Estimating Judicial Ideal Points in Bi‐Dimensional Courts: Evidence from Catalonia

• Author: Nuno Garoupa,Marian Gili,Lucia Dalla Pellegrina
• Date: 01 June 2020
• Published date: 01 June 2020
• DOI: http://doi.org/10.1111/jels.12251

Journal of Empirical Legal Studies

Volume 17, Issue 2, 383–415, June 2020

Estimating Judicial Ideal Points in

Bi-Dimensional Courts: Evidence from

Catalonia

Lucia Dalla Pellegrina,*Nuno Garoupa, and Marian Gili

The empirical literature has consistently shown that judicial ideal points can be estimated

in a one-dimensional space that reflects the traditional conservative-progressive dichotomy.

In this article, we develop an empirical methodology to analyze some features that may

characterize bi-dimensional courts when dissenting opinions are not frequent. We apply

the analysis to the particular case of the Consell de Garanties Estatuta`ries de Catalunya

(Catalan Constitutional Court). The results illustrate that judicial preferences on

conservative-progressive grounds are likely to affect the decision outcome of the Court on

issues having significant public policy content. Conversely, judicial preferences regarding

Spanish-Catalan sovereignty tend to encroach on judgments concerning public policy,

thus affecting the outcome of all judgments of the Court regardless of content type. Fur-

thermore, we find that judicial preferences in the Spanish-Catalan sovereignty dimension

are pervasive enough to shift the outcome of all types of decisions in favor of Catalan insti-

tutions. Policy conclusions are derived.

I. Introduction

Ideal point estimations are still viewed as a helpful measurement to predict judicial

behavior. Although empirical results suggest that judges might not have temporally con-

stant ideal points (Martin & Quinn 2002), they seem to correlate significantly with gen-

eral perceptions of whether a judge is conservative or liberal (see, e.g., Segal & Spaeth

2002). In the context of civil rights, civil liberties, economics, and issues related to feder-

alism, judicial ideal points could evolve over time, although there is no consensus on the

absolute determinants of their evolution. Therefore, different explanations have been

*Address correspondence to Lucia Dalla Pellegrina, Department of Economics, Management and Statistics

(DEMS), University of Milano-Bicocca, Italy; email: lucia.dallapellegrina@unimib.it. Garoupa is at George Mason

University, Scalia Law; Gili is at Universitat Oberta de Catalunya.

We are grateful to two anonymous referees, Vanessa Casado-Pe´rez, Gabriel Dome´nech-Pascual, Carlos Go´ mez-

Ligu

¨erre, Jose´ M. Reis, Laura Salamero-Teixido´ , Joan Solanes-Mullor, and conference participants at the annual

meetings of LAWLE (Brasilia 2017), SIDE-ISLE (Rome 2017), ALEA Meeting (Boston 2018), and AEDE (Lleida

2018) for excellent comments. Constance McKinnon has provided excellent research support. The usual dis-

claimer applies.

383

developed to contextualize static or dynamic preferences as revealed by ideal point

estimations.

The U.S. literature has now been expanded to other courts, thus providing for a

refined method for estimating an individual ideal point based on how each judge mani-

fests his or her views in dissenting and concurring opinions.

1

Technically, the empirical

method of estimation reveals those points in some n-dimensional space of politically rele-

vant choices that judges prefer over all other points in that space. The spatial voting

approach allows scholars to estimate judicial ideal points by ranking them in one or more

dimensions.

The conventional literature reduces the n-dimensional space of politically rele-

vant choices to one revealed dimension, usually characterized by a conservative-

progressive axis. Nevertheless, multidimensionality has warranted scholarly attention.

For example, political scientists working on the preferences of the U.S. Congress have

long argued that a better empirical representation is inevitably multidimensional

(Poole & Rosenthal 1985, 1991; Clinton et al. 2004; Poole 2005; Bailey 2007). How-

ever, empirical estimations of multi dimensions in the judicial context are still in early

stages, partially because the conservative-progressive axis seems to capture the dynam-

ics of U.S. federal courts.

2

In this article, we advance the empirical literature on multidimensional courts.

We analyze a specific example of an institution known to be part of a bi-dimensional

policy space, which is quite unique (Aragone`s 2007a, 2007b). In fact, the Consell de

Garanties Estatuta`ries de Catalunya (hereafter CGEC; we use the English

translation—Catalan Constitutional Court—subject to appropriate qualifications

about the nature of this institution

3

) has two fundamental axes in the relevant policy

dimensions. One is the traditional conservative-progressive axis, common to most con-

stitutional courts around the world. Inevitably, there are judges close to the right-wing

conservative parties and judges close to the left-wing progressive parties. However, a

second significant dimension coexists: those who favor Spanish sovereignty versus

1

See mainly Martin and Quinn (2002) as well as Jackman (2001), Bafumi et al. (2005), Peress (2009), Hanretty

(2012a, 2012b), Clinton et al. (2004), Dalla Pellegrina et al. (2012, 2014, 2017), and Gonza´ lez Bertomeu

et al. (2017).

2

Ferreira and Mueller (2014) identify two dimensions in the Brazilian Supreme Court. Rather than aligning along

the usual conservative-progressive dimension, the authors find that the main dimension is for or against the execu-

tive branch’s economic interests. A second uncovered dimension is for and against states’ autonomy to set their

own laws. Desposato et al. (2015) suggest that granting more authority to the Brazilian Supreme Court has encour-

aged broader ideological principles.

3

The institutional nature of the CGEC is a matter of debate. Strictly speaking, it is not a constitutional court since

Catalonia does not have a sovereign constitution and the decisions are not binding. Within the Catalan Statute of

Autonomy, it is a body of administrative review that controls the activity of the Catalan government and parlia-

ment. Its institutional role in no way competes or overshadows the Spanish Constitutional Court. However, given

its importance within the political dynamics of Catalonia, we have translated CGEC to “Catalan Constitutional

Court” in a very broad sense. A similar reasoning applies to its members. Technically, they are not judges in law

but advising councilors or council members exercising abstract review. However, again, due to the function they

play, we refer to the CGEC members as “judges.”

384 Dalla Pellegrina et al.

those who favor Catalan sovereignty.

4

Consequently, traditional Catalan politics is

dominated by two different dimensions, public policy and sovereignty. However, the

identification of these two dimensions is challenged by the fact that they are likely to

be correlated, although not perfectly. This creates serious empirical difficulties in the

adoption of standard spatial models, particularly when there is a small number of dis-

senting opinions from which to infer judges’ ideological preferences.

5

Our empirical analysis covers the period 2009–2015. The CGEC was created in

2009 in the context of the new Statute of Catalonia (Estatut d’Autonomia de Catalunya),

approved by the Catalan Parliament in 2006 (which resulted in a major constitutional cri-

sis in Spain). This is a period of Catalan political life dominated by constitutional chal-

lenges to the new Statute of Catalonia and the breakdown of the progressive/left-leaning

coalition that ran Catalonia since 2003 alongside the impact of the recession and finan-

cial crisis (including significant budget cuts). In fact, we can say that the strategies devel-

oped by the different political actors in this period climaxed in a new party system

emerging in the 2015 election (both in Catalonia and Spain) and the unilateral indepen-

dence crisis in November 2017. Accordingly, we use the 2015 Catalan elections as the cut

point delimiting our observation period, since they introduced a major change in the

local party system.

6

Our analysis covers a total of 124 decisions, resulting in 1,098 individ-

ual votes.

In the first part of the empirical investigation we concentrate only on cases with dis-

senting opinions, as this allows for the retrieval of judicial preferences based on their mis-

alignment with the majority. This also helps to interpret which priorities shape judicial

ideology, as some dimensions, typically linked by substantial correlation, could easily can-

nibalize each other. Then, we extend the analysis to the entire database in order to iden-

tify the characteristics of the cases in which one or the other type of ideological

dimension tends to prevail, with particular attention to the content of the issues brought

before the Court (i.e., public policy or sovereignty).

Our results show that both dimensions, conservative-progressive and Spanish-

Catalan sovereignty, significantly explain clustering in the CGEC, but in different ways.

Judicial preferences on one of these two dimensions (Spanish-Catalan sovereignty) affect

all types of decisions of the Court, regardless of the nature of the issues that are chal-

lenged. In addition, when Spanish-Catalan matters are addressed, the whole Court tends

to favor Catalan institutions (petitioners). Disputes dealing with the more conventional

4

By the Spanish-Catalan sovereignty axis, we do not mean a debate over conflicts of sovereignty strictly speaking

(since Catalonia is part of Spain and this legal discussion falls within the jurisdiction of the Spanish Constitutional

Court), but the allocation of powers and competences within legislation that affects Catalonia; fundamentally,

arguments over the defense of Catalan autonomy within Spain. Therefore, alternative denominations could be the

Spanish-Catalan “powers framework,” “competences allocation,” “territorial organization,” or “self-determina-

tion” axis.

5

See discussion by Padro´-Solanet (1996) and Aragone`s (2015).

6

See Aragone`s (2015), Orriols and Rodon (2016), Martı´ and Cetra` (2016), and Simo´ n (2016).

Estimating Judicial Ideal Points in Bi-Dimensional Courts 385