Note: This page contains links to PDF files. For more information, see the web page on Accessing files.
View the transcript as a PDF file - (PDF 242KB)
The paper that I am discussing with you today grows out of interests that were reflected in my 2003 book on the Australian Senate, Platypus and Parliament. More immediately, the themes I will develop here have been stimulated by the paper that I prepared in anticipation of my current visit to the ANU’s Parliamentary Studies Centre.[1]
That paper, which I shall revise with the benefit of comments I am receiving from some of you here today, examines the fate of Senate amendments to government bills during the past eleven years. Using data gleaned from the Senate’s annual publication on the business of the Senate, I have sought to trace the path that each Senate amendment took until its ultimate disposition. My purpose has been two-fold: first, to develop some empirical evidence about how often the Senate, when it had a non-government majority, evidently compelled the government to accept amendments to its legislation that it would not accept voluntarily; and second, to investigate how this record of Senate legislative influence has or has not changed since 1 July 2005, when the government assumed numerical control of the Senate for the first time since 1981.[2]
I will not review my findings in detail now because the full text of the paper will be available for any of you interested in reading it. With regard to my first purpose in looking at the documentary record, I will summarize a more complicated set of findings by saying only that I did indeed find evidence of the Senate’s influence on legislation, manifested in amendments that the House of Representatives initially rejected but that ultimately either were adopted or led to other changes in government bills that both houses accepted. With regard to the second purpose, I will observe simply that, from July 2005 to the end of 2006 at least, the change in partisan control of the Senate has been reflected in the virtual disappearance of Senate amendments that the government opposed but that nonetheless led to changes in government legislation. Although I have not yet had the opportunity to incorporate data for this year, it already is clear that the 2004 Senate elections did make a difference in the government’s ability to work its will in this building.[3]
I doubt that these summary conclusions will come as a surprise to anyone here today. But social scientists like to think that there is some value served by documenting the obvious, especially because every once in a while what we think to be obvious proves not to be. There remain, however, two related questions that the documentary record cannot answer. During the years of non-government control between 1996 and mid-2005, was the Senate’s impact on legislation a little or a lot, and was that impact good or bad? Answers to these questions depend not so much on our interpretation of any data, but on what we think the Senate should do—how it should exercise its powers and responsibilities as a participant in the national legislative process. This judgment, in turn, depends on how we believe national policy decisions should be made under Australia’s constitution or under any other reasonably democratic constitution. There are few questions more fundamental to the governments under which we live, whether in Australia or the United States.
Before going any further, let me acknowledge the bias that I bring with me today. In 2000, in connection with the Commonwealth’s centennial commemorations, Professor Elaine Thompson, who introduced us to the concept of Australia’s “Washminster mutation” of the British and American models of government, wrote a paper for the Parliamentary Library on the first century of Australian parliamentary democracy. In her paper, she commented on how the role of the Senate has changed in recent decades, She concluded by writing that “[t]here is still no convention concerning the limits of the Senate’s powers with respect to the Executive. Indeed it is reasonable to suggest that there is a political convention developing which expects the Senate to play a restrained, but nonetheless active role as a second chamber reviewing and, on occasion, rejecting government.”[4]
This “restrained, but nonetheless active role” is what I had in mind when I referred in the lecture I gave here in early 2003 to “the delicate balance” in the Australian political system. I think it fair to say that there is no dearth of people in this building—especially those who walk on blue carpet, many who walk on green carpet, and even some who walk on red carpet—who will willingly and enthusiastically make the case for restraint. It should come as no surprise, on the other hand, when I say that, as an American who spent most of his professional life working for the U.S. Congress, I prefer to make the case for activism.
Let me summarize the argument I am about to make by saying that there are at least three basic models of the decision-making process that can characterize the way in which democratic—or shall we say republican?—regimes work. One of those models is, to my mind, flawed on both empirical and normative grounds, especially when applied in the sweeping way that its proponents often advocate. The second of the three models is a chimera whose advocates are in need of a radical platitudectomy. Instead, it is the third model that I intend to advocate as being best suited to preserving and improving the long-term health of my society and perhaps yours as well; but it also is a model that is under threat in my country and possibly discredited in yours.
I will proceed by discussing each of these models in turn, asking whether any one of them is more likely than the others to characterize how democratic governments can and should work. In the process, I also will make some comments about the implications of my argument for the United States, and I will even venture some thoughts about what it might mean for Australia.
MANDATE MAJORITARIANISM
The first model that I want to discuss posits that, by the very definition of what constitutes a representative democracy, government policies are to be decided by the majority of representatives who, in turn, speak for and act on behalf of a majority of the electorate. To the extent that constitutions prescribe decision-making by majority vote, this may seem to be little more than a truism. Yet majority control of policy inevitably is constrained—by constitutions, as they may be interpreted by courts; sometimes by requirements for super-majorities for some purposes, such as to override an executive veto of legislation, as in Washington; and always by the inescapable need for unelected officials to make policy themselves in the process of filling in the interstices of enacted law.
Furthermore, the kind of majoritarianism I have in mind goes beyond the recognition that democratic constitutions typically give a majority of MPs the power to work their will. The more ambitious form of majoritarianism, what I will call mandate majoritarianism, begins with recognizing the power of the parliamentary majority, but then moves on to contending that the majority has the right and the obligation, as well as the power, to control parliamentary decisions, because that is what the voters expect and demand. At the moment, my argument is that mandate majoritarianism rests on a collection of assumptions about the behavior and motives of the electorate for which there is little empirical support. Later I will take the argument further by suggesting that this model of decision-making would be undesirable even if there were an adequate empirical basis for it.
I critiqued the notion of electoral mandates at some length in Platypus and Parliament, so I will content myself today with summarizing that critique and the empirical reasons for doubting the existence of any such mandates.
In summary, the argument for mandate majoritarianism is this: As each parliamentary election approaches, each of the parties contesting it puts forth a catalogue of specific and detailed programs that it will enact and implement if entrusted with the power of government. Each voter then studies and evaluates all of these catalogues (otherwise, platforms or manifestos), selects one because he or she endorses all its elements, and then votes for that party. One of the parties receives a majority of votes that translates into a majority of seats in the parliament, or in the house that monopolizes power if it is a bicameral parliament. That party then enacts into law, and without substantive change, the programs enumerated in its electoral catalogue of promises and, once those programs are enacted, the governing party proceeds to implement them in the manner it deems most consistent with the commitments it had made to the electorate.
Where this argument goes beyond recognizing the parliamentary majority’s power to control parliamentary decisions, if that majority is sufficiently unified and determined, is in contending that the majority has both a right and a responsibility—it has a mandate—to do so. In entrusting a party with a majority of seats in parliament, the electorate thereby gives that party not only the power but also the right to enact and implement its program because the election results constitute a blanket endorsement of that program. Moreover, this is more than a grant of discretionary authority to act. The governing party is obligated to enact and implement its program, again without substantive change, because a failure to do so would constitute a breach of trust with the electorate. The party must do what the people have elected it to do. It not only has been given a mandate to govern, it has been mandated to govern.
Two elements of this argument may seem unreasonable: first, that the party must enact and implement each and every one of the policies and programs it advocated during the election campaign; and second, that it must enact and implement them without substantive change. You may think that these are unnecessarily restrictive requirements. Upon closer examination, though, it should be clear that removing these elements from the argument would leave the majority party, once in power, with so much discretion that the individual voters could have no real confidence that they really knew what package of policies and programs they would be “buying” if their party were victorious. If that party, once elected, could pick and choose from among the promises it made during the campaign, or if it reserved to itself the right to embody a general campaign promise—to “control the growth of government”, for instance—in any one of a myriad of legislative forms, the linkage between election and governance would be too weak and unpredictable to be very meaningful.
It should be noted that the potential applicability of mandate majoritarianism is limited. All else aside, it makes sense as a model only when the nation’s electoral system is very likely to produce single-party majorities in government or, as in the case of Australia, the possibility of a majority composed of a stable and durable coalition. This is most likely though not inevitable when MPs are elected from single-member constituencies. Such systems tend to encourage two-party competition—and, therefore, a clear electoral victor—because the election in each constituency is a winner-take-all contest in which parties that are unlikely to win have trouble convincing voters to support them. (Preferential voting weakens the strength of this argument, of course.) Those systems also tend to give the party winning the most votes more than its proportionate share of parliamentary seats, thereby increasing the likelihood that there will be a single-party government.
In presidential-congressional systems, there is the possibility of divided partisan control, with one party controlling the executive and another controlling one or both houses of the legislature. In fact, this has been a common condition in the United States during the past half-century. Under these circumstances, no party can make a convincing claim for a mandate to govern, although American presidents routinely do so when confronted by a Congress in which the other party has a majority in the Senate or the House of Representatives. So too in parliamentary systems in which the party constituting the government does not hold a majority of seats in parliament (or in the house of parliament that matters). If there is a minority government, and non-government parties and independents comprise a majority in parliament, the governing party has no reasonable claim to an electoral mandate. By the same token, if there is a majority coalition government, the only case in which there can be a plausible argument for a mandate is if the coalition is formed before the election and the participating parties agree to campaign on a shared and well-publicized platform. Otherwise, voters cannot know with confidence exactly what catalogue of policy commitments they are endorsing with their votes.
My core argument, however, is that even when mandate majoritarianism is possible, there is little empirical evidence to support the existence of electoral mandates. Let me point to some of the assumptions about parties and voters that are implicit in this model and encourage each of you to ask yourselves whether they truly are characteristic of the parties and voters you know.
First, claims of election mandates require that the parties seeking such a mandate must explain to the voters during the campaign just what they would do if entrusted with control of the government. I reviewed the documents prepared by the Liberal and Labor parties in anticipation of the 2004 election, and I was genuinely impressed by the depth and breadth of information that both parties offered the voters, certainly in comparison with the U.S. party platforms that are adopted when we nominate our presidential candidates and then are immediately forgotten. If the Liberal party documents that I examined were typical of all the others, voters in 2004 were presented with roughly 1,500 pages of policy explication that presented the party’s intentions for the three-year period now ending. And since voters are required to choose among two or more parties, we need to add, at a minimum, the several hundred pages that comprised the 2004 ALP platform.
So we might say that Australia’s parties meet the first requirement for election mandates, or at least that they come closer to meeting it than political parties typically do. However, that in itself poses a dilemma. If the party’s manifesto is cast in more general terms, there can be all kinds of ways to write bills that arguably would implement the various items in it. So the voters might know in general terms what legislation to expect, but not the specifics. And, as we all know, in legislation as in so many other things, “the devil is in the details” (a phrase which, incidentally, has been attributed to such luminaries as Michelangelo and Flaubert, among others). But if, on the other hand, the party platform truly is so specific and detailed that it can be translated into legislative language without difficulty or ambiguity, then the governing party’s commitment to fulfilling its mandate leaves it little or no room to adapt to changing circumstances. No theory of electoral mandates can pass the proverbial “giggle test” if it asserts that a party has the right and responsibility to implement its program of policy promises, but if it also contends at the same time that, of course, the party has to be able to make whatever changes in that program it considers necessary.
Second, mandate claims also must assume that the voters actually understand and evaluate the various parties’ plans and promises. The sheer size of the Australian party manifestos, with their supporting white papers and other documents, gives me absolutely unshakeable confidence that, if comparable documents actually were presented by U.S. parties, very few of America’s voters would have more than the vaguest idea about what is in them. Yet when the victorious party in Australia subsequently insists that one of its bills must be enacted and should not be changed, it can point to a paragraph or bulleted point in these documents as proof positive that a majority of voters must want the bill enacted because the party advocated it during the election campaign and the party won the election, from which it is supposed to follow that the electorate thereby endorsed that particular campaign promise. And if any government feels free to make the same argument with respect to each of its campaign promises (although it would be impossible to enact all of them into law before the next election), it must assume that the electorate understood and supported each and every one of them. This assumption, too, is so implausible on its face that simply stating it suffices to refute it.
Third, the mandate theory assumes that voters base their election day decisions on their evaluations of the parties’ respective programs. In the United States, we know that this is not even remotely the case. Voters may prefer one party’s general approach to domestic and foreign policy to the other’s, but they also are influenced, and often more so, by such factors as their parents’ voting history, their own long-standing party loyalties, and, increasingly in the era of television, how much they like and trust individual candidates. We used to joke about “yeller dog Democrats”—voters who would vote for a yellow dog so long as it was the Democratic party’s candidate. And American voters who claim that they don’t routinely vote for one party or the other proudly claim instead that they vote “for the man, not the party”—and, therefore, not the party platform.
Voters whose choices do reflect their strongly-held policy views often are concerned intensely with one issue, whether it be abortion or gay marriage or gun control or immigration or the war in Iraq, and they pay much less attention, and give much less weight, to the others. Furthermore, voting in America often (perhaps typically) is retrospective; voters’ decisions are based on evaluations of how the party in power has performed since it took office, not on what it promises to do in the future. If most voters are not content with how the government has been performing, they are unlikely to be persuaded by its promises for the future. On the other hand, if most voters are satisfied with recent government performance, they may be reluctant to throw the governing party out of office in favor of another party that can offer only assurances of its good intentions.
In short, the voting behavior of Americans is entirely inconsistent with the assumptions and requirements of mandate theory. I venture to think that much the same may be true of the Australian electorate.
It seems to me, then, that on empirical grounds alone, mandate majoritarianism is a deeply flawed model of decision-making, even when the majority of the electorate has voted for the majority in parliament. It is much more difficult to defend the model when the parliamentary majority does not receive a majority of the popular votes—when the majority of votes goes instead to the opposition and to other parties (and independent candidates) that are not in government. In Australia, for instance, neither Labor nor the Coalition often wins a majority of the votes cast in elections for the House of Representatives. In fact, in the 24 elections since 1949, only three times has the party winning control of the House and, therefore, the government, won a majority of the vote, and on two of those occasions it won by margins of 50.1 and 50.2 percent. What becomes of the argument that the government has both the right and the responsibility to have its legislative program enacted as it sees fit when the majority of the electorate, by that same theory, had rejected that program at the most recent election?
CONSENSUALISM AND ITS ALTERNATIVE
I believe the second model I wish to discuss, decision-making by consensus, is equally flawed, and for the same fundamental reason: it makes assumptions and imposes demands that, most of the time, simply are unrealistic.
In the United States, we frequently hear our elected officials say that we need to reach a consensus on how to address the pressing issue of the moment, whatever it may be. The goal should not be to enact the policy prescriptions of one party or political tendency—usually progressives versus conservatives—rather than those of the other (or others). Instead, the goal should be to bring together both or all parties in the legislature, and to bring together the legislature and the executive, in support of a policy decision that all recognize to be the right thing to do.
Let me share with you a few examples that I gleaned from the Congressional Record, Washington’s equivalent to Hansard, for January and February of this year, during the first days that the Democrats once again had majority control of both houses of Congress.
The Democrats in the House of Representatives flexed their new-found political muscle by passing a collection of bills during the first one hundred hours of session, the implication being that they were improving on the record of the House Republicans when they had taken power in 1995 and passed their collection of favored bills during the first one hundred days. Both the content of the bills that the House passed in early January 2007 and the procedures by which they were considered were frequently contested, and bitterly so, by the House Republicans, who once again were learning the pain and frustration of being in the minority.
One such bill concerned the highly contentious issue of human embryonic stem cell research which, for many, raises the specter of abortion. During the debate, the new Republican leader in the House criticized the Democratic bill and the speed with which it was being propelled through the House by praising what he described as an alternative that “offers the potential for a new consensus approach” to the issue. About a week later, at the conclusion of those first one hundred hours, one Democratic Representative felt impelled to announce that “[w]e have set a tone for the 110th Congress that is one of cooperation, consensus, and compromise that extends beyond party lines.” I recall no such announcements from the Republican side of the House chamber.
Soon thereafter, the new Democratic floor leader in the House of Representatives spoke in a debate about re-adopting a procedural rule that the Republicans had repealed. The rule had allowed, and now again allows, the delegates in the House who represent the District of Columbia and America’s other territorial possessions to cast some votes on the House floor. Referring to two of the most expensive U.S. government programs, Medicare, which provides health insurance for seniors, and Social Security, which provides income support primarily for seniors, the Majority Leader proclaimed for the Democrats that the “residents of the five territories should have a voice in shaping a bipartisan consensus that shores up the financial health of these vital programs.” He said this, notwithstanding the fact that Social Security regularly is described as “the third rail of American politics”[5]—to touch it is to risk almost certain political death—as well as the universal recognition that Democrats look for every possible opportunity to accuse Republicans of wanting to cut, gut, privatize, or otherwise attack the Social Security program.
Such paeans to consensus were not limited to the House of Representatives by any means. In February, the senior Republican Senator on the Finance Committee, with responsibility for reviewing and recommending bills affecting taxes, was discussing a provision of the income tax code known as the “alternative minimum tax.” The distinguished Senator wanted to “remind people,” he said, “that in 1999 we passed a repeal of the alternative minimum tax, but President Clinton vetoed it and we haven’t been able to repeal it since....” Moments later, though, he went on to assert that “[t]here is a bipartisan consensus that only complete repeal is an adequate solution to this problem” of the alternative minimum tax.
Finally, what was the most contentious issue in American politics in early 2007, and now for that matter? The war in Iraq, of course. You may recall that the House of Representatives adopted a resolution expressing the opinion, without attempting to embody that opinion in law, that the President’s troop “surge” was not a good idea. The Democrats in the Senate, with the support of a handful of Republican colleagues, attempted to bring a similar resolution to a vote, but they were stymied by a filibuster—a debate prolonged indefinitely—that was supported by most Republicans. One of the leaders of this debate was the former Republican chairman of the Senate’s Committee on Armed Services, who had broken with the President on this issue and who had been instrumental in drafting a resolution that the Democrats ultimately supported in opposition to sending more troops to Iraq.
During the debate, this Senator emphasized that he and his allies had no intention of promoting legislation that could in any way jeopardize the safety and well-being of American military personnel already in Iraq. He insisted that “[w]e solidly support that concept of no cutoff of funds.” “What do we do short of that?” he continued. “Well, we have a debate. Somehow you have to have some focal point, something written down, some document in writing as to the ability of this institution, the Senate, to reach a consensus, and a bipartisan consensus, on how best we go forward with a new strategy in Iraq.”
There is one thing that all these references to consensus have in common. They all are nonsense.
I always have understood “consensus” to refer to a meeting of the minds—a group of people all coming to a common understanding about something.[6] That agreement may be the result of a collective process of deliberation. Or it may be that each member of the group deliberates independently and then they come together to discover that they have reached the same conclusion. Whatever the process, there is implicit in the notion of consensus, to my mind at least, the idea that all members of the group, or at least the overwhelming majority of them, share the same understanding as to what is good, or what is right, or what is the best thing to do. Central to any consensus is, first, that it is supported by all, or almost all, of those involved, and, second, that they support it by choice, not because they are in any sense constrained or compelled to do so.
Conceiving of consensus in this way immediately reveals just how unlikely it is for us to expect to find a consensus on almost any issue of national significance that engages the attention of the Congress in Washington or the Parliament in Canberra. Political decision-making rarely is a process of politicians reasoning together until they all agree that there is a right answer to the question before them. In support of this contention, I need only refer to the thought of a distinguished but unrecognized American philosopher, my father, who used to say that when two people always agree, it’s certain that one of them isn’t thinking.
I will return to this theme shortly. In a practical sense, though, the line of argument I’ve pursued on this subject is irrelevant. Did that Republican Representative really believe that there was a consensus about stem cell research that was waiting to be revealed and embraced? Did that Democratic Representative really expect that his chamber would discover a consensus about how to address the exploding costs of Medicare and Social Security? Did that Republican Senator really believe that there was a consensus in favor of repealing the alternative minimum tax when recent and repeated attempts to repeal it had failed? And did his Republican colleague really expect that out of the Senate debate about Iraq would emerge a common understanding among Senators about what to do there? Did they really believe that consensus was likely, or even possible, in light of the different values that their colleagues hold most dear and the different interests, preferences, and needs of the people they represent? Of course not.
Instead, or so it seems to me, these references to consensus imply almost the opposite: that the problems under debate were so difficult, so divisive, and so intractable that, instead of confronting and addressing them as best they could, it was far easier for these legislators to talk about how wonderful it would be somehow to find solutions that everyone would prefer to the alternatives.
Let me take a momentary detour that, whether you believe it or not, will get me where I want to go. One effect and, to my mind, a benefit of electing MPs from individual constituencies is that each is linked to a particular geographical area and the people who live there. Consequently, it is possible to address MPs in debate not by name but by reference to the constituency—district or state—that each represents: not as “Mr. Jones,” for example, but as “the Member for Buncombe” (which, incidentally, is in North Carolina, and is the original source for “bunk”). This is the practice in the British House of Commons as well as the Australian House of Representatives, and, though sometimes honored in the breach, the U.S. House of Representatives as well.[7] Although this form of address sometimes sounds stilted and artificial to visitors, as do many of the formalities of parliamentary practice, it is explained and justified on the grounds that it de-personalizes debate and reduces the level of animosity that otherwise might develop in the chamber. While that may be true, I think it also serves another related but distinguishable purpose.
It has been said that members of a durable parliamentary assembly need to believe and remember that the members of other parties may be opponents but they are not enemies. At the extreme, representative government is all too likely to collapse if members of parties or parliamentary groups believe that their personal well-being and security, and those of their supporters, are in jeopardy because they are in the minority. (Those who are convinced they are bringing democracy to Iraq might bear this in mind.) The concept of “enemy” evokes images of war with victors and vanquished; the concept of “opponent” evokes images of a game with winners and losers, but a game that will be played again and again so that today’s loser can hope to become tomorrow’s winner.
One reason parliamentarians find themselves opponents is because they have fundamentally different philosophies of government (or ideologies, if you prefer). What is the appropriate role of government in the society and economy, for example, and to what extent and for what purposes should the government intervene in the choices and behavior of individuals as well as collective entities such as corporations? Another reason is that parliamentarians may have different understandings of how the world works. They may differ, for example, over whether international disagreements are amenable to negotiated resolutions or whether international actors respond only to the threat or application of force. Similarly, they may differ over whether helping the disadvantaged in society is best done by targeted government programs that, by definition, do not benefit everyone equally if at all, or whether economic growth and a favorable business climate is the surest way to promote prosperity in which all will share. On their most dispassionate days, it even may be possible for MPs to agree that they seek the same ends even if they have fundamental disagreements about the best means to achieve them.
There is a third reason why MPs disagree which is related to the other two but which is reflected in the impersonal and indirect way in which they often are expected to address each other. Under most party and electoral systems, including the form of proportional representation used in Senate elections, MPs represent geographic constituencies and those constituencies may have different needs and interests. MPs usually are expected to represent those needs and interests in the sense of speaking and advocating for them: “I rise to speak for Buncombe,” as that Congressman is supposed to have said in 1820. Even in systems that elect all MPs from single national party lists, the parties may assign MPs of their party to develop strong ties with a particular community or region, perhaps the one in which each MP resides or was born. Generally speaking, we can expect this linkage between MP and a constituency to be minimal when national party organizations control the selection and re-selection of parliamentary candidates to the virtual exclusion of local influence, and when citizens base their voting decisions on national issues, without regard to how the various parties’ programs would affect their local areas.
If we accept a constituency linkage as a typical characteristic of parliamentary life and work, we also must accept that MPs’ constituencies differ and that, very often, the differences among them are not differences of ideology or worldview; they are measurable differences. Some constituencies are richer than others; some have an atypical racial or ethnic composition; some depend more on industry and others more on agriculture; some rely more than others on exports for jobs and local prosperity. These possible differences could be multiplied. But the point is that, because of such differences, many proposed laws will benefit some constituencies more than others, and sometimes they even will benefit some constituencies at the expense of others. When one MP refers to a colleague who disagrees with him by referring to that colleague’s constituency, it is a way of reminding the MP that his colleague may be taking a contrary position in order to reflect and promote the real interests of his constituents.[8]
Examples abound. Free trade can benefit consumers by increasing the availability of lower-cost imports—clothing made in China or Bangladesh, for instance—but that competition can cost workers their jobs if the textile plants where they work cannot compete successfully. Farm price supports can help keep family farmers in business, but only at the cost of higher food prices and higher taxes. Increasing corporate taxes can reduce the need to increase individual income tax rates, but only at the cost of reduced corporate profits and, it is argued, reduced capital investment and shareholder value. Drilling for oil in the Alaskan wilderness may eventually help control how much it costs me to heat my home and fuel my automobile. And so on.
It does not follow from examples such as these that politics is always a zero-sum game—that for every winner there is a loser. It does follow, though, that government policies and actions often have differential effects on different groups and regions and, therefore, on different constituencies. The challenge of law-making rarely lies in deciding who’s right and who’s wrong. Instead, the challenge usually takes the form of having to strike the most appropriate balance among competing needs and interests, even perhaps in debates over abortion: preserving the life of the mother versus protecting the life of the child. That is why a true consensus on important policy choices rarely is possible, because consensus implies a virtually unanimous agreement as to what is right, not what is the best we can do under the prevailing circumstances. And, returning to my earlier discussion, that is also why I believe that mandate majoritarianism is undesirable on normative grounds, in addition to being unrealistic on empirical grounds.
COMPROMISE AS A VIRTUE
When I was a boy, my schoolmates and I had to memorize the Preamble to the U.S. Constitution, which is so brief I’ll take a moment to read it to you:
We the people of the United States, in order to form a more perfect union, establish justice, insure domestic tranquility, provide for the common defence, promote the general welfare, and secure the blessings of liberty to ourselves and our posterity, do ordain and establish this constitution for the United States of America.
These words had become so familiar to me that it was only years later that I stopped to think about what they mean, and about the act of faith that my constitution represents: that the particular set of political institutions it created somehow would lead to the achievement of all those goals laid out in the Preamble, such as establishing justice and providing for the common defense. The authors of Australia’s constitution saw no need to include such a statement of goals in the preamble to their document, but I’m confident that the same assumption can be attributed to them: that the structure they were creating was more likely to promote the same future for Australia than any other on which they might have been able to agree.
It is on two of these goals that I want to concentrate for a moment, and those are the goals of promoting the general welfare and insuring domestic tranquility, or social harmony as we might call it today. In brief, my argument to you is that majoritarianism based on claims of electoral mandates is not the most promising basis for promoting the general welfare or for insuring domestic tranquility in the long run.
If by majoritarianism we mean only that law-making decisions are to be made by majority vote, then, as I said earlier, we are almost defining a core element of what we mean by democratic governance.[9] Mandate majoritarianism is another matter altogether, because it transforms the power of the governing party (or coalition) to work its will in the parliament into both a right and a responsibility to do so. If that party is entrusted by the voters with the power of government, it is only so that it can implement the legislative program that the majority of the voters endorsed when they cast their ballots. Any concessions to the party or parties that are not in government would come at the expense of the government’s ability to fulfill the mandate it sought and received.
What then of the needs and interests of those constituencies whose representatives are not members of the governing party? Defenders of electoral mandates would argue, I suppose, that the victorious party already had taken those needs and interests into account in the process of formulating the program it presented to the voters. This is true to some extent, I’m sure. I would not want to imply that any responsible political party would deliberately ignore the needs and interests of any numerically significant segment of the population. However, I do question the notion that the leaders or members of any party are the ones best able to decide what should be done for or to those groups—whether regional, economic, social, or whatever—that largely opposed their party in the past and are likely to oppose it in the future.
It is an underlying tenet of democratic government that the people themselves are the ones best able to determine their needs and interests and what government actions will best serve them—what will promote their welfare as part of the general welfare. When it is impractical for us as citizens to speak for ourselves as our laws are being made, we rely on the people whom we have elected to speak for us. If all those who voted for non-government parties and independents believe that, once the election is over, they have no effective voice in government decisions until the next election rolls around, this situation cannot, in my judgment, effectively insure domestic tranquility. On the contrary, long-term social harmony benefits from a generally-shared belief that the needs and interests of all segments of the population are being expressed forcefully in parliament and that their representatives are able to have a modicum of influence over the decisions made there.[10]
This finally brings me to the third of the three models of democratic decision-making to which I referred at the outset. If mandates are a myth, and a dangerous one at that, and if consensus is a chimera, then compromise is a virtue. The alternative to mandate majoritarianism is not “consensualism,” it is the recognition that compromise is a good thing.
Here I am juxtaposing compromise against consensus. If the search for consensus is the search for what is right, the search for compromise is the search for what is best or most generally acceptable (which may or may not be the same thing) under the prevailing circumstances. Compromise requires the governing party to accept some limits on doing what it would like to do, and what it has the numbers and the formal constitutional authority to do, in order to take into account—legislatively, not just rhetorically—the needs and interests of those who voted against it. As my authority on this point, I will quote John Stuart Mill, who wrote that “[o]ne of the most indispensable requisites in the practical conduct of politics, especially in the management of free institutions, is conciliation: a readiness to compromise; a willingness to concede something to opponents, and to shape good measures so as to be as little offensive as possible to persons of opposite views.”[11]
I don’t mean for a moment to suggest that the minority, or the Opposition, should have as much influence over legislative decision-making as the majority, or the government. What I do mean to suggest is that the long-term interests of a nation are best served when the governing party has to ask itself what changes in its legislation it can accept that will ameliorate the detrimental effects that bill may have on certain constituencies, whether defined geographically or otherwise, without sacrificing the principles it has committed itself to promoting. And if the governing party is not inclined to view such compromises as desirable, which is what I would anticipate in the real world of politics, then it is good if the nation’s political institutions give some person or entity the power to make the government accept them as necessary.
Let me illustrate my contention with one concrete example. In the United States, there has been a recurring debate over whether Congress should increase the statutory minimum wage, the hourly wage that federal law requires employers to pay most of their employees. Most congressional Democrats have argued that the minimum wage has been too low for many families to support themselves adequately. Most congressional Republicans have argued that increasing the minimum wage would put too much pressure on many small businesses (which, they argue, are the engine that really drives job creation in the U.S.), forcing them to lay off employees or even close their doors. I’m not competent to say how many people would benefit from the wage increase or how many jobs or small businesses would be lost. But it does seem reasonable to me to believe that there is some truth on both sides of the argument. The pity is that it has been so difficult to agree on a compromise that hardly rises to the level of rocket science: coupling a minimum wage increase with tax breaks for small businesses; in effect, making the taxpayers pay for part of the cost of the minimum wage increase. Such a compromise—socializing part of the costs of a new or increased government benefit program, and thereby offsetting a focused benefit with a cost that is so widely distributed that it cannot really be felt by those paying for it—is an approach that often is available to government policy-makers.
COMPROMISE BY CHOICE OR NECESSITY?
The last issue I want to address today is how such compromises are to be achieved.
In this context, let me first note that compromise seems to be taking on more of a pejorative connotation in Washington these days, sometimes being equated with the abandonment of principle. This reflects, I believe, the increased polarization that has come to characterize American politics in recent years, as each of the two parties has become more homogeneous internally and as the policy disagreements between the two parties are thought to have widened. Party unity in congressional voting has increased as has the level of vituperation in political discourse. Compromise has become more difficult as the members of each party either have come to believe their own rhetoric about the other or they have become so trapped by their rhetoric that they are unable to justify compromises they otherwise might be willing to make. Furthermore, it is so much easier and perhaps even more satisfying for Representatives and Senators to use legislative debates as a way of appealing to their supporters in anticipation of the next election than it is to negotiate compromises with their political opponents from whom they have become increasingly estranged.
Although I have been away from Australia for four years now, it also seems to me that Australian politicians, at least those in government, often are reluctant to speak publicly about legislative compromise. The tenor of question time and the tone of each major party’s public comments about the other would be hard to reconcile with a visible approach to legislating that revealed the government to be taking the Opposition’s concerns seriously and accepting amendments that took them into account, while not satisfying them fully. As I documented in 2003, the voting record in the Senate does not reveal a government and Opposition always opposing each other; it is striking, in fact, how often the government and Opposition voted together on divisions during 1996-2001. That, however, is a hidden story of Australian politics. The kind of compromise approach to law-making that I am advocating involves a public recognition that compromise is taking place and that those compromises benefit the nation.
It is often said in my country, and I would not be surprised if it were equally true in Australia, that the time horizon of politicians is limited to the date of the next election. What may contribute to the long-term health of the polity (the general welfare) and its domestic tranquility is going to be less important in practice to most elected officials than what will contribute to their shorter-term electoral success and their ability to promote their values and the interests they represent while they enjoy the power to do so. I acknowledge, therefore, that the winners of parliamentary elections are unlikely to concede voluntarily any of the fruits of their victories to the losers: the minority in the U.S., the Opposition in Australia. If so, the question then becomes whether there are conditions under which policy compromise is likely to be necessitated by the structure of political systems or the dynamics of electoral competition.
In a presidential system such as mine, the constitutional divide between the executive and the legislature can promote a competition for influence and an unwillingness of the congress to accept the president’s legislative proposals without change.[12] That is so even when the same party controls both branches of government. It is even more likely to be the case when the constitutional divide is exacerbated by divided party control, which has been more the norm than the exception in the U.S. during the last half-century.[13] When these factors are supplemented by relatively weak party discipline and an electoral system that encourages legislators to think first about how their electoral constituency, not their party leaders, want them to vote, policy compromises become almost inescapable, even when compromise is lamented as a necessary evil, not as a positive good.
The problem we have been having is the reluctance of the president to accept the inevitable need for compromise, and the temptation for members of the other party in Congress to insist on too much. As a result, we have had much too much talk of consensus and much too little willingness to engage in the search for practical compromise.
On the other hand, when both branches of government are controlled by the same, highly disciplined party whose unquestioned leader is the president who controls the political future of his party’s legislators, it would be surprising for him to sanction any voluntary concessions to other parties and the constituencies and interests that they represent. This has been the problem encountered in some quasi-, pseudo-, or proto-democracies, especially in Africa, where the dominant party has been largely the personal vehicle of its dominant leader, the president, and the legislature has thought of itself as a parliament, not a congress.
In a parliamentary system characterized by two dominant parties, there is little reason to expect a compromise approach to legislating, especially if each party is quite homogeneous and the two parties are quite polarized. One case in point has been Bangladesh, where there is raw hostility between the two parties, even though the policy differences between them are more difficult to pinpoint. When the leader of each party has accused the other of being complicit in murder, we should not expect them to relish the prospect of compromise decision-making.
In a multi-party system, which often is the product of proportional representation, we need to look at how and when governing coalitions are formed. Such a coalition can be the product of pre- or post-election negotiations among disciplined parties, negotiations that culminate in a specific and detailed political “treaty” that binds the participating parties to a legislative agenda from which the dominant coalition partner must hesitate to diverge for fear of losing its parliamentary majority. If political parties are much weaker, on the other hand, and if coalition negotiations focus more on the allocation of portfolios and patronage, the dominant partner has more room for maneuver and more opportunity to make policy compromises if it chooses to do so.
It probably is fair to say that a compromise approach to legislating is most likely when parties are weakest and legislative majorities have to be assembled one vote at a time. I understand that, in the Kingdom of Jordan, for example, there are no parliamentary parties worthy of the name. MPs are elected as independents and do not then coalesce into unified and stable parties. This situation might seem to give the advantage to the government (putting aside the fact that the King really is the de facto policy-maker) because it is unlikely to face any organized opposition. However, I have heard it argued that the actual result is just the opposite, because the government does not have a parliamentary majority on which it can rely or that it can hope to mobilize. Instead, it must find the votes it needs where it can, and this situation allows each MP to bargain for whatever concessions he is able to extract.
The title of my paper speaks of mandates, consensus, compromise, and the Senate. What does this discussion of the first three mean for the fourth? I think the answer will be obvious to those of you who have followed my argument thus far.
In 2003, I argued that the Senate constituted the forum in which the Australian government can best be held accountable for its actions and decisions. The House of Representatives, controlled as it is by a highly disciplined government majority party whose members’ political futures depend largely on public support for their party’s government, have little practical incentive to question, probe, challenge, oversee, or investigate it in the public forums of Parliament. However much government party members might deny this, it is hard to deny the political logic that challenging their own party’s government in public can only damage the party and their own political prospects.
The Senate, on the other hand, has been in quite a different situation precisely because it has had a non-government majority for most of the last half-century. The combination of close party competition and the form of proportional representation that has been used in Senate elections since 1949 has put the government, whether Coalition or Labor, at a disadvantage in Senate elections, to the extent that the results of the 2004 Senate elections came as a surprise to many (including, I suspect, members of the Coalition government itself). Between 1996 and mid-2005, the non-government Senate majority had the institutional control, the constitutional power, and the political incentive to hold the present government to account in a way that could not reasonably be expected of the House of Representatives. Although, in my judgment, the Senate did not live up to its full potential in this regard, it did contribute to the health of the regime by complementing the formality of government responsibility to the House with a more meaningful degree of accountability to the Senate.
Today I wish to argue that, for the same reasons, the Senate with a non-government majority has the same institutional control, constitutional power, and political incentive to compel the government to accept legislative compromises that it would be unlikely to make of its own volition. If the compromise approach I have described is preferable to the alternatives, and if the House of Representatives is very unlikely to be, and be seen to be, the forum for conciliation of which Mill spoke, then it is again to the Senate that we must turn. That is largely why I chose to look at the disposition of the Senate’s legislative amendments since John Howard became Prime Minister.
It would be unfair for me to speak at length here about findings that you do not have before you. So I will say only three things.
First, there is clear evidence of what we might call “compelled compromise.” The government obviously has had to find additional votes in the Senate to pass its legislation, and certainly this has forced it to make some unwelcome compromises. However, I have not yet discerned a practical way to distinguish amendments made for this purpose from amendments that the government has proposed or accepted willingly in order to address weaknesses it had come to recognize in its own bills. Still, more often than not, government bills have survived Senate legislative consideration unscathed; and also more often than not, the Senate has not insisted on its amendments when the House has disagreed to them. My tentative conclusion is that this is a story of a glass that is half-empty, not half-full.
Second, I did not find much procedural evidence of legislative compromise after government bills have left the Senate and been returned to the House for further action. When the House has disagreed to a Senate amendment, the upshot usually has been either that the Senate has chosen not to insist on its amendment or the House eventually has chosen to accept it. There have been relatively few instances of the third possibility—the two houses agreeing on a presumed compromise in the form of an alternative to the Senate amendment to which both the Senate and the House of Representatives then agreed. It is unwise to ask a small body of data to support too heavy a load of inference. Even so, what I have found does suggest that what often is called the process of reconciling legislative differences between the two bodies has been less a process of reconciliation (that is, compromise) and more a process of allocating victories and losses between the Senate and the government acting through the House.
And third, July 1, 2005, did mark a turning point. After the government took control of the Senate, almost no non-government amendments were approved by the Senate, so there was no need for the government, acting through its House majority, to accept policy compromises. Virtually the only legislative amendments that Parliament has made since that date have been amendments proposed by the government itself. So has the change in party control in the Senate made a difference? I think so. Has that change contributed to the long-term health of the political system? I fear not.
[1] “Senate Majorities and Legislative Outcomes: The Fate of Senate Amendments,” October 2007. I thank Wayne Hooper, Anne Lynch, and John Uhr for their helpful comments on an earlier version of this text.
[2] An unfortunate consequence of Australia’s electoral system is that the Senators elected in 2004 did not take office until more than six months later.
[3] The potential effects of those elections are not limited, of course, to the Senate’s influence on legislative decisions. However important the Senate’s legislative responsibilities are, it is equally important for the Senate to hold the government accountable for how it has, or has not, implemented existing laws. In this respect Harry Evans has reviewed the first year of government control of the Senate and found that “[t]he government majority in the Senate has greatly increased the ability of the government to do what it likes and not to explain itself except to the extent it chooses.” (Harry Evans, “Having the Numbers Means Not Having to Explain: The Effect of the Government Majority in the Senate”) Specifically, he notes such developments as reduced time for Senate committee review of government bills, regular defeat of motions ordering the government to produce documents, and, perhaps most depressing of all, fewer sitting days for a body that already was not exhausting itself with the number of days on which it worked. Since then, the Senate also effectively abolished the references committees that had been chaired by non-government Senators.
It should be noted that, in recent years at least, the government was not always responsive to Senate attempts to obtain information about its decisions and activities, even when non-government Senators were in the majority. To this outside observer, in fact, the government’s attitudes toward these efforts sometimes seemed to be dismissive or even contemptuous. But to the extent that these attitudes did prevail then, and may prevail now, it must be in part because the Senate has been reluctant to assert vigorously its full range of constitutional powers and to insist on due respect for its constitutional responsibilities and prerogatives. If the government has “dissed” the Senate, as young people in America now are prone to say, the Senate itself must accept some of the responsibility by allowing it to happen.
[4] Elaine Thompson, “Australian Parliamentary Democracy After a Century: What Gains, What Losses?,” available at the website of the Parliamentary Library at www.aph.gov.au/library/pubs/rp/1999-2000/2000rp23.htm.
[5] A reference to a third rail that once ran alongside the tracks of some mass transit rail systems powered by electricity. The third rail was needed because it carried the power, but touching it usually was fatal.
[6] In fact, the dictionary on which I’ve happily relied for more than four decades offers “unanimity” as its chosen synonym for “consensus.”
[7] One reason the practice is not always followed in the last of these is that referring to the “Gentleman from California,” for instance, is not very helpful when there are quite a few men representing California districts. I suppose that is why Australian Senators are identified by name; with 12 Senators from each state, referring to the “Senator from Tasmania” would not be very discriminating.
[8] The larger the constituency, the more diverse it is likely to be; and the more diverse constituencies are, the more they will tend to be like each other in their needs and interests. Sometimes what is most important, however, is not the nature of the constituency as a whole, but the nature of the winning MP’s electoral constituency—the majority that has elected him or her to office. In the U.S. House of Representatives, it is not unknown for a very progressive member to be succeeded by a very conservative one, or vice versa, because the district is fairly evenly divided politically and election outcomes are decided by the swing voters in the middle who are most likely to vote differently from one election to the next. The conservative MP will perceive the needs and interests of the constituency quite differently from his or her progressive predecessor because each has a different mental picture of it in mind, no matter how an outside observer might view it.
[9] Although the size of the required majority can be an issue. As I observed earlier, constitutions and parliamentary standing orders may require absolute or larger majorities, such as two-thirds votes, for certain purposes, and it can be said today that the U.S. Senate effectively requires a three-fifths vote to bring any contentious proposition to a vote.
[10] Even if the government party could and did take all needs and interests into account as it formulated its election manifesto, that would not suffice, because it is important for all segments of the population to see themselves as having some effect on government policy-making.
[11] Quoted by David Hamer in his Can Responsible Government Survive in Australia? 2nd Edition. Canberra: Department of the Senate, 2004; p. 347.
[12] During times of crisis, the U.S. Congress sometimes has set aside its capacity for independent judgment, but not for too long. Examples are the Great Depression of the 1930s, when Congress rubber-stamped the first legislative initiatives of the Roosevelt Administration, but then had second thoughts about the risk of excessive government intervention into the economy, and the national security crisis following September 11th, when Congress approved executive powers that, upon reconsideration several years later, have raised serious questions about excessive government intrusions into personal privacy and civil liberties.
[13] When Senate Republicans were in the majority in Washington before the 2006 election, they became frustrated by Democratic filibusters that prevented votes to confirm some of President Bush’s judicial nominees. (A filibuster is an extended debate in the Senate that cannot be ended by a simple majority vote.) Republicans discussed exercising what came to be called the “nuclear option” that would effectively have changed Senate rules to undermine or prevent at least some filibusters. In response, some Senate Democrats spoke of the right to filibuster on the Senate floor as if it were an integral part of the constitutional scheme of checks and balances. It was implied that Congress can check and balance presidential power only when the party opposing the president has the ability to block the president’s legislative initiatives (and nominations) in at least one house of Congress. Such an argument has no basis in constitutional theory or constitutional law, but I admit that it did, and does, have some practical political resonance.
