CONSTITUTIONAL RIGIDITY: A VETO PLAYERS APPROACH

By

George TsebelisKellogg Fellow

CONSTITUTIONAL RIGIDITY: A VETO PLAYERS APPROACH

This paper uses the theory of veto players to measure constitutional rigidity, which is the difficulty of revising a constitution on the basis of its own rules. The low correlation among indexes characterizing previous approaches, as well as the restricted number of countries they cover, generates the need for a theoretically-based assessment of all democracies. I explain the rules of assessing constitutional rigidity, apply them to a complicated case with many veto players, qualified majorities, and alternative procedures (Italy) so that the reader understands the procedures. Then, I apply the rules to all (92) democratic constitutions. Constitutional rigidity is important not only for constitutional revisions, but also to assess the power of constitutional courts.

CONSTITUTIONAL RIGIDITY: A VETO PLAYERS APPROACH

“…Amending clause…describes and regulates…amending power. This is the most important part of the constitution.”

(John W. Burgess, Political Science and Constitutional Law 1890: 137).

Country constitutions systematically involve (besides Burgess’s “most important” amending clauses) two categories of items: individual rights, and the rules of the political game. These two categories of items require time consistency. That is, individual rights and the rules of the political game must be well known in advance and respected by all participants in the political game. In addition, they need to remain constant, so that all participants know their rights and obligations. For this reason, constitutions protect their text from change by making modification difficult. Indeed, constitutions include provisions requiring qualified majorities of one body—or concurrent majorities of several bodies, or both—in order to be modified. In addition, many constitutions specify alternative procedures for constitutional amendments (for example, provide for less stringent majorities in the legislature if the modifications are supported by a referendum). Finally, many constitutions involve provisions that cannot be modified at all (like human rights or the political regime existing in a country).

The multiplicity of these constitutional amendment provisions is extremely important for the way the political game is played in different countries, since if the restrictions are stringent, then, at any point in time, institutions have to be considered almost “exogenous” (the outcome resulting from decisions made in the past and imposed on the current players); on the other hand, if these restrictions are weak, any time the actors disagree with the constitutional rules, they will include a constitutional revision in their agenda.

For example, in some countries, such as France (art. 77), the rules concerning the electoral system are set by customary law. This is why the electoral system became the subject of political competition in 1985 (from plurality to proportional) and 1986 (back to plurality). In other countries, like Slovenia (art. 80), a majority of two-thirds is required for a change of the electoral system. In Greece (art 54), the electoral system is protected either by the requirement of a two-thirds majority in Parliament, or “shall be applicable as of the elections after the immediately following ones.” Preliminary constitutional debates in Greece indicate that some government officials want to include proportional representation in the constitution. Because such rules are evidently consequential for the political game, students of constitutions have been studying the issue of “constitutional rigidity.”

These studies have been done at the normative and/or theoretical level, starting with the debate between Jefferson (who advocated frequent changes of constitution) and Madison (who prevailed in establishing a long-standing one), as well as the empirical one, trying

to assess the level of constitutional rigidity in different countries. Given the variety of locking mechanisms of constitutions, and given the ability of founders to combine them either as supplements or as substitutes, the range of constitutional rigidity is extremely high, and the conclusions of the empirical literature are quite diverse.In this paper, I will focus on the empirical literature, and review the different methods that have been used to assess rigidity (section I). Then, I will analyze the different mechanisms that ensure rigidity, as well as their combinations (as supplements or substitutes) and the effects they produce on the basis of the veto players theory (Tsebelis 2002; section II). I will apply this methodology specifically in the case of the Italian constitution (section III). Then, I will provide the resulting rules of empirical evaluation of constitutional rigidity based on the theory of veto players (section IV). Finally, I will apply this methodology to 92 countries of the world that are considered “democratic” since they scored above 6 in Polity (Marshall 2016) (section V).

I. Constitutional Rigidity Literature

Following Lutz’s (1994) pioneering work on constitutional amendments, a series of authors tried to provide empirical measurements of constitutional rigidity. The challenges of the enterprise are many, and this is the basic reason that the results differ. Constitutional rigidity may differ from one article of a constitution to the next1 as the constitution itself may provide for different provisions for the modification of different articles (not to mention the prohibition of amending certain articles like human rights, or the regime type); for the same article or sets of articles, alternative methods of revision may provide a choice of alternative political institutions. Finally, the array of revision procedures applicable is incredibly wide, ranging from multiple bodies, to referendums, to time delays sometimes involving intermediate elections, and sometimes even creating special bodies (assemblies).Focusing on institutions, some authors consider only a subset of the issues. For example, Lutz (1994) focuses on the qualified majorities required in the amendment process (Lorenz 2005: 341-2) and Anckar and Karvonen (2002) focus manly on the political actors involved (Lorenz 2005: 344-5). Lutz studied 82 constitutions and found that “the amendment rate is highly correlated with the degree of difficulty (Lutz 1994: 364).” As noted above, he examined two different sets of constitutions. The first set was the US state constitutions (n=50), and then he looked at 32 countries in different time periods in a second round of analysis (Lorenz 2005: 347). But Lorenz (2005) was not able to apply Lutz’s index successfully to “measure the rigidity in countries which are not mentioned by him…(Lorenz 2005: 342).” Anckar and Karvonen’s work establishes a “two-dimensional typology of amendment methods by distinguishing between three kinds of legislative votes (none, ordinary, qualified majority) and the same kinds of popular votes (Lorenz 2005: 344).” On the other extreme, Lijphart (2012) bases his assessment mainly on the threshold required for the approval of the amendments creating a fourfold typology: The classes are 1 On the basis of it, Albert (2014) distinguished constitutions as either “comprehensive” (if the whole constitution can be modified with the same rules), “restricted” (if different provisions are subject to different rules) or “exceptional” (where different rules are used exclusively for one provision or a set of related provisions).

defined as approval by an ordinary majority as the easiest way (1.0); by more than an ordinary majority but less than a two-thirds majority or an ordinary majority plus a referendum (2.0); by two-thirds majorities or equivalent (3.0); and by more than a two-thirds majority or a two-thirds majority combined with other requirements, such as approval by state legislatures (4.0) (Lijphart 2012: 207). He finds that this classification correlates with the strength of judicial review (which he assesses independently, and also classifies in a fourfold group) (Lijphart 2012: 214-5). Schneier (2006) uses a similar method, and classifies 101 constitutions in five categories and nineteen subcategories. Tsebelis (2016) uses each one of these extreme methods: either the number of actors involved or the required majorities. I comment on Schneier’s indicators later.Most of the authors use a combination of the actors involved and thresholds required, while a few (Elster (2000), Lane (2011)) use also non-voting criteria such as time delays. La Porta et al. (2004) examine a group of countries whose constitutions have remained unchanged since 1980 (La Porta et al.: 448). They measure constitutional rigidity on a scale from 1 to 4, which is broken down in La Porta et al.’s Table 1: “One point each is given if the approval of the majority of the legislature, the chief of the state, and a referendum is necessary in order to change the constitution. An additional point is given for each of the following: if a supermajority in the legislature (more than 66 percent of votes) is needed, if both houses of the legislature have to approve, if the legislature has to approve the amendment in two consecutive legislature terms, or if the approval of a majority of the state legislature is required (La Porta et al.: 451).”

Rasch and Congleton (2006) combine data on amendment rates from Lutz (1994) with the institutional information they have on formal amendment rules in countries. They then “create indexes of consensus and of the number of central government veto players or points of agreement required to secure a constitutional amendment (Rasch and Congleton 2006: 546).” The number of governmental veto players is coded as 0-3, with a single point awarded for each center of institutional authority beyond parliament that must agree to a proposed amendment (bicameral, presidential, and federal). They consider the number of veto points to be the “number of governmental veto players plus an additional point if an intervening election is required and another if a referendum is normally used to ratify constitutional amendments (Rasch and Congleton 2006: 546).” They find that “amendment rates fall as the number of veto players increases and with requirements for intervening elections and/or referenda [and also that] the requirement of supermajorities in the legislature and the age of the constitution have no systematic effect within the present sample…the regression results are consistent with the existence of a significant cost-effect on the demand for constitutional amendments. As the cost of passing an amendment increases, fewer amendments are adopted (Rasch and Congleton 2006: 546).” Furthermore, they find that the “frequency of lawful constitutional changes, unfortunately, cannot be understood by focusing on the number of veto players and degree of required consensus alone (Rasch and Congleton 2006: 548).” Rasch and Congleton argue that “the political demand for constitutional reform reflects economic, political, and cultural circumstances, as well as the magnitude of unresolved problems at any given point in time (Rasch and Congleton 2006: 548).” Lorenz (2005) focuses on a mix of institutional and contextual variables and combined elements of Lutz’s, Lijphart’s, and Anckar and Karvonen’s work to identify “the type of

majority rule with the number of voting arenas or actors (Lorenz 2005: 346).” She uses the Lijphart scale (1 to 4) to score the majority rules in each voting arena, but she modifies it by treating the three-fifths majority like two-thirds majorities (3.0 points), and then assigns four points to majorities that are greater than three-fifths. She adds one point for an approval by a constitutional court, a government, or the President. She also includes pre-final votes in her scoring rule: “if a required vote does not relate to the final document but only to the declaration of a need to amend the constitution, a first, still general draft or similar, then this vote is scored with half of the normal points for the respective action (Lorenz 2005: 347).” Lorenz found that the Lutz index “tends to explain only those amendment data which he presented (Lorenz 2005: 354),” while the Anckar and Karvonen index “tends to confirm the hypothesis about an inverse correlation between the difficulty and the contemporary amendment rates (Lorenz 2005: 347).” Lorenz found Lijphart’s index “tends to result in the consistently weakest relationship between rigidity and amendment rates (Lorenz 2005: 355).” However, a more interesting distinction in the literature is if the approach uses institutional measures exclusively to assess constitutional rigidity, or if it combines it with other criteria that measure the frequency of amendments (and indicators this frequency depends on). Lutz has used this combined approach, but the most recent and sophisticated effort has been done by the Comparative Constitutions Project (CCP) of Elkins, Ginsburg, and Melton. This is why I have chosen to focus on it. Here I will summarize the argument, and later, in the end of section II, I will use the theoretical arguments to scrutinize it. The CCP authors start with the assessment that constitutional rigidity should be calculated by the combination of the institutional procedures required to amend and the actual (in)frequency of amendment. According to them, neither of these two components is sufficient on its own. And while the institutional component can be assessed (with difficulties that they enumerate and that this lit review corroborates) by looking at the constitution, the frequency of amendment depends on a host of social and historical factors: “Thus, we regress the amendment rate on a set of amendment procedure variables as well as a host of factors that should predict political reform more generally, including those factors included in our model of constitutional duration (Ginsburg and Melton 2015: 695).” Such factors include percentages of different ethnic groups, economic development, amendment rate, amendment rate squared, etc. (Elkins et al. 2009: 227-28). I will argue below why the inclusion of social and historical indicators may affect amendment frequency, but cannot affect constitutional rigidity. And that institutional arguments like: “constitutional rigidity has a negative effect on amendment frequency,” which most of this literature shares,2 cannot be accurately evaluated with the use of measures that include amendment frequency as an ingredient of constitutional rigidity. This is the reason that Tsebelis (2016) replaced the CCP measures by purely institutional variables to extend the Tsebelis and Nardi (2014) results from OECD countries to all democracies.

2 Lutz (1994): 365-366; Lijphart (2012): 211; Rasch and Congleton (2006): 542, as well as Dixon (2011: 106).

Given the variety of variables included, the reader will not be surprised by the table that follows, where I present the correlation coefficients among the different methods. For reasons of economy of space, I have included in this table the measure I propose in this article. Before I turn to this measure, however, let me make some remarks on the table that follows. All the authors mentioned above consider constitutional rigidity as an impediment to constitutional revisions. This is a rational expectation, because the reason that the founders of a constitution create all these restrictions is to make amendments more difficult, and therefore infrequent. Most of the literature finds a negative correlation between constitutional rigidity and the frequency of amendment (however each one of them is measured), although sometimes quite small. The three indexes used in Tsebelis (2016) called “constitutional rigidity 1, 2, and 3” are calculated on the basis of either the qualified majorities required for a successful amendment (either in percentage terms, or as a categorical variable), or the number of institutions required to approve. The fact that they are negatively correlated indicates that the founders in each country use mainly one of the two methods to protect the constitution. Actually, 89% of the countries that require just one body for constitutional changes also require a two-thirds majority or greater. Among countries that require two bodies, that percentage decreases to 63%, whereas only 52% of countries that require three bodies also require a two-thirds majority or greater. The most extreme countries (using only one of the two methods and generating the negative correlation) are Bulgaria and Mongolia on the one hand (requiring ¾ qualified majority from a single chamber) and Australia, Canada, Denmark, France, Iceland, Ireland, Italy, and Paraguay on the other, requiring the simple majority approval of 3 different bodies (usually including a bicameral legislature).It is interesting to note that “constitutional rigidity 1 and 2” are positively correlated with Lijphart’s indicators since they refer to the size of the required qualified majority alone, while the number of constitutional actors correlates with the other indicators. The fact that despite the negative correlation among them they are all associated with “time inconsistency” of long constitutions: as Tsebelis (2016) demonstrates, should be considered as an indication of the robustness of the finding. The fact that the CCP indicator of rigidity has many negative correlations with other indices indicates the significance of the departure from institutional variables, and the inclusion of social, economic and other ingredients in the calculation.

INSERT TABLE 1The index I introduce in this paper has a high correlation with the other indexes measured on the basis of institutional constraints (the number of veto players recognized by the constitution and the required majorities to each one of them), particularly with Lutz,

Lorenz, and Rasch and Congleton. But even if the correlation is low, its statistical significance remains high, despite the small number of observations (the intersection of countries selected by different authors). The major exception is with the CCP index, where the number of countries is high, the correlation negative, and its significance nonexistent. The numbers reflect a difference in approach (institutions in the case of my index, and a mixture of institutions and frequency in the case of CCP).

II. Constitutional Veto Players and Constitutional Core

Every constitution includes a series of articles that specify the rules concerning its future revision. The rules of constitutional revision specify a series of collective or individual actors (like one or both chambers of a legislature, or a special assembly, or a referendum, or an elected president) that are required to agree on the revised text in order for a revision to be valid. In other words, they specify the set of constitutional veto players. In the same or subsequent articles, the constitution specifies a series of provisions (required quorums, majorities, time constraints, and additional rules, such as intermediate elections) to make the decision of each one of these constitutional veto players valid. On the other hand, it is often the case that the constitution presents a series of alternative mechanisms for constitutional revisions (what Albert (2014) calls “multi-track”). If these alternative paths of revision are followed, the previously specified constitutional veto players are rescinded or quashed (partially or completely). In other words, the articles of constitutional revision set and (may) rescind constitutional veto players.

This section will analyze constitutional rigidity on the basis of the “constitutional veto players” defined by the constitution and the “constitutional core” that they produce. The constitutional core is the set of provisions that cannot be changed on the basis of the rules specified by the constitution. On the basis of this definition, it is obvious that the higher the required majorities (say 2/3 instead of 3/5, or simple majority), the more rigid (protected) the constitution. It is also obvious that the higher the required majorities, the more provisions will remain unchangeable, that is, the larger the constitutional core. As a special case, if the requirement for revision is unanimity (like the EU Treaties), then revisions become almost impossible. Similarly, the more bodies that are required to agree for a revision (for instance, a bicameral parliament with equal weight, as opposed to lopsided with one chamber making the final decisions, or unicameral), the more difficult it is to change the constitution. In this section, I will specify the implications of different rules on the size of the constitutional core.

Let us start with the location of the constitutional core. It is placed in the center of the political spectrum, because anything located farther away from the center (in any direction) will be overruled by the specified qualified majority. Consequently, in a democracy, one should not expect the constitution to include “extreme” positions, that is positions not shared by an overwhelming majority of the citizens. While this statement is true in general, the precise study of the institutional provisions by which constitutional revision become possible will enable us to deepen our understanding of constitutional rigidity.

The fundamental mechanism of constitutional revisions is a qualified majority of an existing legislative body (parliament) or some specifically elected institution (constituent assembly). To this qualified majority, additional institutions may be added, either as complements or as substitutes. If they are complements, the number of constitutional veto players increases; if they are substitutes, the previously defined constitutional veto players are rescinded. Let us examine these mechanisms of adding or subtracting veto players and the effect they have on the constitutional core.

Adding and subtracting veto playersLet us first consider the case of adding a veto player. Think of a requirement of a qualified majority of two different institutions (say a House and a Senate required to agree in identical text, or a unicameral legislature required to produce the same text twice, both before and after an election, or a referendum required for the approval of a text produced by a unicameral legislature). Figure 1 presents the cores C1 and C2 of the two required institutions.It is obvious that any outcome located inside the core of either institution belongs also to the bicameral constitutional core. Indeed, points X and Y are part of the constitutional core, since any modification of X would fail to command the required majority inside veto player 1, and any modification of Y would fail to command the required majority inside veto player 2. But the new constitutional core is not restricted to the combination (union) of the two cores. It includes also the whole area between the two cores. Indeed, if one connects X and Y with a straight line, the segment XX’ is located inside the core of veto player 1, and consequently cannot be modified; similarly, the segment YY’ is located inside the core of the veto player 2 and cannot be modified. The segment X’Y’ also cannot be modified, because each one of the two veto players would like to pull points in this segment in opposite directions. Consequently, the whole segment XY is part of the bicameral core (no matter where X and Y are located inside the cores of veto players 1 and 2). In other words, the addition of a second institutional veto player does not simply extend the core to include the core of this second veto player, but all the area between the two cores (the area A1A2B1B2 in the Figure). This is a general statement relating to any additional constraint the constitution may introduce. It never reduces the previous existing core. INSERT FIGURE 1Let us now consider the opposite case: what happens if the constitution, instead of adding constraints, adds alternative methods of revision. Figure 2 presents a case of alternative provisions for constitutional revisions. Consider that besides three-fourths approval by a bicameral legislature (represented by chambers A1A2A3 and B1B2B3), the constitution requires either an approval by a referendum (represented by player P) or by an elected president of the Republic (represented by player Q). On the basis of the previous analysis, the bicameral core would be the whole area A1A2A3B3B2B1. The additional requirement of a referendum would expand the core to the area A1A2PB2B1, while the alternative route of asking for the approval of the President of the republic would generate the core A1A2QB2B1. However, the dotted areas in the picture are not parts of the constitutional core of the country. The reason is that the points in the dotted areas can be modified by either one of the two permissible mechanisms (either the referendum or

the president). The constitutional core will be the intersection of the two possible cores (the shaded area in Figure 2).

INSERT FIGURE 2The two Figures demonstrate the logic of constitutional revisions. Their extent depends not only on the institutional provisions, but also on the positions of the actors involved. For example, in Figure 1, the constitutional cores of the two chambers could be smaller or overlap, leading to a reduction of the size of the constitutional core, or could be larger and further away from each other, leading to its expansion. Similarly, in Figure 2, one of the two procedures could become easier than the other (for example, if Q is inside the triangle PA2B2, then the intersection of the two cores will be identical with the core requiring approval by Q, (and P will become irrelevant)).However, there are two rules that will produce stable effects on constitutional cores. The first is that adding constraints will never reduce a constitutional core (although it may not affect it depending on the positions of the actors). The second is that adding alternatives will never expand it (although, depending on the position of actors, it may result in no change). These are the two rules that I will use extensively in the calculation of constitutional cores of the different countries.On the basis of this analysis, the institutional structure specified in the constitution is the requirement for constitutional amendments, and cannot be ignored in a democracy. Whether a country requires revisions of the constitution or not will depend on many factors, like the preferences of the different actors, social or economic changes, etc. However, how successful the attempts to change the constitution will be depend on constitutional rigidity (that is, the rules of how the constitution can be changed, or, in our analysis, the size of the constitutional core). No matter what the prevailing conditions in a country, constitutional revisions will be less successful if the qualified majority threshold increases (say from 3/5 to 3/4) or the number of actors required (veto players) increases (say from 2 to 3). Mixing constitutional constraints with social conditions is likely to create confusion.For example, Ginsburg and Melton (2015) regress the frequency of constitutional amendments in a country’s constitution on a series of institutional constraints taken separately (amendment threshold, number of proposers, number of approvers, multiple sessions, judicial review, etc.), as well as the frequency of constitutional amendments of the previous constitution, which they call “political culture.” They find that their “political culture” variable is the only significant one in their regression. There are several problems with the approach. First, what they call “political culture” is a combination of many different economic, social, and political variables. Second, the inclusion of such a lagged dependent variable (which may account for the “demand” of revisions) is likely to absorb not only the variance attributable to these omitted social and economic variables, but also to absorb the effect of the included ones, particularly serial correlation increases by the inclusion of so many distinct institutional variables.3

III. The constitutional core of a country: Italy3 Achen (2001:1) argues that “When one or more lagged values of the dependent variable are added “as a control … in many instances the autoregressive terms are strongly significant and the fit improves sharply, but the original sensible substantive effects of other variables disappear. This pattern frequently occurs even when the lagged variables have no plausible causal interpretation.”

Let me now apply the above rules to the calculation of the constitutional core of Italy, as a hands-on experiment, which involves several restrictions as well as alternative procedures. According to the Italian Constitution (Part II, Title VI, Section II, article 138): “Laws amending the Constitution and other constitutional laws shall be adopted by each House after two successive debates at intervals of not less than three months, and shall be approved by an absolute majority of the members of each House in the second voting.”Said laws are submitted to a popular referendum when, within three months of their publication, such request is made by one-fifth of the members of a House or five hundred thousand voters or five Regional Councils. The law submitted to referendum shall not be promulgated if not approved by a majority of valid votes.A referendum shall not be held if the law has been approved in the second voting by each of the Houses by a majority of two-thirds of the members.”This set of provisions specifies two alternative procedures. Constitutional revision can be achieved either by two-third majorities in both chambers of the legislature, or by simple majorities plus a referendum. For each one of these procedures, there are some additional constraints (i.e. double passage by the same legislature after a time interval). Let us try to calculate the core of each one of these alternatives, as well as assess the final outcome of article 138.Figure 3 demonstrates the core of the first procedure (two-thirds majorities in both chambers). I represent each chamber by a pentagon, and look for agreement of four out of the five members to achieve the required threshold. The pentagon CL is the core of the lower chamber. Indeed, any point outside this pentagon can be defeated by its projection on the pentagon, since four out of the five members of the Lower chamber would prefer this solution. Similarly, the pentagon CU is the core of the Upper chamber. On the basis of the argument presented for Figure 1, the bicameral constitutional core is comprised not only by both of these cores, but also by the area between them (the shaded area in Figure 3).In this Figure I have deliberately selected two chambers with preferences that do not overlap, in order to have a clear picture. If the preferences were overlapping, the two pentagons would come closer, and the bicameral core would shrink.

INSERT FIGURE 3Now let us look at the second possible procedure for constitutional revision, which requires a simple majority in both chambers plus a referendum. In this case, the bicameral core by simple majority would be along the line L2U2. Indeed, this line is a “bicameral median,” that is, it has a majority of members of both houses on it and on either side of it. It follows that there is a majority in both houses that would prefer, over any point outside this line, its projection on the line itself. The bicameral core, however, does not extend from point L2 to U2. Anything outside the solid line in the figure does not command a majority in one of the chambers. So, the bicameral core is the solid line between L2 and U2. In order to calculate the whole constitutional core, that is, include the referendum requirement, one must add the core of the voters (point P) and expand the core the way Figure 1 indicates, that is, include the whole shaded area that looks like a triangle.Again, I have selected on purpose to consider distinct preferences. If the two houses come closer to each other, or if the voters have preferences similar to one of them, the triangle will shrink, but it will not affect the logic of the argument.

INSERT FIGURE 4In order to visualize the core of the Italian constitution, it is now time to consider the intersection of Figures 3 and 4. This intersection is represented by the shaded area in Figure 5. Again, changing the location of the preferences of the different actors could significantly change the area of the core. In particular, if the preferences of the public move somewhere between the two chambers, it will be significantly easier to modify the constitution with a referendum than with the three-fifths majorities of both chambers. And this is a point that we should retain for the rest of the analysis. It cannot be a priori assessed which one of the two procedures is easier to be used. It depends not only on the institutional rules, but also on the actual preferences of the actors.

INSERT FIGURE 5

The section above provided a specific example of how veto players theory can be used to calculate the constitutional core of a country, as well as to point out that, with respect to specific issues, one would need to know the preferences of the actors in order to make specific assessments. The algorithm of the calculations, however, remains the same,4 and it is to this issue we now turn in order to calculate the constitutional rigidity in all democratic countries.

IV. Calculating the constitutional rigidity of democratic countries.For this calculation, I use the constitutional text (date 2013) of all the “democratic” countries, that is, countries scoring above 6 in the Polity indicator. When there are several alternative procedures, I measure only the one that is presented in the constitution first, because, as I have argued, there is no consistent way to evaluate which one is “easier,” given that the size of the core depends on the policy positions of the different actors. I also focus on the “usual” method of constitutional revisions, not exceptional ones (where different procedures are used for explicitly enumerated articles of the constitution).The fundamental method is the summation of approval thresholds of different elected institutions. For all countries, any elected body that must approve a constitutional amendment is included in the formula with the value representing the threshold by which approval must be granted. Included in this formula, if applicable, are the executive, legislative (counted as a single body), people, and states. For example, if a legislative body must pass an amendment by a simple majority, 0.50 is added to the formula. If an intervening election is required between two rounds of majority approval, 0.5+0.5 is added. The only exception to this rule is that bicameral legislatures are calculated as 1.5 of the required threshold. The reason is that it is significantly more difficult to pass measures from a bicameral legislature than from a unicameral one, however, in bicameral countries, the founders (being used to the existence of a second chamber), simply specify the threshold. A more precise approach would be to take into account the electoral system used for the election of each chamber, and increase or decrease the coefficient as a function of the dissimilarity of the composition of the two chambers.The epsilon rule. In addition, I am incorporating any modification of the rules that makes constitutional amendments more or less difficult than specified in the fundamental

4 Tsebelis (2002: 32) demonstrates that even if the core does not exist (which will happen if the issues under consideration are highly multidimensional) the same argument can be made in terms of the winset of the status quo, which exists when the core is empty.

method by adding or subtracting an epsilon (a small number (=.01)) for any provision that would increase or decrease rigidity. For example, if there is a provision of the percentage of members for a quorum, if there is a requirement for a double passage, or if there is a delay from one passage to the next, an epsilon is added to the formula. If there is an alternative procedure specified, an epsilon is subtracted (see section II).This way, all the rules specifying constitutional rigidity are incorporated. What is left out because, by definition, it consists of contextual factors, is the policy positions of the different actors. For example, it is possible that in a country that requires approval by a bicameral legislature and a referendum, the position of the people is between the House and the Senate, in which case, they should not be included in the calculations (they would be absorbed as a constitutional veto player). Yet, the formula would include an additional 50% for the referendum, despite the fact that, if the measures are approved by the two houses, they would not be rejected by the referendum.The above rules use Veto Players analysis to produce the column “VP Rigidity”. This indicator can be used to produce a more accurate idea about the effect of constitutional rigidity. Tsebelis (2016) has used constitutional rigidity (as measured by the provisions of the constitution) and amendment frequency (the number of times a constitution was amended divided by the years of existence of the constitution under a democratic regime) in order to calculate the “time inconsistency of a constitution,” that is, the frequency of revisions despite locking. He used easily available indicators, like the qualified majority required for constitutional amendments, or the number of actors required to be involved in a revision. He found that long constitutions were time inconsistent, but he concluded: “As for constitutional research, the elaboration of a more detailed (in terms of procedures) as well as encompassing (in terms of countries) measure of constitutional rigidity would certainly improve one’s understanding of the issue of time inconsistency.” As calculated above, VP Rigidity can be used to produce a more accurate indicator of time inconsistency and examine the relationship with the length of a constitution. Figure 6 provides the visual image of the relationship, and corroborates the Tsebelis (2016) findings: indeed, long constitutions are more time inconsistent than shorter ones. The effect is more pronounced for OECD countries than for the rest of democracies, as Tsebelis and Nardi (2014) anticipated.

CONCLUSIONS

Constitutional rigidity has been considered by many authors in the literature as a method to protect constitutions from change. While the empirical literature only slightly corroborated the expectation, two different attempts to gap the discrepancy developed. One is the “culture of amendment” approach by Ginsburg and Melton (2015), and the other is the “time inconsistency” approach by Tsebelis (2016). According to the first, amendment institutions are irrelevant, and should be replaced by culture. According to the second, the lack of correspondence between locking and frequency of amendments characterizing long constitutions is due to their restrictive character, which forces the countries to undertake the significant required efforts to achieve constitutional revisions.However, all these analyses were based on a limited number of countries, or on not very well-developed indicators. This article aims at producing a theoretically-based,

empirically accurate, and encompassing indicator. The use of this indicator corroborates the Tsebelis (2016) findings about time inconsistency of long constitutions.In addition, constitutional rigidity is associated with the importance of Constitutional Courts in different countries. Tsebelis (2002) has made the argument, at the theoretical level, that the larger the size of the constitutional court, the more fearless the judges that they will not be overruled by the political system. The evidence for this proposition is usually restricted to ordinary courts (Andrews and Monitola (2004). The only exception is Santoni and Zucchini (2004), who have examined the Italian Constitutional Court from 1956-1992 and found that the frequency that it disputes the constitutionality of laws increases when the constitutional core increases (the core in their case increases as a function of the policy positions of the same parties). The index I present in this article can be used to assess court independence and/or significance across the democratic countries of the world.

BIBLIOGRAPHY

Albert, R. (2014). “The Structure of Constitutional Amendment Rules.” Wake Forest Law Review (913-975).

Anckar, D. and Karvonen, L. (2002). “Constitutional Amendment Methods in theDemocracies of the World.” Paper delivered at the 13th Nordic Political ScienceCongress, Aalborg, Denmark, 15–17 August 2002.

Andrews, J. T., and Montinola, G. R. (2004). "Veto players and the rule of law in emerging democracies." Comparative Political Studies 37.1: 55-87.

Dixon, R. (2011). “Constitutional Amendment Rules: a Comparative Perspective.” In The Research Handbook in Comparative Constitutional Law. Edited by Rosalind Dixon and Thomas Ginsburg. Northhampton, MA: Elgar Publishing.

Elkins, Z., Ginsburg, T., and Melton, J. (2009). The endurance of national constitutions.Cambridge University Press.

Elster, J. (2000). Ulysses Unbound: Studies in Rationality, Precommitment, andConstraints..Cambridge University Press.

Ginsburg, T., and Melton, J. (2015). "Does the constitutional amendment rule matter atall? Amendment cultures and the challenges of measuring amendment difficulty."International Journal of Constitutional Law, 13(3), 686-713.

La Porta, R., Lopez-de-Silanes, F., Pop-Eleches, C., and Shleifer, A. (2004). “JudicialChecks and Balances.” Journal of Political Economy, 112(2), 445-470.

Lane, J. E. (2011). Constitutions and Political Theory. Second edition. ManchesterUniversity Press.

Lijphart, A. (2012). Patterns of democracy: Government forms and performance inthirty-six democracies." Second edition. New Haven, CT: Yale University Press.

Lorenz, A. (2005). "How to Measure Constitutional Rigidity: Four Concepts and TwoAlternatives." Journal of Theoretical Politics, 17(3), 339-361.

Lutz, D. S. "Toward a Theory of Constitutional Amendment." American Political

Science Review 88, no. 2 (1994): 355-370.

Marshall, M. G. “Polity IV Project: Political Regime Characteristics and Transitions,1800-2013.” Published by Societal-Systems Research Inc. Last accessed January 25, 2016. Available at: www.systemicpeace.org/polity/polity4.htm.

Rasch, B. E. and Congleton, R. D. (2006). “Amendment Procedures and Constitutional

Stability,” Edited by Roger D. Congleton and Birgitta Swedenborg. In Democratic Constitutional Design and Public Policy: Analysis and Evidence. Cambridge, MA: MIT Press, 319-342.

Santoni, M., and Zucchini, F. (2004) "Does policy stability increase the constitutional court's independence? The case of Italy during the first republic (1956–1992)." Public Choice 120(3-4), 439-401.

Schneier, E. (2006). Crafting Constitutional Democracies: The Politics of Institutional Design.

Tsebelis, G. “The Time Inconsistency of Long Constitutions: Evidence from the World.”Forthcoming.

Tsebelis, G., and Nardi, D. J. (2014). "A Long Constitution is a (Positively) BadConstitution: Evidence from OECD Countries." British Journal of PoliticalScience. Available on CJO, doi: 10.1017/S0007123414000441.

TABLE 1Correlations among different measures of constitutional rigidity

FIGURE 1

FIGURE 2

FIGURE 3

FIGURE 4

FIGURE 5

FIGURE 6