Differences in Institutional Quality across Euro Area Countries: Which Factors Contribute Most to Inequality?

In recent years the differences in the institutional structure across the Euro area countries have become a cause of concern, both for some individual Member States and for the functioning of the Economic and Monetary Union (EMU). In this paper we analyse the inequality in institutional quality across Euro area countries and estimate which factors of public and private institutions contribute most to overall inequality in institutional quality. To this end, we consider the institutional indicators of the Global Competitiveness Index (GCI) from the World Economic Forum (WEF) during the period 2007-2017, with the most disaggregated data possible. Our findings support the call for structural reforms, particularly in the areas of ethics and corruption (in the public sphere but also in the business environment), undue influence on the judiciary and government decisions, and protection of property rights, as the major sources of inequality in institutional quality.

A vast literature shows how institutions matter for economic performance, focusing on aspects such as the relationship between institutions and economic development (see e.g. Dani Rodrik, Arvind Subramanian, and Francesco Trebbi 2004; Daron Acemoglu and James A. Robinson 2012; Augustin K. Fosu 2016) or the effects of institutions on income inequality and welfare (see e.g. Alberto Chong and Mark Gradstein 2007; Philip Arestis and Ana R. Gonzalez-Martinez 2016;Kosta Josifidis, Novica Supić, and Emilija Beker Pucar 2017). In general terms, there is a consensus in the argument that good institutions lead to good economic outcomes, and vice versa. This way, it is frequently pointed out that the institutional inequalities among countries would be a relevant element behind the differences in the economic performances among countries, even though institutional convergence has not necessarily come with a process of economic convergence (see e.g. Jesus Ferreiro et al. 2017).
In the context of the Economic and Monetary Union (EMU), nowadays the substantial differences in institutional quality across Euro area countries, apart from being a cause of concern for individual Member States, are increasingly regarded as a risk for the smooth functioning of the Eurozone. Thus, institutional convergence tends to 1. Data and Methodology

Data
The GCI is based on a weighted average of 12 pillars, which represent different dimensions of competitiveness. The first one is dedicated to institutions (P1), distinguishing between public institutions (A) and private institutions (B) (first level of disaggregation). Each indicator is rated on a scale of 1 to 7, where 1 represents the worst possible situation and 7 the best, according essentially to the views of business leaders 2 (Klaus Schwab 2017).
The WEF assess five components of public institutions (second level of disaggregation), each one including various subcomponents (third level of disaggregation) (see Table A1 of the Appendix summarising components, subcomponents and weights): 1) Property rights (A1), which rates the level of protection of property rights, including financial assets (A1.1), and intellectual property protection (A1.2); 2) Ethics and corruption (A2), which deals with three questions: (a) diversion of public funds to companies, individuals, or groups due to corruption (A2.1); (b) the ethical standards of politicians (A2.2); and (c) undocumented extra payments or bribes connected with imports and exports, public utilities, annual tax payments, awarding public contracts and licenses, and obtaining favourable judicial decisions (A2.3); 3) Undue influence (A3), which captures the extent to which the judiciary is independent from influences of members of government, citizens, or firms (A3.1.), and the extent to which government officials show favouritism to well-connected firms and individuals when deciding upon policies and contracts (A3.2); 4) Government efficiency (A4), which assesses five points: (i) the extent to which the composition of public spending is extremely wasteful or highly efficient in providing necessary goods and services (A4.1); (ii) how burdensome it is for businesses to comply with governmental administrative requirements (A4.2); (iii) how efficient the legal framework is for private businesses in settling disputes (A4.3); (iv) how easy it is for private businesses to challenge government actions and/or regulations through the legal system (A4.4); and (v) how easy it is for businesses to obtain information about changes in government policies and regulations affecting their activities (A4.5); 5) Security (A5), which addresses the extent to which the threat of terrorism (A5.1), the incidence of crime and violence (A5.2), and organized crime (mafia-oriented racketeering, extortion) (A5.3) impose costs on businesses, as well as the extent to which police services can be relied upon to enforce law and order (A5.4). As regards private institutions, two components (second level of disaggregation) and several subcomponents (third level of disaggregation) are provided, as follows: 1) Corporate ethics (B1), which refers to ethical behaviour of companies in interactions with public officials, politicians, and other firms (B1.1); 2) Accountability (B2), which jointly rates four aspects: (a) auditing and reporting standards (B2.1.); (b) efficacy of corporate boards (whether management is accountable to investors and boards or not) (B2.2); (c) protection of minority shareholders' interests by the legal system (B2.3); and (d) strength of investor protection (it rates transparency of transactions, liability for self-dealing, and shareholders' ability to sue officers and directors for misconduct) (B2.4).
Whereas other studies such as European Central Bank (2015) focused on disparities in indicators of governance in the Euro area, our analysis adopts a broader perspective, considering a wide range of indicator or public and private institutions. Thus, Figure 1 shows the evolution of the mean overall institutional quality in the Eurozone, discriminating between public and private institutions, while Figure    In general terms, there is a decreasing pattern in the assessment of institutional quality, particularly between 2008 and 2014, coinciding with the period of economic downturn. In fact, it is widely known that the state of institutions can be (negatively) affected by economic downturns, as was the case during the Great Recession (  This study goes further and focuses on institutional differences across Euro area countries. The objective is to examine the extent to which national institutional changes in recent years have increased or decreased cross-country inequality in institutional quality in the Eurozone, and which institutional components and subcomponents contribute most to overall inequality, in order to provide specific policy recommendations to reduce large institutional differences in the EMU.

Methodology
The mean value of the institutional quality can be used to summarize its distribution across Euro area countries, but it is not a complete description, since institutional quality differs from one country to another. In this paper we measure the inequality in institutional quality through the well-known Gini coefficient. We consider a fixed homogeneous population N {l, 2, . . . , n} of n (n ≥ 2) countries. They are identical, but generally differ in their institutional quality. A feasible distribution is given by a vector ( , , . . . , ) ∈ where is country i's institutional quality index at period , = 1, 2, … , , ≤ ≤. . . ≤ and the mean value of the institutional quality at time . The Gini coefficient (Corrado Gini 1912) at time is: (1) is always between 0 (no inequality) and 1 (maximum inequality). Alternatively, the Gini coefficient can be written as: (2) This expression shows that the Gini coefficient is a linear function, in line with Farhad Mehran (1976). The Gini coefficient can be decomposed to assess the contribution of each component or subcomponent to inequality. Let denote the value of country for component or subcomponent of the institutional quality and let = ∑ represent the overall institutional quality with ∑ = 1 . Table 1 shows how the so-called first pillar of the GCI "institutions" is simply the sum total of weighted components or subcomponents; therefore, Expression (2) may be alternatively written following Anthony F. Shorrocks (1982) as: where is the mean of and is known as the pseudo-Gini for factor k. This expression seems to have been introduced by John C. H. Fei, Gustav Ranis, and Shirley W. Y. Kuo (1978) and was subsequently used by Jacques Silber (1989), among others. Note that the pseudo-Gini is not the Gini of the component or subcomponent k, as is arranged in ascending order of ,which does not necessarily agree with the ascending order of .
We can evaluate the contribution of component or subcomponent k to overall institutional inequality as: a combination of the weight attached to each component or subcomponent, the mean quality of such component or subcomponent, the mean overall institutional quality and the pseudo-Gini.

Results
The mean value of each institutional indicator gives partial information on the evolution of institutional quality across Euro area countries that can be complemented with a measure of cross-national disparity. We report the Gini coefficient in Table 1, highlighting that the heterogeneity in the assessment of institutional quality across countries increased from 2007 to 2017 in all the components of public and private institutions. This increasing pattern of heterogeneity shows a steeper slope for Government efficiency (A4), precisely the subcomponent worst mean rated (Figure 2), and the least pronounced slope for Accountability (B2). This component, together with Security (A5), is the most homogeneously assessed component during the period, while the most heterogeneous are Ethics and corruption (A2) and Undue influence (A3). By comparing Figure 2 and Table 1, we can verify that there is a negative correlation (-0.79) between the mean value of the institutional components and their inequality, so that those with better assessments are also the components with less heterogeneity. In order to know the contribution of each institutional component and subcomponent to overall inequality in the quality of institutions, we apply the methodological approach described above. This computation may provide relevant policy findings, as it allows identifying the main sources of inequality in institutional quality across Euro area countries, and can therefore shed some light on which factors require more attention for reducing cross-country disparities in institutional quality. Table 2 Table 2, and its contribution for each year is shown below 3 . The contribution of public institutions to overall inequality is greater than the expected contribution imposed by its definition (75%) in all years, reaching its highest value in 2015. By contrast, the contribution of private institutions to the overall inequality is obviously always lower than the expected contribution (25%) As shown in Figure 3, the components Ethics and corruption (A2) and Undue influence (A3) are the main sources of inequality. Their contribution slightly decreases during the period examined, although they continue to be the main sources of inequality in overall institutional quality. While the countries with the greatest decreases in institutional quality are Spain, Malta and Greece for A2 and Slovakia, Malta and Slovenia for A3, for both indicators the countries that improve the most are Estonia, Ireland and Lithuania (see Table A2 of the Appendix). On the other hand, Security (A5) and Accountability (B2) are the components that contribute least to inequality, showing a lower contribution at the end of the period. Finally, A4 (Government efficiency) and A1 (Property rights) increase their contribution to inequality during the period, even though they remain in intermediate positions. In this sense, Greece, Slovakia and Latvia for A4, and Greece and Germany for A1, are the countries that to a greater extent get worse over the period examined (see Table A2 of the Appendix). Source: Own construction based on WEF (2017).

Figure 3 Evolution of the Contribution of Each Component to Overall Inequality in Institutional Quality (%)
Given that each component comprises a variety of institutional aspects, we go a step further and estimate the contribution of each subcomponent (third level of disaggregation) to overall inequality. As shown in Table 3, by subcomponents the main source of inequality in institutional quality is Ethical behaviour of firms (B1.1) (Spain and Slovakia are the countries with the greatest declines in this subcomponent according to Table A3 of the Appendix), followed by Judicial independence (A3.1) until 2015, when the second highest contribution to inequality becomes Property rights (A1.1) (in this case, Greece, Italy and Latvia are the countries showing the worst evolution). On the contrary, the smallest source of inequality in almost all years is Strength of investor protection (B2.4) (Slovenia and Italy are the countries with the highest increase), followed by Business cost of terrorism (A5.1) (Spain being the country with the greatest improvement), both reducing their contribution during the period analysed. We report the evolution of the contribution of these selected subcomponents in Figure  4.
A useful way to summarize the ranking of sources of inequality during the years analysed, that is, the contributions of the components and subcomponents to overall inequality in institutional quality, is with a dominance diagram ( Figure 5 for institutional components and Figure 6 for institutional subcomponents).

Source: Own construction based on WEF (2017).
Source: Own construction based on WEF (2017).    In Figure 5 we can identify three substantial groups of institutional components during the period 2007-2017 in accordance with their contribution to overall inequality in the quality of institutions. Security (A5) and Accountability (B2) are the institutional components that contribute least to overall inequality, while Ethics and corruption (A2) and Undue influence (A3) are the ones that contribute most to inequality. Government efficiency (A4), Property rights (A1) and Corporate ethics (B1) are in the middle ground, with intermediate contributions.
If we take into account a higher level of disaggregation (third level), Figure 6 shows that it is not easy to rank the contributions of the respective institutional subcomponents consistently during the period analysed, especially those in the upper part of the ranking (those with a lower contribution to inequality).
In any event, we can identify with some clarity four groups of subcomponents. First, Ethical behaviour of firms (B1.1) is at the bottom of the dominance diagram, showing that this subcomponent is the one that contributes most to overall inequality in institutional quality, irrespective of the year examined. Then, Judicial independence (A3.1), Favouritism in decisions of government officials (A3.2) and Property rights (A1.1) form the second group of great contributors to inequality, followed by a third group of subcomponents comprising Intellectual property protection (A1.2), Diversion of public funds (A2.1), Public trust in politicians (A2.2) and Irregular payments and bribes (A2.3).
The remaining subcomponents constitute the group of lowest contributors. We observe that there is not a clear dominance in this group, so that only Business costs of terrorism (A5.1) can be highlighted by its dominance over most of the other elements of this group.

Concluding Remarks
This study highlights that institutional quality not only decreased on average across the Euro area countries during the period 2007-2017, but that its inequality also increased considerably, as measured by the Gini coefficient, particularly in the case of public institutions.
In order to identify the institutional factors that contribute most to overall inequality in institutional quality, we disaggregate institutions into public and private institutions and their respective components and subcomponents (three levels of disaggregation). Our analysis reveals the significant contribution to overall inequality of the components of public institutions related to ethics and corruption and undue influence, both highly distorting and detrimental practices for economic performance. Specifically, seven countries worsen their institutional quality in both indicators simultaneously between 2007 and 2017, namely: Austria, Cyprus, Greece, Malta, Portugal, Slovakia and Slovenia; whereas other countries such as France, Germany, Italy, Latvia and Spain worsen in only one of these institutional components, although in some cases (Germany, Latvia or Spain) significantly.
Conversely, security and accountability are more homogeneous across Eurozone countries, showing lower contributions. Going a step further in the components of institutional quality, we disaggregate at a third level and conclude that both judicial independence and favouritism in government decisions (undue influence) and the threesome public trust in politicians, diversion of public funds and irregular payments and bribes (ethics and corruption) act as major sources of inequality, jointly with other factors such as ethical behaviour of firms, property rights and intellectual property protection. In general, most countries have worsened their overall institutional quality, except six countries that have improved it, namely: Estonia, Finland, Ireland, Lithuania, Luxembourg and Netherlands.
These empirical findings tend to support the call for structural reforms to reduce the institutional gap among Euro area countries, as large differences in institutional quality endangers the sustainability of the EMU and puts its long-term prosperity at risk. In particular, the decline in institutional quality of the countries with the lowest levels such as Greece, Italy, Latvia or Slovakia should not be only a concern for these countries, but also for the Eurozone as a whole. Nevertheless, in line with Diaz del Hoyo et al. (2017), this does not imply the need of converging towards a single institutional model for all countries; rather, it points to a need to find solutions that are tailored to country-specific situations (e.g. differing political or cultural preferences of citizens, pre-existing institutional settings, etc.).
Although much of the focus in the reform debates usually is on the product and labour market, reforms enhancing institutional quality, particularly in the areas of ethics and corruption (not only in the public sphere avoiding diversion of public funds or irregular payments and bribes, but also in the business environment), undue influence on the judiciary and government decisions, and protection of physical and intellectual property rights, should be a priority for both EU-wide and domestic policies. The Eurozone countries need to strive for convergence towards high institutional quality particularly in these areas in order to achieve a less vulnerable and more resilient EMU.