CONSTITUTIONS
The foundational act of political engineers is the creation of a constitution to govern a country. Or at least this has been increasingly the case since 1789, to the point that it is an almost universal process.1 Indeed, for democratic polities in particular, constitutions form a container in which we can find the basics of most (if not all) of the other institutional variables discussed in this book, such as the design of legislative, executive, and judicial power or how policy-making authority is divided between central and local authorities.
This chapter will address the question of the basic options open to political engineers regarding constitutions, define the basic concepts associated with constitutions, and conclude with an explicit comparison of the US constitution to those in our other thirty democracies.
THE INSTITUTIONAL OPTIONS
When we consider the options open to the founders of the American republic, it should be noted that from a historical perspective, a written constitution was not a foregone conclusion. While it is retrospectively obvious that the Philadelphia conventioneers should come together to write a constitution, historically speaking such an approach was almost novel at the time. What we think of as the norm in the early twenty-first century was a fairly new enterprise at the time. Indeed, the models available to the framers were limited. As Elkins, Ginsburg, and Melton note, “at least since Aristotle, the word ‘constitution,’ in its various translations, has been invoked to refer to the higher law of political jurisdictions. . . . Nevertheless, no universal model existed until the rise of the modern state at the turn of the 18th century” (2009: 41). As such, the number of models that could be consulted was limited, with the best pool of such examples being, in fact, the states themselves. It is therefore not surprising that The Federalist Papers make numerous references to those documents.
At least in theory, the United States could have followed the examples of other existing countries and could have functioned sans a formal, written constitution, as was the case in Great Britain. Under that option, a collection of documents and practices could have formed the US Constitution. Of course, there were several reasons this was not the case. First, as a relatively young entity, there were no distinct American traditions and practices to be followed in the way analogous to the British experience. Instead, the procedures needed to be spelled out. Further, the pathways toward a written constitution had already been set by the constituent states of the union (as early as 1776) as well as in the guise of the Articles of Confederation. As a result, the United States was a forerunner in a trend that would move from novelty to norm within fifty years.2 Perhaps most important, the complex political relationships that exist in a federal and presidential system required a specific set of written parameters. Since such a system delegates specific powers to specific actors who control veto gates within the policy-making and policy-implementing apparatus, it is necessary to spell out the basic relationships. Imagine, for example, trying to determine the specific powers of the president and Congress without a written document.
A key need for a constitution for the United States can be found in the fact that the type of government that existed post-independence was one that was vested in the basic notion of popular sovereignty. This means that it was a government that derived its authority to govern from the people, not from a king or aristocratic class that could assert power over a territory, or from a religious order that could claim a divine right to rule. Rather, when sovereign power comes from the populace, it becomes quite vital to define what that actually means in a practical sense. Along those lines we can consult Federalist 51, wherein Madison wrote, “In framing a government which is to be administered by men over men, the great difficulty lies in this: you must first enable the government to control the governed; and in the next place oblige it to control itself. A dependence on the people is, no doubt, the primary control on the government; but experience has taught mankind the necessity of auxiliary precautions.” While not in the context of an argument for constitution writing, per se, the passage reflects the notion that for a government to properly function, it cannot rely solely on popular sovereignty, but that proper design of that government is needed, and by extension we can argue that having the basics of that design codified into a constitutional order is a desirable process.
Of course, it should be noted that having a constitution does not make a state democratic, nor does it mean that the government ordained by that constitution derives its power from the people. Indeed, in the constitutional era is it quite common for nondemocratic states to have constitutions. Many of the countries under study in this book have experienced authoritarian governments, including ones that installed their own constitutions. As such, some of the constitutional orders identified in table 3.2 count nondemocratic constitutions alongside democratic ones.
DEFINING THE CONCEPT
Constitutions are documents (with the notable exceptions in New Zealand and the United Kingdom, which famously have unwritten constitutions)3 that establish a basic framework for governance. Constitutions are, by definition, the highest law in a polity and therefore the source of all other laws. The creation of an efficacious constitutional order is understood to be part of the basic establishment of the rule of law in lieu of the arbitrary rule of aristocrats or some other source of authority answerable only to itself or some narrow segment of the society. In the democratic tradition constitutions translate the abstract notion of popular sovereignty to paper and serve as a foundation upon which to build government. This concept of constitutions has become directly linked to the development of modern democracy since the late seventeenth century (see, for example, Casper 1989, Dippel 1996, and especially Sartori 1962). Larry Diamond expressly links the notion of a fully democratic state to a constitutional order: “A constitutional state is a state of justice, a Rechtsstaat in the German, in which the state acts predictably, in accordance with laws, and the courts enforce restrictions on popularly elected government when they violate the laws, or the constitutional rule” (1999: 12). In other words, the notion of constitutional government is one of government limited to an agreed-upon set of structures and responsibilities as recorded in a fundamental document. The state then flows from that document (at least in terms of an ideal interpretation of the process and recognizing that the practical applications of parchment principals are often messier than their authors might like).
Constitution writers have to make fundamental choices about how to structure the relationship between citizens and their government, as well as to determine how the internal structure of the government will function. It is within constitutions that we expect to find the basic outlines of the rights and privileges of citizens and how the rights of citizenship can be used to influence the selection and behavior of politicians in government. As such, constitutions outline the basic parameters of the principal-agent relationship outlined in chapter 1 (and illustrated in figure 1.1). Further, constitutions detail the basic design of government and how the agents of the people will interact with one another as they attempt to make and implement public policy. A clear illustration of this idea can be found in a fundamental constitutional choice: between a parliamentary system, which fuses the legislative and executive powers, and a presidential system, which creates separation of powers. In a parliamentary system the legislature chooses the executive (prime minister and cabinet), and the relationship that is created is a hierarchical one in which the cabinet is ultimately answerable to the legislature, and thus the policy preferences of the cabinet and the legislature are typically aligned.4 However, in a presidential system, wherein the legislature and executive are elected separately through differing methods and by different constituencies, the relationship becomes transactional. This means that the president must somehow secure the consent of the legislature, which may have different preferences, to achieve policy (more on this in chapter 8, “Executive Power”).
Indeed, constitutions typically establish the basic parameters for practically all of the items under discussion in this text, that is, the division of power between the central government and localities, the locus and basic functioning of legislative, executive, and judicial power, and the parameters (if not the precise rules) for the electoral system. The functioning of all these institutions then directly shapes the development and behavior of political parties and interest groups and, ultimately, the production and implementation of public policy.
While constitutions are typically single, unified documents, this is not always the case. Some constitutions are considered “unwritten” because they lack such a single, unified document. The quintessential example of this is the British constitution, for while it does contain written materials, it is a collection of multiple sources and traditions rather than a singular document. The earliest constitutional document of the United Kingdom, the Magna Carta, dates from the year 1215. However, the Magna Carta is only a small part of the British constitution, and the British constitution is a so-called “unwritten” constitution, consisting of “an agglomeration of statutes, judicial interpretations, conventions, laws and customs of Parliament, common law principles, and selective jurisprudence, such that even scholars may disagree on what is and what is not a part of the constitution” (Maddex 2008: 470). The British constitution is therefore not included in many of the tables included in this chapter. Likewise, New Zealand’s constitution consists of a number of documents, dating from the Treaty of Waitangi in 1840, but lacks a unified document. Israel’s constitution is also often called “unwritten,” because instead of a single constitutional document, it has a series of separate basic laws. Israel’s 1948 declaration of independence has been called the “earliest quasi-constitutional document” (Maddex 2008: 226), and the oldest basic law is the 1958 basic law on the Israeli parliament.5
In addition to being the fount of many topics for discussion, constitution writing and reform is a fundamental example of the concept of political engineering as noted in chapter 1 (for a direct discussion of such, see Lijphart 2004). Indeed, one cannot find a clearer example of the principles of practical political experimentation and political engineering than the process of constitution building in the early United States (as we discussed in the previous chapter and outlined in its appendix). At the onset of the revolutionary period, the thirteen colonies that became the thirteen states originally organized themselves under the Articles of Confederation. However, the failure of that constitution to create an adequate governing structure for the United States led to the Philadelphia Convention of 1787, one of the most famous examples of political engineering of all time. Indeed, The Federalist Papers contain, primarily, arguments against one constitution (the Articles) and in favor of the one that would be ratified in 1789.
Constitutions tend to operate on a general level, providing a basic blueprint for governance without delving into the specifics of day-to-day public policy, although some do make explicit policy statements. The precise working out of this blueprint as well as the daily function of government is left to the legislative and executive powers of the state.
The comparative study of constitutions is complicated by the fact that over time constitutions come and go for a variety of reasons. Several factors ought to be considered. There are straightforward issues, such as the date of establishment of the current constitution governing a given country and therefore the age of a given constitution. Beyond that there are the questions of when the practice of using a written constitution was established in a given country (a practice, as noted earlier, that only started to become a global norm in the nineteenth century) and how many constitutions a given country has had.
Even the process of counting constitutions can be tricky, insofar as a given constitution might be used at different times in a given country’s history. For example, a constitution might be abrogated by a military coup or war and be replaced by one or more constitutions before being restored at some later date. Let’s consider a hypothetical case in which a country establishes a constitution in 1850, has a coup in 1860, which results in a new constitution, which in turn is replaced as the result of the country being occupied by a foreign power in 1870, only to finally restore the 1850 constitution after a successful war of liberation in 1880. How many constitutions did that country have? In terms of documents, the country had three (1850, 1860, 1870), but in terms of constitutional systems, four (1850, 1860, 1870, and 1850 again). Along these lines we adopt the counting process employed by Elkins, Ginsburg, and Melton: “A constitutional system encompasses the period in which a constitution is in force before it is replaced or suspended” (2010: 2).
In general the number of constitutional systems that a given country has experienced can be a measure of its long-term political stability as well as perhaps its vulnerability to outside political influence. Of course, not all constitutional orders are replaced as a result of violence or systemic failure but rather may be the result of political reform of a democratic nature. As we seek to understand political development in our democracies, it is useful to note when they adopted their first constitutional system as well as how many they have had. Table 3.1 contains this information for twenty-nine of our democracies (excluding New Zealand and the United Kingdom).
Several factors are worth noting. First, as described earlier, some constitutional orders are nondemocratic, imposed by authoritarian governments or occupying powers (such as France’s 1940 constitution). Other times the result of loss in war can lead to the adoption of a democratic constitution (such as in Germany in 1949 and Japan in 1946), as can the exit (from various means) of an authoritarian regime (such as Brazil’s new constitution in 1988 or Argentina’s restoration of its constitution of 1853 in 1983).
Table 3.1. Historical constitutionalism in twenty-nine democracies (1789−2006)
Source: Based on data from Elkins, Ginsburg, and Melton 2010.
Second, many replacements come about via a process of democratic reform, replacing one democratic constitution with another entirely without abrogating the democratic nature of the regime (for example, Colombia’s replacement of its 1886 document with one in 1991, or France’s move from the Fourth Republic constitution to that of the Fifth Republic in 1958).
Other new replacements are even less dramatic: Entirely new constitutions may be written and formally adopted that are substantively not very different from the preceding constitution. For instance, the 1999 Swiss constitution was largely an aesthetic exercise for the purpose of deleting outdated provisions and integrating accumulated amendments into more streamlined constitutional texts. A similar—hypothetical—rewriting of the US Constitution would entail such changes as (1) moving the two-term limit for presidents, specified in the Twenty-Second Amendment, to Article II dealing with the executive power; (2) deleting the no-longer-relevant provision in the Twenty-Second Amendment that the two-term limit does not apply to the president in office at the time the amendment was proposed; (3) removing the sections in Article I concerning how slaves and Indians would be counted (in Article I, Section 2); and (4) deleting both the Eighteenth Amendment and the Twenty-First Amendment (except Section 2) that respectively impose and repeal prohibition.
Third, many states have experienced a large number of constitutions in a short period of time, marking initial instability and then moving into a period of sustained constitutionalism (such as Colombia’s eight constitutions from 1821 to 1886 and then entering into a period of 106 years without a replacement).
THE UNITED STATES CONSTITUTION IN COMPARATIVE PERSPECTIVE
All constitutions provide the basic rules for the organization and operation of governments. This is also the basic purpose of the US Constitution. In addition, however, the US Constitution serves as a source of national pride—arguably more so than the constitution in any other democracy. Most knowledgeable Americans know that the US Constitution is the oldest constitution in the world, and they are proud of this special characteristic of their Constitution. They also know that it is a remarkably brief document, and they generally admire this feature. Because of their reverence for their Constitution, they tend to be very reluctant to change it; hence they also approve of the fact that the Constitution is very difficult to amend, and they generally know that in practice it has only rarely been amended. Finally, they also see the Constitution as the foundation for the establishment of democracy in the United States at a relatively early date, and they often call the United States “the world’s oldest democracy.” All of these judgments imply comparisons with other countries. This section of the chapter is devoted to an explicit and systematic comparison with thirty major democracies on these subjects.
1. An old and durable constitution. The proud claim that the United States is governed by the world’s oldest written constitution is clearly justified. It was adopted in 1787 and became effective in 1789. In fact, the advent of the US Constitution was such that the Comparative Constitutions Project uses 1789 as the starting point for its study of worldwide constitution making.6 If we look at the current constitutions of our other democracies (as listed in table 3.2), we can see that the US Constitution is the only one, from the twenty-nine democracies with written constitutions, that dates back to the eighteenth century and only one of four from democracies that were not established (or reestablished) in the twentieth century.
If we look back to table 3.1, we can further see that the United States has had a very long experience with constitutionalism, but one that shares eighteenth-century origins with France, the Netherlands, and Switzerland. It also has had a remarkably durable constitutional system. Of the pre–twentieth-century constitutional systems, only two others share the distinction of still utilizing that same constitution (Belgium and Canada).
Table 3.2. Date of origin of current constitutions in twenty-nine democracies
Source: Based on data from Elkins, Ginsburg, and Melton 2010.
2. A short constitution. Determining the length of a given constitution is not as easy a task as it may appear at first glance. One problem, but only a minor one, is that in order to obtain comparable counts, English-language versions of all constitutions have to be used, and translations by different translators can obviously contain varying numbers of words. A more serious problem is that some constitutions affirm the validity of earlier constitutional texts without repeating them in the constitution itself. For instance, the 1958 constitution of the French Fifth Republic begins by solemnly proclaiming the attachment of the French people to the 1789 Declaration of the Rights of Man and of the Citizen, a separate document of more than eight hundred words. Another example can be found in the 1949 German constitution, which lists five articles of the previous constitution of 1919 and states that these articles “are an integral part” of the new constitution. The second problem is whether provisions that are clearly no longer relevant (such as in the example of the Eighteenth and Twenty-First Amendments to the US Constitution mentioned earlier) should be included in the count. Our decision was to be as inclusive as possible, but not to claim a degree of accuracy that is not warranted. Table 3.3 therefore shows the length of the constitutions of twenty-nine democracies (all of our democracies except the United Kingdom and New Zealand) in thousands of words.
The United States is near the top of the table, but not at the very top: both Japan and Denmark have shorter constitutions. Overall, the differences in the length of constitutions are not as great as the differences in their ages. Three constitutions in the middle of the table (those of Switzerland, Israel, and Spain) have constitutions that are roughly twice as long as the US Constitution, and only eleven constitutions are more than three times as long. The Indian constitution is an anomaly among democratic constitutions, with more than one hundred thousand words. The longer and more detailed constitutions are, the more often they need to be amended—a subject to which we shall return at the end of this chapter.
Table 3.3. Length of constitutions (in thousands of words) of twenty-nine democracies
Source: Based on data available at ConstitutionMaking.org 2010; Political Database of the Americas at http://pdba.georgetown.edu/Constitutions/constudies.html; and Tschentscher 2010.
3. The world’s oldest democracy? In his 1993 inaugural address, President Bill Clinton called the United States “the world’s oldest democracy.” He was not the first American, and will not be the last, to make this claim. How valid is it?
The best-known analysis of the paths that the countries of the world have taken toward (and away from) democracy is that of Samuel P. Huntington (1991). Huntington identifies a long first wave starting as early as 1828 and lasting until 1926, a short second wave from 1943 to 1962, and a third wave starting in 1974. Two reverse waves, in which democracy collapsed in many countries, occurred between the three waves of democratization.
It is worth noting that as time has progressed, the definition of democracy has deepened. Huntington’s choice of 1828 as the beginning of the first wave, for example, illustrates this fact. Specifically he uses the following criteria: (1) 50 percent of males eligible to vote, and (2) a chief executive responsible either to parliamentary majorities or directly to the electorate (1991: 16). From a contemporary point of view, these are extremely inadequate measures of democracy. From a historical perspective, however, such a reliance on popular will was innovative, as was a chief executive who obtained power via some connection to that popular will (as opposed to, for example, being a monarch or military leader).
The United States is listed first among the democracies in table 3.4, because Huntington explicitly identifies the United States as the country where the first wave of democratization began in 1828. The other seven countries in the first group are listed alphabetically, as are the countries in the other groups with the same pattern of waves and reverse waves in the table. Of our thirty-one democracies, twenty-five can be called “old” democracies, in the sense that they all participated in the first wave of democratization, but only eight, including the United States, can claim to have been continuous democracies since the first wave. It is also worth noting how many of our democracies have experienced reversals: twenty of the thirty-one current democracies. Four countries—Argentina, the Czech Republic, Greece, and Hungary—have the checkered democratic history of having participated in all of the waves and reverse waves.
Table 3.4. Waves of democratization and reverse waves in thirty-one democracies
Source: Adapted and expanded from Huntington 1991: 14–17.
Some of Huntington’s judgments may be questioned. In particular, he argues that the German occupation of France, the Netherlands, Belgium, and Denmark during the Second World War constituted breaks with democracy in the first of his reverse waves, even though these breaks were externally imposed and lasted a relatively short time, after which democracy was immediately restored. Similarly, he includes India in his second reverse wave because of the so-called Emergency in India from 1975 to 1977, although this clearly undemocratic interlude was quickly ended by free elections after only about a year and a half.
A more serious objection to Huntington’s analysis, and to his conclusion that the United States is the oldest of the world’s democracies, is that he uses an extremely lenient definition of “universal” suffrage: the right to vote for at least 50 percent of adult males.7 As a result, he ignores the exclusion of women and members of racial and ethnic minorities from the franchise in many countries. Truly universal suffrage, including the right of women and minorities to vote, is not a sufficient condition for democracy—for instance, the fact that all citizens of Iraq under Saddam Hussein had the right to vote obviously did not make Iraq a democracy—but is clearly a necessary condition.8
Table 3.5 lists our thirty-one democracies according to the year in which universal-suffrage democracy was established or reestablished. In just over half (sixteen) of our countries, especially the more recent democracies at the bottom of the table, such full democracy was instituted after a period of completely nondemocratic rule. An additional two cases established universal suffrage at the time of national independence. However, thirteen countries, generally toward the top of the table, already had extensive but not fully inclusive voting rights and adopted universal suffrage in the years indicated. In most cases, the final extension of the suffrage entailed the admission of women to the suffrage—as early as 1893 in New Zealand, but as late as 1948 in Belgium, 1954 in Colombia, and even 1971 in Switzerland. In three countries, the final step in making the franchise universal involved the grant of secure voting rights to ethnic and racial minorities: Indians (or First Nations, as they are now known) in Canada in 1960, Aborigines in Australia in 1962, and the African American minority in the United States in 1965, when the Voting Rights Act was passed.9 In terms of the establishment of universal-suffrage democracy—which is really a redundant term because, as stated earlier, universal suffrage is a sine qua non of democracy—the United States is not at the top, but only roughly in the middle of table 3.5. It is therefore more accurate and more modest to claim that the United States is the oldest country with a broadly representative government—which, however, did not become fully democratic until the 1960s.
4. How democratic is the US Constitution? Democracy requires adherence to the principle of “one person, one vote” but also to the principle of “one person, one vote, one value”—that is, all votes should have equal weight. In his book with the provocative title How Democratic Is the American Constitution? Robert A. Dahl (2001) identifies several antidemocratic elements in it and, in particular, the serious inequality in the value of citizens’ votes that results from the equal representation of the fifty states in the US Senate. He points out that the vote of a resident of Nevada is worth about seventeen times that of a resident of neighboring California, and that the vote of someone in Connecticut is worth almost six times the vote of his or her neighbor in New York. “In the extreme case, the ratio of over-representation of the least populated state, Wyoming, to the most populous state, California, is just under 70 to 1.” Dahl (2001: 49–50) argues that such inequalities in representation constitute “a profound violation of the democratic idea of political equality among all citizens.”
Dahl recognizes that, in all federal systems, there is an inherent tension between the idea of equality of citizens and the special representation of states, provinces, and cantons that are the constituent units of the federation. This special representation frequently means the over-representation of the smaller units and hence the underrepresentation of the residents of the larger units.10 Dahl’s criticism is therefore not so much that a degree of citizen inequality exists, but that this inequality is unusual and extreme in the United States. In most of our other federal democracies, the inequality is indeed more modest. We shall return to this topic in greater detail in our chapters on federalism (4) and legislative power (7).
Table 3.5. Establishment or reestablishment of universal suffrage in thirty-one democracies
Sources: Based on data in Karatnycky, Piano, and Puddington 2003; Inter-Parliamentary Union 1995; and Mackie and Rose 1991.
5. A constitution that allowed, then prohibited, slavery. The Constitution of the United States is only one of eight that mentions slavery, and the only one that contained constitutional provisions that supported the “peculiar institution” that it then had to amend out of the document.11
In comparative terms, the eight constitutions that mention slavery in some capacity are those from Argentina, Brazil, Chile, Colombia, Japan, Mexico, South Africa, and the United States of America. Of these eight, four (Chile, Colombia, Mexico, and South Africa) contain a prohibition on slavery as an original constitutional right, with two (Chile and Mexico) also providing immediate emancipation for any foreign slaves who enter their national territories. Two constitutions (Brazil and Japan) make reference to slavery without any particular rights being defined. Japan’s preamble simply asserts a “desire to occupy an honored place in an international society” and to “striv[e] for the preservation of peace, and the banishment of tyranny and slavery, oppression and intolerance for all time from the earth,” while the Brazilian constitution speaks of preserving “documents and sites” of historical significance to runaway slave communities. The Argentina constitution provides both a prohibition of slavery and an emancipation proclamation: “The few [slaves] who still exist shall become free as from the swearing of this Constitution.”
The US Constitution, on the other hand, has three references to legal slavery from its original 1789 incarnation (and the wording remains in the document). It is worth noting that the word slavery was not mentioned until it was abolished by the Thirteenth Amendment (the fourth reference to the institution of slavery in the Constitution), but the inclusion of the concept was quite clear. The first reference is in Article I, Section 2, Clause 3, which contains the infamous “Three-Fifths” compromise in which persons “bound to Service for a Term of Years” (that is, slaves) would count as three-fifths of a person in terms of taxes (which were then based on head count) and representation in the House of Representatives. The second reference is in Article I, Section 9, Clause 1, which prohibited the regulation of the “Migration or Importation of such Persons as any of the States now existing shall think proper to admit” (that is, slaves) until 1808. The third reference is to runaway slaves, which can be found in Article IV, Section 2, Clause 2, which states (in its entirety): “No Person held to Service or Labour in one State, under the Laws thereof, escaping into another, shall, in Consequence of any Law or Regulation therein, be discharged from such Service or Labour, but shall be delivered up on Claim of the Party to whom such Service or Labour may be due.”
As noted, all of these clauses were made moot by the Thirteenth Amendment, which prohibited slavery and involuntary servitude.
All of this, however, does mark the US Constitution as having a unique place on this issue, which illustrates that political impasse over the slave issue at the framing and also underscores the degree to which historical sequencing (that is, when a constitution is written) can substantially influence the contents of that document (and the pathways its institutions are then set upon). Further, this aspect of the origins of the United States highlights why issues of race have been of great salience to the development of US politics, even to the present day.
Of course, it is not surprising that the bulk of the constitutions that mentioned slavery in one form or another (six out of eight) would be from the Americas, where the issue of slavery was a key component of both colonization and development. Along those lines, it is noteworthy that other cases in our study did utilize slavery, especially in colonial holdings. For example, the Dutch did not ban slavery in its colonies until 1863, and while slavery was outlawed in France during their revolution, it was restored in the colonies by Napoleon. Also noteworthy: The end of slavery in Brazil in 1888 marked the latest emancipation of slaves for countries in our study.
6. A constitution that is difficult to amend. All democratic constitutions contain provisions for their own amendment, and almost all of them make it more difficult to change the constitution than to enact or change ordinary laws. But they differ a great deal with regard to how flexible or rigid they are in this respect. Procedures for amending a constitution vary in their numeric thresholds for approval, as well as the institutional complexity of the process. Thus we have two dimensions, where the numeric dimension refers to “how many must agree,” whereas the dimension of complexity refers to “who must agree.”
On the numeric dimension, we find that constitutional amendment processes often require some sort of supermajority of the legislature, although not all do. The range of thresholds runs from an absolute majority (half plus one)12 up to three-quarters. Many constitutions also require a certain number of legislators to be present (a quorum). Numeric thresholds that make approval of amendments more difficult can also be applied to referenda, which are votes by the public.
The second dimension, that of complexity, reflects the number of actors/veto gates in the process. In pure legislative supremacy in a unicameral setting, complexity is low, as the number of veto gates is one: the legislature itself. However, most amendment processes go beyond just the legislature. Some require multiple votes, forcing legislators to take responsibility twice for voting for change, such as in one of the procedures under the Colombian constitution. Others required two votes and an intervening election. Instead of one actor, a given legislative body, such a process includes three: the originating legislature, the electorate, and then a newly elected legislature chosen by those voters, who will likely have had their vote influenced by their views on constitutional reform. Other systems require a referendum, which directly inserts the voters into the decision-making process, by having them vote for or against the proposed changes. Further, federal systems often include subunit legislatures or have referendum provisions that take account of the states or other units. For example, Switzerland requires a majority of voters nationwide and also majorities in a majority of subunits (cantons). The complexity of such procedures is such that the combination of subunits and a supermajority requirement can empower minorities to band together to block changes to the constitution. We tend to find that federal states have more complex processes for amending their constitutions, which are therefore typically more rigid than their unitary counterparts.
Figure 3.1 plots the constitutional amendment procedures (in some cases more than one, as is elaborated upon in the appendix) for our thirty-one democracies on these two dimensions. Some countries (such as India) have different processes depending on what part of the constitution is being amended. As the chart moves from the top left cell to the bottom right, the constitutions as plotted in the chart move from flexible to rigid.
The dotted diagonal line in the chart is the line of constitutional rigidity, moving from the least rigid (legislative supremacy with a 50 percent plus one process) to the most rigid (to a two-level approval process involving a referendum and a 75 percent threshold for passage). Of course, the most rigid arrangement within the confines of this model is a theoretical one, as none of our democracies have those provisions in place. Only one of our democracies fits in the most flexible category (Israel). In the past, New Zealand and the United Kingdom could also have been included in the least rigid category, but increasingly in both cases the referendum possibility has become the consensus position.
Among our thirty-one democracies, the United States is the only example of a constitution with a three-fourths supermajority requirement for all amendments. Indeed, it is only one of two cases that uses a three-fourths supermajority at all, as South Africa’s constitution employs it for amending some of the foundational aspects of its constitution. The much more common supermajority, found in most (eighteen out of thirty) of our democracies, is the two-thirds majority—based on the idea that supporters of a constitutional change have to outnumber their opponents by a ratio of at least 2 to 1.
Fig. 3.1. Constitutional amendment strategies and supermajority requirements in thirty-one democracies
Key: [bi] Indicates a bicameral legislature, meaning two votes are needed within the process. Italicized states are federal.
*Under France2 the parliament votes in a joint session. So while the legislature is technically bicameral, this process requires only one vote (Article 89).
**In Argentina, the legislature indicates the need for amendment, but must call an elected assembly to deal with the actual amending.
CONCLUSIONS
Constitutions are foundational documents wherein a polity can establish its basic institutional framework. Therefore, constitutions are quintessential examples of political engineering, which set the stage for the subsequent process of policy creation within a given state. Indeed, the rest of this book discusses various aspects of democratic governance that directly or indirectly grow out of the constitutions of our thirty-one cases.
Like the other institutional variables that we are analyzing herein, the US Constitution stands out in comparison to its counterparts in our other cases: It is the oldest and longest-lived of such documents. In these ways it is unique among our cases. It is also one of the shortest constitutions under study. In terms of its democratic quality, it has legitimate claim to being one of the oldest democratic constitutions, although when it comes to the full establishment of universal suffrage, its performance is less singular.
APPENDIX: CONSTITUTIONS WITH MULTIPLE AMENDMENT PROCESSES
Here is a listing of the specific processes noted in figure 3.1:
Austria1: two-thirds of both chambers
Austria2: two-thirds plus referendum for total revision
Chile1: three-fifths of both chambers
Chile2: two-thirds if amendment to Chapters I, III, VIII, XI, XII, or
XV (Referendum Threat by president in both cases)
Colombia1: Passed by 50 percent +1 by two consecutive regular sessions (no intervening election)
Colombia2: Passed by 50 percent +1 of legislature and then submitted to referendum (25 percent of electorate must participate for measure to pass)
France1: 50 percent +1 of both chambers followed by a referendum
France2: three-fifths of parliament convened in congress
India1: two-thirds of legislature
India2: Two-level requirement if changes affect:
a. Article 54, Article 55, Article 73, Article 162, or Article 241, or
b. Chapter IV of Part V, Chapter V of Part VI, or Chapter I of Part XI, or
c. any of the Lists in the Seventh Schedule, or
d. the representation of States in parliament, or
e. the provisions of this article (that is, Part XX, 368)
Poland1: two-thirds in the Sejm, 50 percent +1 in the senate
Poland2: Amendments to Chapter I, II, or XII can also be submitted to a referendum
South Africa1: Other amendments require a two-thirds vote in the National Assembly and six of nine (two-thirds) of the National Council of Provinces
South Africa2: Section 1 and Subsection 74 require a three-fourths vote in the National Assembly and six of nine (two-thirds) of the National Council of Provinces
Notes
1. See Elkins, Ginsburg, and Melton 2009, esp. ch. 3. Specifically, the authors point out that while in 1800 almost no countries had formal constitutions in the modern sense of the term (which they meticulously operationalize), over the course of the nineteenth century the gap between countries in existence and countries with constitutions rapidly closed. Indeed, for new states formed after 1789, the adoption of a constitution was seen as an “obvious” action (41).
3. These constitutions are unwritten at least in the sense of a unified document. This is further discussed below, along with the case of Israel.
4. If they are seriously at odds, the provision for a no-confidence vote means that the legislative majority can change the composition of the executive, or an early election can be called.
5. Israel’s basic laws can be viewed, in English, at the Knesset’s Web site: http://www.knesset.gov.il/description/eng/eng_mimshal_yesod1.htm.
6. The Comparative Constitutions Project is aimed at “understanding the origins, characteristics, and consequences of written constitutions for most independent states.” It contains “nearly all national constitutions from 1789 onward” (Elkins, Ginsburg, and Melton 2009: ix). The project can be accessed online at http://www.comparativeconstitutionsproject.org/index.htm.
7. Huntington (1991: 14) concedes that he includes both democratic and “semidemocratic” systems.
8. This is well illustrated by the October 2002 Iraqi election, in which the only candidate on the ballot was Saddam Hussein. See the BBC article, “Saddam ‘wins 100% of vote’” available online at http://news.bbc.co.uk/2/hi/2331951.stm. Examples of authoritarian governments allowing elections are numerous. Certainly we could refer also to the old Soviet Union and various Soviet satellite states, such as East Germany. Sometimes the disconnect between elections and governance is more subtle, such as in Mexico under the Institutional Revolutionary Party (PRI) before the democratizing reforms of the 1990s.
9. The principle of universal suffrage was also violated by the United States, the United Kingdom, France, the Netherlands, and Belgium while these countries were colonial powers, because the inhabitants of their colonies lacked voting rights; by the three Allied Powers while they were occupying Germany and Japan; and by post-1967 Israel, on account of its control over the occupied territories. Postwar control of conquered countries or areas is the least serious violation of the universal-suffrage standard because such control is meant to be temporary; the longer such control lasts, however, the more it creates a dilemma for democracy.
Further, in the contemporary United States, US citizens residing in territorial holdings such as Puerto Rico and Guam lack voting rights as they pertain to the Congress and the presidency. Residents of the District of Columbia have had voting rights for the presidency since the passage of the Twenty-Third Amendment in 1961 but have no voting member of the Congress.
10. However, as we shall see in ch. 4, it does not necessarily mean the completely equal representation of all units, regardless of population.
11. One could also argue that the British constitutional system had to amend itself to abolish slavery, as was done with the Slavery Abolition Act of 1834.
12. As distinct from an ordinary majority, which means more “yes” than “no” votes, which may be less than a majority of the members, due to abstentions.
