Convergence in corporate governance: a leximetric approach.

• Author: Siems, Mathias M.

1. Introduction II. Methodology. III. Difference Analysis of Individual Countries A. Differences Between French and German Law 1. Observations 2. Explanations B. Differences Between U.K., U.S., and Indian Law 1. Observations 2. Explanations IV. General Analysis on Legal Origins and Convergence V. Conclusion I. Introduction

   Has there been convergence of corporate governance systems? The current debate started with Hansmann and Kraakman. They expect "substantial convergence in the practices of corporate governance as well as in corporate law" because the Anglo American model of corporate governance has won the day. (1) Others object that historical and cultural differences between countries are likely to persist, reflecting different types of market economies. (2) This debate will re-emerge in the context of the financial crisis of 2007 to 2009. Assuming that corporate governance failures have contributed to the financial crisis, (3) commentators are likely to use the slogan that we need global solutions to this global problem to justify (further) convergence in corporate governance. (4)

   This Article examines whether national laws on corporate governance have become more similar. This focus on legal rules provides a link to the literature on the convergence of legal systems. Here too, some contend that in the modern world, legal differences, in particular the distinction between civil law and common law legal origins, have become less marked, (5) whereas others object that different legal mentalities still play an important role. (6) Modified positions are also possible. For instance, it could be said that today legal systems do not primarily differ because of different legal origins but due to their belonging to the EU, or due to the question of whether a country is developed or developing. (7) Another modification is that there is only a "weak legal origin" effect, which means that the effect of belonging to a particular legal origin varies over time, depending on the strength of pressures for convergence and for the "endogenization" of law to local conditions. (8)

   A crucial element of corporate governance is how well shareholders, creditors and workers are protected. This Article uses a new quantitative methodology ("leximetrics") (9) in order to answer the question of whether there has been convergence, divergence, or persistence of the rules on shareholder, creditor, and worker protection. In particular, this Article will analyze whether there are deep differences between the Anglo-Saxon countries (common law countries) and Continental Europe (civil law countries). The bases for this Article are three comprehensive indices for shareholder, creditor, and worker protection, which code the legal development of France, Germany, India, the United Kingdom, and the United States from 1970 to 2005. Part II describes these datasets and explains how they can be used to measure convergence or divergence of the law. Part III examines the differences and similarities between the five countries in detail. This will be supplemented by Part IV, which provides a more general analysis on convergence and legal origins. Part V concludes.

   Two caveats must be made. First, this Article is mainly interested in the legal rules that determine national differences of corporate governance. This focus on legal rules does not deny that their enforcement may differ between jurisdictions, or that non-legal considerations also determine corporate governance at the firm level. Secondly, this Article does not try to answer the "causality problem," namely whether legal convergence (or divergence) is mainly the result of factual changes (the "law follows" thesis), or whether law is predominately a source of factual changes itself (the "law matters" thesis). This point is well discussed in previous literature. (10) This Article will, however, analyze why particular legal changes have taken place. This does not mean that the law merely reacts. Rather, this analysis makes the realistic assumption that, at least to some extent, the law is influenced by factual changes.

2. Methodology

   The bases of this Article are three indices that code how well countries protect shareholders, creditors, and workers. These indices cover a wide range of variables: 60 variables for shareholder protection, 44 variables for creditor protection, and 40 variables for worker protection. (11) The indices are therefore very detailed in their legal coverage, with 144 legal variables coded for each country-year. The indices also extend over a relatively long time period: 36 years (1970 to 2005). As a limitation, however, only the laws for a small number of countries have been coded: France, Germany, the United Kingdom, the United States, and India. These countries are of particular interest because they include three "parent" legal systems: the United Kingdom, France, and Germany; (12) the world's largest economy: the United States; and the world's largest democracy: India. Thus, in total, these three indices code for 25,920 observations. (13)

   The full text of these indices and data (plus detailed explanations) can be found online. (14) Here, for purposes of illustration, it is sufficient to present extracts of the shareholder protection index (Table 1) and of the French and U.K. codings (Tables 2 and 3).

   We have explained the indices and coding methodology in more detail in other articles, in particular, in which instances we allow intermediate scores and how our indices differ from indices of the previous literature. (18) We have also used these indices to determine the strength of shareholder, creditor, and worker protection in France, Germany, India, the United Kingdom, and the United States. (19) Furthermore, it has been examined whether the strength of legal protection is reflected in a country's financial development. (20) These articles therefore respond to the "law and finance" literature which, based upon cross-sectional studies, claim to have proven that the "greater the protection afforded to shareholders and creditors by a country's legal system, the more external financing firms in that jurisdiction will be able to obtain." (21)

   The methodology and content of this Article is different from these previous ones. Here, the interest is not on the aggregates of legal protection but on the differences between the five countries. For this purpose, research for this Article involved calculating the differences between each variable in the law of a particular legal system, and the same variable in the law of the other countries and adding together the absolute values of these differences. For example, the formula for the differences between shareholder protection in France and the United Kingdom is

   [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]

   where SPFrance1970 and SPUK1970 stand for the 60 variables on the strength of shareholder protection in 1970. (22) Equivalent formulas have been used for the other 35 years, the other nine pairs of countries, and the other two indices. All of these mathematical operations therefore lead to 36 x 10 x 3 = 1080 observations, which form the basis of this Article.

   These observations indicate whether the laws of two legal systems have converged or diverged. For instance, in the first set of diagrams--concerning the differences from French law (23)--the score of "0" would indicate that the law of a particular country would be identical to French law. Using time-series allows tracing differences between countries that have developed in the last three and a half decades. For instance, the downward trend of the curve which displays the differences between shareholder protection in France and the United Kingdom (24) means that French and U.K. law have converged in the last decades.

   The past literature has distinguished between various types of convergences. In particular, in the context of the debate on globalization of corporate governance trends, a distinction is drawn between formal, functional, contractual, hybrid, normative, and institutional convergence. Gilson and Coffee assume that functional convergence is likelier than formal convergence. (25) "Functional" in this context means that a comparable result is produced with, say, bad managers being dismissed, but along different statutory paths. Alternatively, according to Gilson there may be contractual convergence, where the formal differences may be functionally relevant, but equivalent effects can also be reached through contractual arrangements. (26) Furthermore, the dualism between formal and functional convergence is supplemented by Rose with the concept of hybrid convergence. (27) Hybrid convergence concerns the situation where a firm "escapes" domestic law by shifting its registered seat to another country. (28) Outside the legal sphere, one may, with Milhaupt, raise the question of "normative convergence." (29) Here, "normative" means that the viewpoint of convergence is applied to extra-legal norms. (30) Further, Charny employs the term "institutional convergence," where de facto the structures in firms become more similar. (31) This point concerns, for instance, the question of whether the shareholder ownership structure of firms changes, or firms are more frequently exposed to market influences (such as the possibility of hostile takeovers).

   The methodology of this Article can show only whether there is a formal convergence, persistence, or divergence of legal rules. However, the question about formal convergence is also relevant for the other types of convergence. As long as there is formal convergence, the question of whether other forms of convergence may step in as substitutes becomes obsolete. Also, as long as there is formal convergence but de facto persistence, this can lead to further research regarding whether a "convergence of law and reality" may be expected in the future. (32)

3. DIFFERENCE ANALYSIS OF INDIVIDUAL COUNTRIES

   This Part uses the...