, the House and Senate versions of the bill may take such radically different approaches to the same subject that it is not possible to line them up against each other and to identify specific differences between them.

In Washington as in Canberra, the two houses must treat each Senate amendment to a House-passed bill (or each House amendment to a Senate-passed bill) as a single entity. They may not disaggregate an amendment into two or many component parts and reach agreement separately on each of those parts. Instead, they must reach agreement on each amendment as a whole. Consequently, the individual approach to reaching legislative agreements is not an available option when, procedurally, there is only one full-text substitute amendment on which agreement is needed, even if that amendment differs from the bill it amends in ways too numerous to count. (And, as I shall discuss in the next section, this is precisely what now typically occurs in Congress.)

On the other hand, this procedural situation is very well-suited to the collective approach to reaching agreement, especially in a conference committee, because the conferees have before them every element of the House bill as well as every element of the Senate substitute amendment. They may pick and choose from among the general approaches and specific provisions that are found in either the House or the Senate versions of the bill. In fact, they have no alternative. The conferees’ final agreement ultimately takes the form of a new, third version of the bill that they propose for the House and Senate to accept in place of the versions of the bill that the House and Senate originally had passed.

This procedural situation does not arise in Canberra. If the Senate receives a bill from the House and does not refer it to a Senate committee, then, of course, it is that bill that the Senate debates and amends in plenary session. The result is a House bill that the Senate passes with separate discrete amendments (assuming the Senate amends the bill at all). And even if the Senate does refer a House bill to a Senate committee, and even if the committee is less than pleased with the bill, it does not report to the Senate instead on a bill that the committee itself has written or that a senator has introduced. This practice flows logically, and even necessarily, from the understanding shared among all members of both houses and all parties that the government should control the legislative agenda in both houses, and that this control extends beyond identifying the subjects that deserve legislative attention and includes writing the exact language of the bills that the House and Senate begin to consider in plenary sessions.[64]

As a result, whenever the Australian Senate agrees to two or more amendments to a government bill that the House already has passed, each of these amendments remains procedurally separate from the others. Never are they all combined into a single Senate substitute amendment, as occurs so often in Washington. This situation does not preclude recourse to the collective approach in Canberra; the two houses could refer all the Senate’s separate amendments to a conference or a conference committee with the hope that the conferees could reach agreements with respect to all of them. However, the legislative process in Canberra never produces a situation in which recourse to the collective approach is so suitable because there are not separate amendments to be addressed individually.

On the other hand, there is good reason to think that the Australian Senate and House of Representatives do reach agreement with respect to many Senate amendments by means of an approach that is collective in a political sense, though not in a procedural one. So long as there is a non-government majority in the Senate, the government must attract the support of enough non-government senators to pass each of its bills, which obviously may involve accepting one or more amendments to each bill. Clearly, if the government agrees to some unwelcome Senate amendments in order for its bill to pass the Senate, it thereby commits itself, explicitly or implicitly, to accepting those same amendments when the Senate sends them to the House for its concurrence. If the government were to accept some amendments in the Senate and then fail to support those same amendments in the House, it would be seen to be reneging on its commitment and, as a result, seriously jeopardising its ability to negotiate the passage of other bills in the Senate.

So it is reasonable to assume that when the House takes up Senate amendments to one of its bills, the House in effect acts collectively, in a political if not in a procedural sense, in agreeing to all those amendments that the government already had committed itself to supporting. There is no such prior commitment, however, with respect to any other amendments that the government opposed in the Senate but that the Senate nonetheless approved.

Procedural consequences of political change in Washington

In sum, there is a fundamental difference in the formal procedures by which the US Congress and the Australian Parliament have gone about reaching bicameral agreements on legislation. Both of the assemblies have relied most of the time on a process of exchanging messages and amendments with the hope that this process eventually would produce a version of each bill on which both houses (or, in Australia, both parties) could agree. The difference is that the Australian Parliament has relied on this process exclusively, whereas the US Congress has tended to rely on conference committees (that is, temporary joint committees) to negotiate a proposed package settlement of all the differences between House and Senate versions of the most important and contentious legislation.

The reasons for this difference, as I have tried to identify and explain them here, are various. However, they all can be traced, directly or indirectly, to the constitutional, institutional, and political contexts in which the US and Australian legislative processes take place. Of particular importance at the constitutional level are the parliamentary sources and underpinnings of Australia’s national political system, reflected, for example, in the government’s control of the legislative agendas of both houses, even though the government usually does not control a majority of seats in the contemporary Senate. As for the institutional context, the different roles that Australian (Senate) and US standing committees play at the front end of the legislative process are reflected, not surprisingly, in the part they play at the back end. And of particular importance at the political level is the strength of party discipline in the Australian Parliament, even when compared with the historically high levels of party cohesiveness in the contemporary US Congress. From these differences, and others closely related to or flowing from them, emerge what seem to me to be a satisfactory set of explanations why the collective approach to reaching bicameral legislative agreements has been an available and valued procedural option in Washington but not in Canberra.

Neatness would be served if it were possible to conclude the analysis here. In recent years, however, there have been two related—and, unfortunately, complicated—changes in congressional procedures and practices of which we need to take account. One is unlikely to be reversed; the durability of the second remains open to conjecture. For as long as they persist, however, they affect the foregoing analysis in two ways. First, they reduce the clarity of some of the differences that I have drawn between politics and procedures in Washington compared with those in Canberra. But second, they only strengthen the underlying argument that the legislative procedures in a democratic national assembly will reflect the political context in which they are employed, and sooner or later they will be adjusted to reflect changes in that context.[65]

The first of these two developments is one that I introduced earlier: the possibility that one house in Washington may pass a bill received from the other, but pass it with a full-text substitute. This substitute constitutes an entirely different version of the bill and it often embodies a significantly different approach to the subject of the bill. Even so, there is only one amendment on which the two houses now have to agree, no matter how many policy differences may be reflected in that amendment.

If the House passes a bill that takes one approach to an issue and the Senate then passes it with a full-text substitute amendment that takes a different approach, the two houses usually have no choice but to negotiate a third approach—presumably some kind of compromise between the approach proposed by the House and the approach preferred by the Senate. This third approach can be presented by a conference committee and embodied in its report which then is put before both houses for their approval. Alternatively, though, the same new version of the bill can take the form of a new full-text substitute that the House is asked to approve in lieu of the Senate’s previous amendment to the House’s bill. If the House approves the new, negotiated version of the bill, and the Senate does so as well, the two houses will have reached agreement and the bill will be ready for it to be presented to the president for his approval. The various differences between the two houses’ approaches to the issue will have been resolved collectively, not individually, whether by means of a conference committee or the exchange of amendments, and for the reason that the approaches which the two houses initially took to the same issue could not be reduced to a series of specific differences that lent themselves to being resolved individually.

Increasingly during the closing decades of the last century, the US House and Senate found themselves in precisely this situation—having before them a bill first passed by one house and then passed by the other with a very different full-text substitute amendment. One important reason may well have been the perceived time pressures that made it seem impractical for the second house (usually the Senate) and its relevant standing committee to await the arrival of a House-passed bill and then to base their deliberations on that version of the bill. Instead, it became increasingly likely for the Senate to debate and amend its committee’s own bill on the same subject and then to substitute the text of that bill for the text of the House’s bill, whenever it arrived.

Smaller, less important, bills often were passed by one house and then passed by the other house with several discrete amendments, on which the two houses could negotiate and reach agreement individually, through the exchange of amendments. By the 1990s, however, the differences between the two houses over most major bills had to be resolved collectively because of the prevalence of full-text substitutes from one house for the texts of bills already passed by the other. The primary exceptions were the annual appropriations bills, generally thirteen in number, which the House always passes first. Until the 1990s, the Senate’s general practice had been to await the arrival of each of these bills from the House. The Senate’s Committee on Appropriations and then the Senate as a whole would consider and vote on amendments to each House bill, passing it with many, sometimes hundreds, of individual amendments, rather than with a single full-text substitute.

Consequently, these appropriations bills were much better suited to the individual approach to reaching agreement. Even so, in the absence of severe time pressures, congressional practice was to send each appropriations bill to a conference committee. It was not unusual, however, for the reports of these committees to leave some of the individual Senate amendments remaining in disagreement.[66] In that case, first the House and then the Senate would vote on the conference report, proposing a collective resolution of most of the bicameral differences over the bill, and then act on each of the remaining amendments individually. The result was to combine the use of both the collective and the individual approaches to resolving legislative disagreements.

Congressional practice regarding appropriations bills now has changed. Today, the Senate is as likely to pass an appropriations bill with a full-text substitute amendment as any other bill. Consequently, because of this procedural development, the individual approach to reaching agreement now tends to be limited to only those bills of lesser significance on which the differences between the two houses are relatively small and easily identifiable.

There is no obvious political reason for the development that I have just discussed. By contrast, the second procedural development, which is even more recent, clearly has its roots in the changed partisan differences that have come to characterise the contemporary US Congress.

The standard logic of American legislative politics assigns a triad of motives to the members of US congressional conference committees. First, each member (or ‘conferee’) seeks to advance individual goals and interests that are some combination of making what he or she conceives to be good national policy, protecting or advancing the interests of his or her constituency, and enhancing his or her own power and influence within Congress. Second, each conferee also seeks to promote a conference agreement that advances the policies and programs of his or her political party. And third, each conferee is to advocate the position of his or her house vis-à-vis that of the other.

This triad of individual, partisan, and cameral (as opposed to bicameral) motives, of course, usually are closely linked. Conferees’ own policy views typically are very close to, and may be indistinguishable from, those of their party, and US representatives and senators historically have demonstrated a remarkable ability to conclude that their own policy preferences and those of their constituents are compatible or identical. Also, conferees enhance their own standing within Congress by delivering back to the House or Senate, as the case may be, a conference report that represents a victory both for their party and for their house over the other party and the other house.

However, there always has been the potential for conflict between conferees’ partisan interests and their cameral responsibilities, between their incentive to negotiate with their party’s policies in mind and to negotiate with a view toward reaching a conference compromise that resembles the original position of their house more than that of the other house. The cameral dimension of conference negotiations is reflected in the fact that conference agreements must garner the support of a majority of the conferees from each house (as well as in scholars’ attempts to decide whether the House or Senate ‘wins’ more often in conference). And the partisan dimension is reflected in the fact that a majority of the conferees from each house invariably are members of the majority party.

The Speaker appoints all the conferees from the House of Representatives, but his discretion is constrained in three ways. He is expected to draw most, if not all, of the conferees from the membership of the standing committee or committees with jurisdiction over the bill in question. He also is expected to accept and appoint the minority party’s choices for its conference committee members. And finally, the rules of the House admonish the Speaker to appoint as House conferees ‘no less than a majority who generally supported the House position as determined by the Speaker’, including those who were ‘primarily responsible for the legislation’ and ‘the principal proponents of the major provisions’ of the bill being sent to conference (clause 11 of rule 1).

Although this rule cannot be enforced, its purpose and intent are clear: to ensure that most of the House’s conferees (i.e., the majority who are drawn from the majority party) will promote a conference agreement, and in particular, an agreement that reflects the position of the House on the major matters in disagreement with the Senate. Although the Speaker cannot assure that the minority party representatives on the conference also will seek the same kind of conference agreement, it was not supposed that the Democratic and Republican representatives (or senators) on a conference automatically would be at odds with each other. For much of the last century, most members of some standing committees of the House and Senate often were able to reach bipartisan agreements on legislation that they then joined forces to defend against amendments during plenary sessions of their own house and against members of ‘the other body’ in conference. Other standing committees, not surprisingly, tended to be more partisan, in part because of the subjects with which they dealt, so the House and Senate conference delegations that their members tended to comprise also were more likely to divide along party lines.

In short, there was a shifting balance between the influence in conference committees of party preferences as compared with chamber preferences. The degree to which conferees emphasised their preferences as party members or their responsibilities as the agents of their houses depended on such things as the subject of the bill in question, the controversy that it evoked, its centrality to the fundamental policies of either or both parties, the similarities and differences in constituency interests among committee members of the two parties, perhaps the proximity of the next election, and even the personalities of the leading conferees and their working relationships with the other conferees from their house but the other party, as well as with those from the other house but the same party.

In recent years, the growing strength of party in US congressional voting has affected this balance. As each party in Congress has become more homogeneous and as the two parties have become more polarised, the notion that the Republicans and Democrats from the House or Senate who are appointed to a conference committee are expected to work together in support of their house’s version of the bill in question has come to be seriously challenged, if not reduced to a polite fiction. The high levels of partisanship that now frequently characterise votes in the House and Senate on passing the two versions of a bill that then are sent to conference are likely to carry over into the work of the conference committee itself. Minority party conferees who voted against passing the bill in the House or Senate have less incentive to argue in conference in favour of their house’s position than they would if they had voted for it. So majority party conferees have less reason to expect bipartisan support for their efforts to defend their house’s position against the position of the other house. In short, the heightened party polarisation in Congress has tended to encourage lines to be drawn in conference committees between the two parties, not between the two houses.[67]

The result has been to cause what may prove to be a temporary or lasting change in how the US Congress now attempts to reach bicameral agreements on bills that previously would almost certainly have been committed to conference committees. The change has been much more a change in practice than a change in the formal procedures of either house; in fact, one reason for the change in practice has been a new-found willingness among senators to take advantage of procedural opportunities that previously had been ignored.

First, more decisions in conference came to be made along party lines, and fewer minority party representatives and senators were willing to endorse conference committee reports. Walter Oleszek quotes a Senate majority party conferee as telling the minority party conferees in 1991 that ‘[w]e don’t expect you to sign [the conference report], so we don’t expect you to be needed’ in the negotiations.[68] And second, minority party conferees came to find that they sometimes were being excluded from the informal negotiations that almost always precede official conference committee meetings, and that are intended to reduce those meetings to nothing more than formal presentations and endorsements of agreements already negotiated. To illustrate, Oleszek quotes three other laments from senators, made in 2000, 2001, and 2005:[69]

I have been appointed to conference committees in the Senate in name only, where my name will be read by the [presiding officer] and only the conference of Republicans goes off and meets, adopts a conference report, signs it, and sends it back to the floor without even inviting me to attend a session.

After much talk of bipartisanship, the other side locked out the Democrats from the conference committee ... We were invited to the first meeting and told we would not be invited back, that the Republican majority was going to write the budget all on their own, which they have done.

On issue after issue, we have had conferences where the minority was excluded so that the majority could ram through unpopular provisions as part of an un-amendable conference report.

Third, therefore, minority party senators, frustrated by what they viewed as their increasing exclusion from the conference process, began to delay or prevent bills from being sent to conference in the first place.[70]

It has long been established that there are three steps the Senate must take in sending a bill to a conference committee. First, it must disagree to the House amendment to a Senate bill, or insist on its own amendment to a House bill. In doing so, the Senate reaches what is known as ‘the stage of disagreement’, and both houses must reach this stage before a conference committee can be created. Second, the Senate must either request a conference with the House, or it must agree to the request for a conference that the House already has made. And third, the Senate must authorise its presiding officer to appoint the Senate’s conferees, if that is the Senate’s wish, as it always is.[71]

For decades, it had been the invariable practice of the Senate to take these three steps as if they were one, and to take them with the unanimous consent of all the senators present, and without any delay or debate. It probably never occurred to many senators that, by objecting to such a unanimous consent request, they could require the Senate to take each step by agreeing to a motion that is debatable and, therefore, subject to being filibustered.[72] Even if the Senate is prepared to invoke cloture on each of the three motions as soon as its rules permit, that still can leave as much as thirty hours for senators to debate each motion after invoking cloture on it. Consequently, a large and determined Senate minority can delay for days or even weeks what had been an almost unnoticed process of completing the procedural formalities necessary to send a bill to conference.[73]

By asserting procedural rights that they had implicitly been waiving, minority party senators discovered an effective way to respond to what many of them saw as their increasing irrelevance to the process of resolving legislative disagreements in conference. By threatening to compel the Senate to resort to making and agreeing to the three pre-conference motions, minority party senators (whether Democratic or Republican) could protest what they saw as the majority’s unwillingness to take satisfactory account of their concerns during conference negotiations, and even the majority’s apparently deliberate decisions to exclude them from those negotiations altogether. And by making the same threat or, even worse, by carrying through on it, the Senate minority could seriously delay or even prevent the establishment of conference committees on bills that they opposed in the form in which the Senate had passed them, and which the minority did not believe would emerge from conference in a form they could support. In turn, not surprisingly, the majority could conclude that the minority was interested only in delaying or blocking legislation, and would have nothing constructive to contribute to bicameral negotiations, whether in a conference committee or otherwise.

By compelling the Senate to comply with long-established procedures that the Senate had gotten into the habit of bypassing consensually, minority party senators were able to make the collective approach to resolving legislative disagreements with the House much more costly in both time and effort. As a result, the advantages of the collective approach, compared with the individual approach, declined during the past decade, and Congress turned to the process of exchanging messages and amendments to reach bicameral agreements on some bills that, in years past, almost certainly would have been given over to conference committees. As I noted earlier, even before these developments, less than ten per cent of the bills that became law had been committed to conference committees, so this recent trend may not have made much of an impression statistically.[74] However, members of both houses unquestionably would agree that it complicated and retarded what already had been a complex and time-consuming legislative process.[75]

For the majority party in the Senate, and especially when the same party has majorities in both houses, there are some compensating advantages in having to rely more on the individual than the collective approach to reaching bicameral agreements on difficult bills. It offers more flexibility: there is no need for a public meeting, however perfunctory, of the conference committee; there is no need to write a conference report and the accompanying explanation; and there is no need to give members of either house time to review the conference report. Also, some restrictions on the kinds of proposals that can be included in conference reports do not apply to the exchange of amendments between the House and Senate.

In principle, for example, conferees are supposed to resolve each disagreement between the two houses by reaching a settlement that falls within the scope of the differences between their positions. To take the most simple kind of obviously hypothetical example, if the House proposes to appropriate $5.00 for a purpose and the Senate proposes $10.00 instead for the same purpose, the conferees can agree on $5.00, $7.50 or $10.00, but they are not to agree on $2.00 or $12.00 because either sum would fall outside the scope of the differences between the House and Senate positions. Furthermore, conferees are not to propose any changes in provisions to which both houses already have agreed, nor are conferees to propose in their report recommendations on any subject that was not addressed in either the House or Senate version of the bill that they sent to conference. All these restrictions apply to conference reports but not to amendments sent back and forth between the houses.[76]

On the other hand, when the Senate considers any new amendment to send to the House, that amendment is not only subject to being filibustered, as a conference report would be, it also might be subject to being amended, which a conference report is not. In other words, whereas the use of the collective approach to resolving differences can present each house with a package settlement that it usually must accept or reject in its entirety, the use of the individual approach does not come with the same guarantee.

In 2007, the Senate acknowledged the problem that had developed with regard to the participation of minority party senators in conference committees. One small part of the much larger and ambitiously named ‘Honest Leadership and Open Government Act of 2007’ states in part that ‘(1) conference committees should hold regular, formal meetings of all conferees that are open to the public; (2) all conferees should be given adequate notice of the time and place of all such meetings;’ and ‘(3) all conferees should be afforded an opportunity to participate in full and complete debates of the matters that such conference committees may recommend to their respective Houses ...’[77]

Notice though that these are admonitions, not requirements: the law states that these things should be done, not that they shall or must be done. In fact, this section of the law is explicitly identified as a ‘sense of the Senate’ provision, which is a way for the Senate to express its opinion or judgement on something without taking any legally binding action. As such, this provision is not enforceable either as law or as a rule or order of the Senate.[78] The provision calls for nothing that should not be, and once was, standard practice in both houses of Congress.

Furthermore, majority leaders of both parties in the Senate sometimes have employed a procedural stratagem by which they could try to have their legislative cake and eat it too. This stratagem allowed them to avoid the newly severe difficulties they were encountering in attempting to send bills to conference committees. It also allowed the Democratic or Republican majority party to continue excluding the minority party, to the extent it wished to do so, from negotiations to resolve legislative differences with the House, and it could preclude amendments to, and compel a single vote on, its proposal to resolve those differences through an exchange of amendments, just as there are no amendments to and only one vote on accepting or rejecting a conference report.

The procedures are complex but they can be summarised briefly.[79] If the House passes a bill and the Senate then passes it with its own full-text substitute, the majority party members of the two houses can negotiate a third, compromise version of the bill, with only as much participation by the minority in the negotiations as the majority wishes.[80] Once this agreement is reached, the House is asked to accept it as a replacement for the Senate’s full-text substitute, and the House can do so by a simple majority vote after no more than one hour of debate and with no opportunities for amendments to the new, third version of the bill. After the House informs the Senate of the action it has just taken, the Senate’s Majority Leader moves that the House also accept the same new version of the bill; formally, he moves that the Senate concur in the House amendment that embodies this version.

Before the Senate votes on such a motion to concur, any other senator may offer a motion to amend the new proposed text of the bill; formally, he or she can move that the Senate concur in the House amendment with a further Senate amendment. To pre-empt that possibility, the Majority Leader makes that motion himself, with the further Senate amendment he proposes being of negligible importance (for example, to change by one day the date on which the bill, once enacted, will take effect as law).[81] Finally, the Majority Leader files a motion to invoke cloture on his original motion to concur.[82]

Until the Senate votes on whether to invoke cloture on the Majority Leader’s motion to concur, the actions he has taken prevent any other senator from offering any amendment, and especially an unwelcome amendment, to the proposed final version of the bill that majority party members of the two houses have negotiated. The critical question then becomes whether three-fifths of all the serving senators will vote for the cloture motion.

If the cloture motion fails, the stratagem collapses. Eventually, there will be a vote on the Majority Leader’s motion to amend the House amendment (formally, to concur in it with a further Senate amendment), which represents the majority party’s negotiated bicameral agreement. Thereafter, other senators can propose one or more amendments of their own to the proposed bicameral agreement. If, however, the Senate invokes cloture by the required three-fifths vote on the motion to accept (formally, concur in) the House amendment, that vote indicates that a simple majority of senators is prepared to accept the bicameral agreement (in the form of the pending House amendment) without change. So no motion to concur in the House amendment with a Senate amendment is likely to prevail.

Any readers who have slogged their way through the last four paragraphs will see, therefore, that these procedures may enable the Senate’s majority party to compel a vote on whatever bicameral agreement between the two houses that the majority party’s political and policy leaders have been able to reach. Furthermore, they can arrange for this vote to take place in a way that undermines any attempts to change the terms of the agreement, just as if the Senate were considering a conference report instead. Implementing this strategy does require the Senate to agree to a cloture motion, but so too may cloture be required to end the debate on a conference report.

The key to the success of this strategy is convincing three-fifths of all senators to vote for the Majority Leader’s motion for cloture on his motion to concur. Most often, the majority party in the Senate holds fewer than three-fifths of the seats, meaning that the strategy cannot succeed if the minority party is united in opposing it. The 2008 Senate elections produced a result that is unusual but not unprecedented in modern US history: the Democratic majority did have 60 votes if it also held the votes of two independent senators (on one of whom it can depend). This made the strategy I have described a viable one, even when the majority was opposed by a united minority party.

As of January 2010, however, the Democrats had lost their 60-seat majority in the Senate, raising doubts about the continuing viability of this strategy. Furthermore, the current distribution of seats in the Senate between the parties is far from being stable and secure. In fact, conventional wisdom, based on previous election patterns, suggests that the Republican minority party will hold more than 41 of the 100 Senate seats when the Senate convenes in January 2011, following the November 2010 elections. Should that be the case, the Democratic majority in the Senate (presuming that the Democrats remain in the majority) may face the unhappy choice between having difficulty overcoming minority party efforts to block bills from going to conference, and having the same difficulty compelling the Senate to vote without amendments on the legislative product of its reliance instead on an exchange of amendments.

It seems likely that the intent of the Senate’s minority party always had been to convince the majority to let it back into the conference negotiations. Filibustering the triad of motions that are necessary to send a bill to conference is not a pleasant undertaking; it requires time, energy, and a considerable degree of coordinated effort. Moreover, if the Senate minority requires the majority to resort to an exchange of amendments instead of sending a bill to conference, the minority can be sure that it will be excluded from whatever negotiations take place. Finally, and in any event, the Senate minority retains the right to filibuster any proposed agreement that is produced, regardless of whether it emerges through a conference committee or an exchange of amendments with the House. Breaking such a filibuster requires jumping the 60-vote hurdle that is necessary to invoke cloture either on a conference report or on a motion that the Senate concur in a House amendment that is the functional equivalent of a conference report.

There are good reasons, then, for the Senate to revert to its practice of sending major and contentious bills to conference committees, unless the pressure of time simply is too great to permit it.[83] But this would not mean a return to the political status quo ante. The intra-party homogeneity and inter-party polarisation that now characterise both houses of Congress create a different context for bicameral negotiations than when both the proponents and opponents of a bill in both houses were more likely to find a significant number of allies on the other side of the aisle. To whatever extent House or Senate conferees of both parties once really did think of themselves as negotiators on behalf of their own house’s version of a bill—and that expectation often was honoured in the breach—today the opponents in conference are much more the members of the other party than the members of the other house. For this reason, the politics of reaching bicameral legislative agreements in Washington will remain different from what they were several decades earlier, even if the procedures for reaching these agreements revert to their earlier pattern.

Procedures in constitutional, institutional, and partisan context

To summarise: in both Washington and Canberra, the national constitutions effectively leave it up to Congress and Parliament to devise procedures for resolving their legislative disagreements, and such disagreements are certain to arise if each House of Representatives and each Senate is not reluctant to exercise its legislative powers. The US Constitution is totally silent on how these disagreements are to be resolved. Perhaps it did not occur to the authors to address this potential problem; perhaps they thought it best to leave it to the imagination and wisdom of future generations of legislators. The Australian Constitution does address the subject, but only by establishing a mechanism—a joint meeting after legislative deadlock on a bill has been reached three times, and with a new election for all members of both houses intervening between the second and third attempts—that they must have known would be entirely impractical to use during the ordinary course of legislative business. So they too left it to future members of Parliament to devise other procedures that are better suited to the workaday business of legislating.

Both Parliament and Congress responded by making room in their standing rules and orders for both the individual and collective approaches to resolving their differences. Their rules are not identical, of course, nor would we expect them to be. For instance, the congressional rules are more explicit and detailed on the subject of what substantive agreements the conferees can present to the House and Senate in their reports. And the parliamentary rules are more explicit in providing for conferences (the collective approach) when reliance on the individual approach has proven unsuccessful. However, any such differences in formal procedures pale in comparison to the key difference in practice that I have described: Congress would find it difficult to do its work without using conference committees; Parliament does not. Although only a small fraction of new US laws are the product of conference committees, Congress has relied on such committees for resolving bicameral disagreements on many (or most) of the most important and complex bills the two houses pass, and, notwithstanding the recent political and procedural developments I have described, it probably will continue to do so. Parliament has created only two conferences since Federation, and shows no inclination to change its practice.

Despite the obvious differences reflected in the presence of a president in Washington and a prime minister in Canberra, the Australian and US Senates share much the same constitutional powers and responsibilities for legislating, and they are differentiated from their companion Houses of Representatives in many of the same ways: for example, the numbers of their members, the ways in which the members of the two houses are elected, the length of their terms, and the sizes and diversity of their constituencies. And the essential procedural rules for addressing the bicameral legislative disagreements that almost inevitably will arise between the two houses of either the Congress or the Parliament would be recognisable in important respects to the denizens of the other.

One reason, I have argued, for the different ways in which the two assemblies do or don’t use these procedures flows from the differences in stature and influence between US congressional committees and Australian parliamentary committees, even in Canberra’s Senate with the relatively active part that its committees play in reviewing legislation. US congressional committees tend to write the bills that the House and Senate consider, and their members guide and often dominate the process of debating and amending those bills in the House and Senate chambers. It would seem only natural, therefore, for the corresponding committees of the two houses to take the lead in reaching bicameral agreements and, historically, conference committees have provided a forum for just that to take place. In Canberra, on the other hand, and even in the Senate, committees can be valuable but ultimately less pivotal players in the design of legislation than their counterparts in Washington. It would be much less logical and appropriate, therefore, for Parliament to rely on a bicameral resolution mechanism that could be a natural extension of its Senate and House committee systems.

For the two other reasons that strike me as critical, we must turn to contextual factors originating outside of Parliament or Congress. One is historical or philosophical, depending on one’s point of view. Since soon after the US Congress first met, the House and Senate both have viewed themselves as important legislative actors, being either equal or superior in importance to the other. Even before US senators came to be directly elected, instead of being formally chosen by the state legislatures, neither the House nor the Senate had been inclined to defer to the other on anything approaching a regular basis. In Australia, on the other hand, I think it fair to say that there always has been some question as to how constitutionally legitimate and appropriate it is for the Senate to press its own legislative judgements when they differ from those of the House of Representatives and the executive government that is constitutionally responsible to it.

The essential legislative negotiations between Australia’s Senate and the government would seem to be those that produce the compromises necessary to persuade the Senate’s non-government majority to pass government bills. These negotiations may in fact resemble a conference in that the resulting agreement is on a package of amendments that must be preserved as a package, even if the Senate adopts them singly, if the agreement for the Senate to pass the bill as amended is to survive. The fact that these negotiations tend to take place informally and in private is consistent with the tendency of both parties when in government to disparage the Senate as an equal legislative partner. Once an Australian government compromises enough to secure passage of one of its bills through the Senate, it should be expected to stand by the amendments to which it agreed in the Senate when it comes time for the House of Representatives to accept those amendments. Beyond this point, however, Australian governments seem disinclined to compromise much further by accepting additional Senate amendments to their legislation.

The other contextual factor on which this analysis has concentrated is the historically greater strength of parties in Parliament compared to Congress. The almost perfect party discipline that has characterised governance in Canberra almost since Federation has made the House of Representatives a dependable agent of the government and discouraged the regular use of any legislative procedures which presume that the House and Senate are equal partners in the legislative process or which assume that their members’ primary interest is in supporting the positions adopted in their chamber against those adopted in the other. In the US, the divisions within each party that prevailed for most of the 20th century and that were reflected in both the House and the Senate lent some credence to the notion that the process of resolving bicameral legislative disagreements involved a real element of House versus Senate as well as Democrats versus Republicans.

When the Australian government recently had a temporary majority in both houses, there were few significant differences between the versions of legislation produced by the House and Senate and, therefore, few contentious legislative disagreements to be resolved. By contrast, for all but four years between 1933 and 1981, the Democratic Party controlled both houses of Congress, yet conference committees were so common and important as to be described as ‘the third house of Congress’.[84] In Australia, partisan politics contributed to making conferences unnecessary; in the US, the lack of equally clear and consistent party differences contributed to making conference committees repeatedly useful if not essential.

The now-higher levels of party polarisation in Congress have disrupted the previously established practices for resolving House–Senate disagreements on legislation. The majority party tended to exclude the minority party from bicameral negotiations in conference committees in the belief that the minority was less interested in promoting agreement than delay and obstruction. The minority responded in the Senate by making it painful if not impossible to send bills to conference, to which the majority reacted by devising a convoluted procedural strategy to force a Senate vote on its preferred bicameral agreement, without amendments and without the need to create a conference committee. However, this strategy depends on the majority being able to amass votes from 60 of the 100 senators, which normally—and once again today—requires some of those votes to come from the minority side of the aisle. In the short term, therefore, the procedures for reaching legislative agreements are almost certain to remain more stable and predictable in Canberra than in Washington.

[*]     This paper was prepared for presentation at the Ninth Workshop of Parliamentary Scholars and Parliamentarians, 24–25 July 2010, Wroxton College, Oxfordshire, UK, sponsored by the Centre for Legislative Studies at the University of Hull.

[1]     My sincere thanks to Elizabeth Rybicki in Washington and David Sullivan in Canberra for alerting me to some inaccurate and intemperate statements that somehow found their way into an earlier draft of this paper.

[2]     In post-Ceausescu Romania, to cite an extreme counter-example, members of the two houses were elected in the same ways and for terms of the same length, and the two houses enjoyed the same powers. The original plan evidently called for a ‘differentiated bicameralism’. ‘The final Constitution called for an undifferentiated bicameralism, however, conferring an identical democratic legitimacy upon both chambers. This unusual choice was partly motivated by the framers’ fear that one institution’s claim to ultimate legitimacy might permit an excessive concentration of power’. Elana Stefoi-Sava, ‘Romania: Organizing Legislative Impotence’, East European Constitutional Review, vol. 4, no. 2, Spring 1995, pp. 78–83. Stefoi-Sava also describes the resulting arrangement as ‘carbon-copy bicameralism’ and ‘monochrome bicameralism’.

[3]     In France, for example. See fn. 32 below.

[4]     In Australia, as I discuss below, there is the constitutional option to resolve legislative disagreements in a joint session. However, this option arises only if the two houses fail to pass the same bill in the same form on three separate occasions, with a ‘double dissolution’ election of all senators and members of the House of Representatives occurring between the second and third occasions. Obviously, therefore, this process is too slow and cumbersome to be used regularly. And in fact, it has been used only once.

[5]     This assumes the existence of a party system, which usually is a safe assumption to make, at least for national assemblies. In non-partisan assemblies, or in assemblies in which parties are inchoate or embryonic, the difficulties of reaching agreement may be greater because of the need for supporters of legislation in each house to assemble majorities one vote at a time. On the other hand, the absence of parties may facilitate agreement because there are less likely to be groups of assembly members whose first instinct is to oppose each other’s positions.

[6]     References throughout to ‘rules’ encompass not only the codified and formally adopted rules of each house, but also the enforceable (and published) precedents and practices by which these rules have been interpreted and applied.

[7]     But see the discussion of legislative autonomy below.

[8]     Also, Australian governments think more than twice before asking the Governor-General for a double dissolution because the elections that follow are more likely to result in the election of minor party and independent senators than are the usual half-Senate elections at which half of Australia’s senators are elected.

[9]     For most purposes here, the Coalition is treated as if it were a single party. While it is true that, during Senate elections at least, the Liberal and National parties may run separate slates of candidates, the two parties almost always vote as one in Parliament.

[10]    There are twelve Australian senators from each of the six states and two from each of the two territories (the Northern Territory and the Australian Capital Territory).

[11]    In elections to both houses, Australia also uses a system of preferential voting, such that the House candidate who wins a plurality of votes in his or her constituency does not always win the seat. However, the details of the two voting systems need not detain us here. These details are readily available elsewhere—for example, in Scott Bennett, Winning and Losing: Australian National Elections. Melbourne, Melbourne University Press, 1996.

[12]    There are no minor parties represented in Congress, and rarely more than one or two independents in either house. Furthermore, independents usually act, and for most purposes are treated, as members of one party or the other.

[13]    One challenge in writing about contemporary political institutions is that they can be moving targets. Days after I first wrote this in 2009, a Republican US senator announced his intention to switch parties. That development, combined with the belated resolution of a contested Senate election in favour of the Democratic candidate, gave Senate Democrats a total of exactly 60 votes, at least until the subsequent death of Democratic Senator Edward Kennedy. Once Kennedy’s temporary replacement was named, the Democrats regained what was, in theory, a filibuster-proof majority until a Republican successor was elected in January 2010, which again reduced the Democratic majority to 59. Regardless, Democratic leaders in the Senate are acutely aware that not all majority party senators can be expected to vote for all cloture motions.

[14]    See John Warhurst, ‘Conscience voting in the Australian Federal Parliament’, Australian Journal of Politics and History, vol. 54, no. 4, pp. 579–96. A recent, rare, and remarkable exception was the carbon pollution reduction legislation that divided the Liberal Party at the end of 2009, with some Liberal senators crossing the floor and the leader of the Liberal Party being replaced.

[15]    It should not be assumed, however, that all Senate divisions pit the government against the Opposition, with minor party and independent senators holding the ‘balance of power’. The government and Opposition have voted together on divisions more often than their combative public rhetoric would suggest. See Stanley Bach, Platypus and Parliament. Canberra, Department of the Senate, 2003, pp. 157–237.

[16]    The bill may be delayed if these differences are intractable.

[17]    But note fn. 14 above as well as the several instances in which Senator Joyce, the current National Party leader in the Senate, crossed the floor to vote with the ALP.

[18]    Quoted in Walter J. Oleszek, Whither the Role of Conference Committees: An Analysis. Congressional Research Service report RL34611, August 2008, p. 10.

[19]    In Washington, however, absences are much more common. Representatives and senators are most likely to be present for the most important votes, but perfect attendance never can be taken for granted and often cannot be achieved. Especially on controversial matters, the majority party leaders in both houses do their best to schedule votes for days of the week when the fewest possible number of their members are away from Washington. However, there have been occasions on which party leaders have encouraged some of their fellow party members to miss a vote, the leaders preferring those party colleagues to be absent rather than having them present and voting against the prevailing party position.

[20]    During the past several decades, there has been a striking increase in the number of Senate filibusters and filibuster threats, as well as in the number of cloture motions proposed to end them. During the 1960s, for instance, filibusters were restricted primarily to a handful of regionally and politically sensitive ‘civil rights’ bills. Today, on the other hand, almost any major bill will inspire some talk of a filibuster, even if it does not materialise. Although there may be disagreement about whether a particular debate is, or is in the process of becoming, a filibuster—the number of cloture motions on which senators vote is a very imperfect measure of the number of filibusters—there can be no disagreement that filibusters are a more pervasive aspect of US Senate life than ever before.

[21]    These all are three-fifths votes of all serving senators—senators ‘duly chosen and sworn’—not just the senators present and voting. So the effect of a senator being absent when such a vote takes place is the same as if the senator were present and voting ‘No’.

[22]    Donald R. Matthews, U.S. Senators and Their World. New York, Vintage Books, 1960.

[23]    ibid. pp. 101–2.

[24]    Quoted in Stanley Bach, ‘Crisp, the Senate, and the Constitution’, Australian Journal of Politics and History, vol. 54, no. 4, December 2008, pp. 548–9.

[25]    Not surprisingly, minor parties that have no realistic prospect of being in government are more likely to be champions of Senate power and activism, at least so long as they hold seats there.

[26]    Unless, that is, the US president vetoes the bill and Congress fails to override the veto by a two-thirds vote in each house, or in the very unlikely event that the Governor-General in Australia exercises his or her constitutional authority under s.58 to withhold royal assent.

[27]    Most of the discussion that follows regarding Australia focuses on amendments and not on requests for amendments, but in most respects, the procedures governing disposition of requests are the same as those governing disposition of amendments.

[28]    The best known comparative and theoretical treatment of the subject is George Tsebelis and Jeannette Money, Bicameralism. Cambridge, Cambridge University Press, 1997. The standing Bundestag–Bundesrat mediation committee as well as the conciliation committees used by the European Union also are discussed by Thomas Konig and Bjorn Horl, ‘Bicameral conflict resolution: an empirical analysis of conciliation committee bargains in the European Union’. Paper presented at the 2003 annual meeting of the American Political Science Association.

[29]    The same US Representatives and senators who might otherwise be appointed to a conference committee on a bill may meet informally, much as they would as members of a conference committee, and reach an agreement on all aspects of the bill. This agreement then is presented to the House of Representatives, for example, as a House amendment to the Senate’s amendment to the underlying House bill. If the House accepts this new amendment, it then asks the Senate to do likewise. If the Senate concurs, the legislative process is complete. In such a case, the final version of the bill emerges from negotiations similar to those that occur in a conference committee, but without the formalities that conference committees and their reports involve. On the other hand, an agreement reached through this kind of informal alternative to conference is not protected, as is a conference report, from amendments when the House and Senate consider it.

[30]    For a discussion of conference committees elsewhere, see George Tsebelis and Jeanette Money, op. cit., ch. 8. The same authors point out elsewhere that reliance on the individual approach to resolution is the norm in international practice. Jeannette Money and George Tsebelis, ‘Cicero’s puzzle: upper house power in comparative perspective’, International Political Science Review, vol. 13, no. 1, pp. 25–43.

[31]    The second house to act on a conference report may not return, or ‘recommit’, it to the conference committee for revision, because the vote of the first house to agree to the report has the effect of dissolving the committee. Thus, there no longer is a conference committee to which the second house might otherwise recommit the report.

[32]    In the French case, which has been well-analysed in English by Tsebelis and Money, the process in broad brushstrokes is similar to that in Congress, in that a bill passed by one house then is sent to the other which in turn may propose amendments for the first house to consider. After both the National Assembly and the Senate have acted twice on a bill (or sooner if the bill is declared to be urgent)—the process of sending the bill back and forth is known in France as the navette or shuttle—the government may call for creation of a conference committee comprising equal numbers of members from the two houses. If the conference committee proposes a compromise that both houses accept, the legislative process is complete. However, if the conference committee fails to agree or if its report is rejected, the government can ask the National Assembly, to which it is responsible, to pass the bill in whatever final form the government proposes. In other words, the government and its parliamentary partner, the National Assembly, have the final word if they want it. See George Tsebelis and Jeanette Money, ‘Bicameral negotiations: the navette system in France’, British Journal of Political Science, vol. 25, 1995, pp. 101–29; and Jeanette Money and George Tsebelis, ‘The political power of the French Senate: micromechanisms of bicameral negotiations’, The Journal of Legislative Studies, vol. 1, 1995, pp. 192–217.

[33]    The report itself is accompanied by an explanatory document which details what the conferees mean by some of the new legislative provisions they propose, and how they expect executive branch officials to interpret and implement them. These ‘statements of managers’ have no legal force, but executive branch officials ignore them at their political peril.

[34]    See note 31 above.

[35]    US conference committees have been the subject of several books, including Ada C. McCown, The Congressional Conference Committee. New York, AMS Press, 1967 (reprint of original 1927 Columbia University Press edition); Lawrence D. Longley and Walter J. Oleszek, Bicameral Politics. New Haven CT, Yale University Press, 1989; Gilbert Y. Steiner, The Congressional Conference Committee. Urbana, IL, University of Illinois Press, 1951; and David J. Vogler. The Third House. Evanston, IL, Northwestern University Press, 1971. They also are discussed in various official House and Senate publications, the most accessible of which are Wm. Holmes Brown and Charles W. Johnson, House Practice. Washington, US Government Printing Office, 2003 (available online at; and Floyd M. Riddick and Alan S. Frumin, Riddick’s Senate Procedure. Washington, US Government Printing Office, 1992 (available online at Even more digestible are the materials available on the website of the House of Representatives’ Committee on Rules at Regrettably, the Congressional Research Service of the US Library of Congress does not make its relevant reports available to the general public. However, the most comprehensive of these reports—Resolving Legislative Differences in Congress: Conference Committees and Amendments Between the Houses, by Elizabeth Rybicki (Report 98–696; November 26, 2008)—can be found at–696.

[36]    One exception is the general requirement imposed by both the House of Representatives and the Senate that meetings of conference committees be open to the public. However, there is less here than meets the eye. It is quite possible, and not at all unusual, for conferees to meet privately—and, therefore, unofficially—at great length, and then have one brief formal—and public—meeting to announce the agreement that they have reached.

[37]    This does not include the handful of ‘private laws’ that were enacted to benefit named individuals or entities. Private bills are rarely, if ever, considered by conference committees.

[38]    Based on data made available by the Congressional Research Service.

[39]    There have been exceptions, to be sure, especially when there has been intense time pressure to reach agreement, because of the time it can take to appoint and convene a conference committee, prepare its written report, allow members of both houses some minimal time to study it, and then for each house to debate and vote on the conference report. See also the discussion of this subject later in this essay.

[40]    Indeed, there is no practical alternative, because the formal rules and procedures of one house differ in some fundamental ways from those of the other.

[41]    This is a political statement, not a procedural one. It has been said that conferees present the House and Senate with a ‘take it or leave it’ proposition, but that is not true. First, a conference report may be recommitted by the first house to consider it. And second, even if one house or the other defeats a conference report, that brings the bill back to the procedural stage it was at before being sent to conference. The House and Senate can agree to create a new conference committee or they can turn instead to an exchange of messages and amendments. That said, in practice the defeat of a conference report can mean the death of the bill, especially for the practical reason that at the end of a two-year Congress, there simply may not be enough time to continue negotiations over the bill.

[42]    House Standing Orders 373–84 and Senate Standing Orders 156–62. This discussion of conferences incorporates with revision part of my discussion of the subject in Platypus and Parliament, op. cit.

[43]    House Standing Order 162 and Senate Standing Order 127.

[44]    The effect of the two houses’ standing orders (House Standing Order 162 and Senate Standing Order 127) is that only the Senate may request a conference on a Senate bill that the House has amended and, conversely, only the House may request a conference on a House bill that the Senate has amended. Since most Australian legislation, and certainly almost all of the most important legislation, originates in the House, the decision to request a conference rests in practice with the House, not the Senate.

[45]    Similarly, ‘[c]onferences between the two houses provide a means of seeking agreement on a bill or other matter when the procedure of exchanging messages fails or is otherwise inadequate to promote a full understanding and agreement on the issues involved’. Harry Evans (ed.), Odgers’ Australian Senate Practice, 12th edn. Canberra, Department of the Senate, 2008, p. 541.

[46]    I. C. Harris (ed.), House of Representatives Practice, 5th edn. Canberra, Department of the House of Representatives, 2005, pp. 452–3.

[47]    There was another instance in which the Senate requested a conference on a Senate bill, the Social Services Consolidation Bill 1950. The House had amended the bill and insisted on its amendment, and the Senate had insisted on its disagreement to the amendment. However, the House did not agree to the conference.

[48]    On an informal conference as well as a 1903 proposal for a conference comprising all members of both houses, see also Evans, op. cit., pp. 541–2.

[49]    House of Representatives, Parliament of Australia, Guide to Procedures, 3rd edn. Canberra, Department of the House of Representatives, 2008, p. 84.

[50]    To reiterate, I refer here to the two different formal, procedural approaches to resolving bicameral disagreements. That the individual approach always is used in Canberra does not mean that the two houses always, or even often, consider each amendment in isolation from the others. It does mean that the Australian Parliament does not create formal negotiating forums to propose how to reconcile their legislative differences.

[51]    For example, John Ferejohn, ‘Who wins in conference?’, Journal of Politics, vol. 37, 1975, pp. 1033–46; John Carter and John Baker, ‘Winning in House–Senate conferences: the ‘revised theory’ and the problem of countertrending conferences’. Paper presented at the 1986 Annual Meeting of the Southwest Political Science Association; Lawrence Longley and Walter Oleszek, ‘The three contexts of congressional conference committee politics: bicameral politics overviewed’. Paper presented at the 1983 Annual Meeting of the American Political Science Association; Gerald Strom and Barry Rundquist, ‘A revised theory of winning in House–Senate conferences’, American Political Science Review, vol. 71, 1977, pp. 448–53; and David Vogler, ‘Patterns of one-house dominance in congressional conference committees’, Midwest Journal of Political Science, vol. 14, 1970, pp. 303–20. The outcomes of conference committee negotiations in state legislatures are the subject of Donald Gross, ‘House–Senate conference committees: a comparative state perspective’, American Journal of Political Science, vol. 24, 1980, pp. 769–78.

[52]    The reason why this question has no answer is that it depends on what each house actually hopes to win in conference. Elementary strategic calculations suggest the likelihood that one or both houses may pass a version of a bill that takes into account the version that the other house already has passed or is expected to pass. The Senate, for example, has been known to include provisions in bills that it fully expects to relinquish during its negotiations in conference with the House. Writing of Senate floor consideration in 1987 of a major trade bill, Stephen Van Beek found that ‘[m]ost of the 160 amendments the Senate considered were adopted without a great deal of debate, as Bentsen [the Senate committee chair] allowed his colleagues to include add-ons that only increased his leverage in conference’. In this way, Bentsen ‘loaded up the bill with proposals to bargain them away’ in conference. Stephen Van Beek, ‘Post-passage politics: the changing nature of bicameralism’. Paper presented at the 1992 Annual Meeting of the American Political Science Association, pp. 11, 14. The fact that the Senate ‘gives away’ such bargaining chips during conference negotiations may be misinterpreted as a victory for the House and a defeat for the Senate when, in fact, the Senate merely is trading away provisions it does not particularly want to retain in return for concessions from the House on provisions that really do matter to the Senate.

[53]    The recent growth, in both size and number, of voluminous and multi-focus omnibus bills also challenges efforts to arrive at a single answer to the question of ‘who won in conference?’. At best one can ask ‘who won’ with respect to one or more particular dimensions of such a bill.

[54]    Except for the very unlikely possibility that a money bill might be considered in a joint meeting of the members of both houses following a double dissolution.

[55]    That is, the appropriate minister or the government’s leader in the House of Representatives or the Senate.

[56]    Woodrow Wilson, Congressional Government. Cleveland and New York, The World Publishing Company, 1956, p. 69 (originally published in 1885).

[57]    This requirement often is waived by unanimous consent or suspended by a two-thirds vote, but almost always for relatively minor bills that enjoy broad bipartisan support. When the committee referral requirement is bypassed for important bills, it is almost always because of the need for extreme speed in passing the bill.

[58]    Stanley Bach, ‘Strengthening Australia’s Senate: some modest proposals for change’, in Papers on Parliament, no. 50. March 2009, p. 87 (available online at, drawing on John Vander Wyk and Angie Lilley, Reference of Bills to Australian Senate Committees, Papers on Parliament, no. 43, June 2005 (available at

[59]    As a general rule, congressional party leaders have not participated very actively and substantively in legislative debates. Their concern has been more with managing and expediting the procedural flow of legislative business. When a party leader in the House of Representatives speaks on a bill or an amendment, his or her purpose usually is to stress the importance of an issue to the party and to rally his or her troops as the time for voting draws near. In the Senate, party leaders have taken on a more active legislative role in recent years, both reflecting and contributing to the increased partisan polarisation that I mentioned above and to which I shall return below.

[60]    Evans, op. cit., p. 544.

[61]    Such an amendment often is called ‘an amendment in the nature of a substitute’, and proposes to replace everything after the ‘enacting clause’ of a bill—the legally mandated phrase at the very beginning of a bill that is necessary for the bill, once enacted, to have legal force.

[62]    This discussion passes over numerous procedural details, including some differences between the rules and practices of the House and Senate, that would unnecessarily complicate and obscure the argument that I am making here. These details are available in, among other sources, Elizabeth Rybicki’s Congressional Research Service report on Resolving Legislative Differences in Congress: Conference Committees and Amendments Between the House (Report 98–696, November 26, 2008), available at document/98-696.

[63]    A system such as this for numbering bills is a great convenience, but one that the Australian Parliament does not use.

[64]    These generalisations exclude the handful of private members’ and private senators’ bills that are introduced each year.

[65]    This discussion has benefited from the more extended Congressional Research Service report for Congress on the same subject by Walter J. Oleszek, Whither the Role of Conference Committees, op.cit. Regrettably, this report is not available to the public, but a version of his analysis appears as ‘Whither the Role of Conference Committees, Or is it Wither?’, Extension of Remarks, Newsletter of the Legislative Studies Section of the American Political Science Association, vol. 33, no. 1, January 2010, available at For a broader perspective on changing bicameral relations in Congress in light of changes in their context, see Edward G. Carmines and Lawrence C. Dodd, ‘Bicameralism in Congress: The Changing Partnership’, in L.C. Dodd and B.I. Oppenheimer (eds), Congress Reconsidered, 3rd edn. Washington, CQ Press, 1985, pp. 414–36.

[66]    There were two reasons why an appropriations conference committee would submit a conference report to the House and Senate with one or more amendments remaining in disagreement: either the conferees actually could not reach agreement on them, or the agreement they could reach on each of them violated their authority as conferees—for example, by exceeding the scope of the differences between the initial positions of the House and Senate. (See the discussion below on p. 35.)

[67]    As recently as 1997, Tsebelis and Money could write of US conference committees that ‘conferees act mainly as representatives of their chambers’. Tsebelis and Money, Bicameralism, op. cit., p. 203. I doubt that such careful scholars would make the same assertion today.

[68]    Quoted in Oleszek, Whither the Role of Conference Committees, op. cit., p. 12.

[69]    ibid.

[70]    It should not be assumed that these developments occurred in such a simple, linear progression.

[71]    When the Senate authorises its presiding officer to appoint Senate conferees, the presiding officer automatically appoints whatever list of names is presented by the party or committee leaders. Should the Senate not authorise the presiding officer to make the appointments, the Senate would have to elect its conferees, and the opportunities for delay that then would arise would be daunting indeed.

[72]    The House of Representatives also has typically agreed by unanimous consent to go to conference on legislation. If there is an objection to doing so, however, the House, unlike the Senate, can agree by a simple majority vote to a motion that accomplishes the same result.

[73]    Walter Oleszek cites one instance of this having been done in 1994, and several steps in the same direction that senators had taken during 1992–94. Oleszek, Whither the Role of Conference Committees, op. cit., pp. 9–10. Actually, the potential for delay is considerably greater. After the Senate authorises its presiding officer to appoint Senate conferees but before he or she does so, an apparently unlimited number of motions are in order to give non-binding instructions to the Senate conferees. Each of these motions is debatable and, therefore, subject to being filibustered. Thus far, it has not been thought necessary to exploit this opportunity.

[74]    In fact, there is no statistical evidence of a declining frequency of conference committees versus amendments between the houses during the period 1993–2007. Data recalculated from Walter Oleszek, Whither the Role of Conference Committees, op. cit., table 1 on p. 4.

[75]    Oleszek includes several instructive case studies in his analysis. Oleszek, Whither the Role of Conference Committees, op. cit., pp. 14–23.

[76]    However, these restrictions on the content of conference reports are not as stringent as they might seem at first. First, the House has procedures available to waive any of them by simple majority vote. Second, the Senate traditionally has adopted accommodating standards for enforcing these restrictions. (‘Rulings and practices in the Senate have left the chamber with a body of precedents that allow the inclusion of new matter as long as it is reasonably related to the matter sent to conference’. Elizabeth Rybicki, Senate Rules Restricting the Content of Conference Reports. Congressional Research Service report for Congress RS22733, 13 February 2009, p. 1.) And third, the Senate recently introduced a new procedure by which a violation of these restrictions need not prove fatal to the other agreements that the conferees reached.

[77]      Section 515 of the Honest Leadership and Open Government Act of 2007, Public Law 110-81, 110th Congress, enacted on 14 September 2007. Full text available at <>.

[78]      The fact that this ‘sense of the Senate’ provision was included in a bill that became law makes no difference. Under the constitutional authority of each house to make its own rules, any provision of law that affects only the Senate or the House of Representatives may be amended or repealed—or suspended, waived, or ignored—by that house alone, acting unilaterally and without the participation or consent of the other house or the president.

[79]    Fortunately, they are explained carefully in Elizabeth Rybicki’s Amendments Between the Houses: Procedural Options and Effects, Report R41003 for the Congress by the Congressional Research Service, 4 January 2010. Unfortunately, this report is not yet available to the public.

[80]      All this assumes that the same party has majorities in both the Senate and the House. Should there be divided party control within Congress, the political dynamics and strategic calculations would be considerably different.

[81]    The Majority Leader also offers an amendment to his own second motion—that is, his motion to concur with an amendment—in order to prevent anyone else from doing so.

[82]    The Majority Leader is able to do all this because the Senate’s precedents instruct the Presiding Officer to recognise him—that is, to give him the floor—in preference to any other senator who is seeking recognition at the same time. This is one of the few procedural powers the Majority Leader has, but it can be a critically important one.

[83]      From time to time, however, the Senate majority party may have its own incentive to rely on an exchange of amendments instead of sending a bill to a conference committee, if the ultimate agreement on the bill that the majority envisions would include elements that would violate the restrictions on the contents of conference reports.

[84]    E.g., David J. Vogler. The Third House. Evanston, IL, Northwestern University Press, 1971.

Back to top

Facebook LinkedIn Twitter Add | Email Print
Back to top