Search

To create this dataset we collected information of the 100 largest non-financial corporations, according to their share capital at four benchmark years (1923, 1937, 1944, and 1954) and to their annual sales at other three-benchmark years (1970, 1990 and 2000). The choice of these seven benchmark years primarily reflects the different stages in Argentina’s economic history, but it also factored in data availability, in that board membership information is not readily or systematically available for most of the century. The final sample included 100 non-financial firms for all benchmark years with the exceptions of 1970 and 1990. In 1970, the sample included 95 firms; we used 102 companies in 1990. We also included the 25 biggest banks per year, though in 1970 we had 28. Total deposits were the main criterion for selecting the banks. We lack information about insurance companies, which were not among the biggest firms. We included other financial companies that were not banks in the main top-100 firm rankings (six in 1923, five in 1937, nine in 1944, eight in 1954, and four in 2000). The total sample for each year thus included 125 companies, except for 1970 and 1990, which contained 123 and 127 companies, respectively. We collected data on board composition using various resources.

This dataset covers the composition of the boards of the 125 largest Portuegese firms (by assets) in several benchmark years. These years reflect the different contexts faced by business firms over roughly a hundred years— 1913 to 2010. The first year— 1913— comes at the end of the ‘long nineteenth century’, which was characterized by institutional transformation, but slow economic growth and weak industrialization. The second benchmark year— 1925— marks the end of the buoyant first half of the 1920s, when several industrial sectors reaped the benefits of tariff protection. The next— 1937— captures the impact on the top of the business structure of the domestic monetary and fiscal stabilization policies of the late 1920s, which was followed by the Great Depression. By 1957, another benchmark year, Portugal was entering a period during which modern economic growth emerged following two decades of a corporatist environment. The year 1973 represents the summit of Portuguese economic growth, and we analyze the corporate network just before the 1974 revolution and the subsequent institutional shock suffered by the business structure because of the nationalizations of 1975. A snapshot of the corporate network after the impact of nationalizations is provided in 1983. The next benchmark year, 1997, captures the corporate network during the privatization process and 10 years after the entry into the European Economic Community. Finally, in the last year— 2010— we look at Portugal’s current corporate structure. The data on big linkers and central firms, that are part of this project, are not part of this data package.

The dataset contains data on the composition of the boards of directors (BoD) of the 250 largest UK companies by net assets for nine benchmark years between 1904 and 2010 (1904, 1938, 1958, 1976, 1983, 1993, 1997, 2003 and 2010). The sample is composed of 50 financial firms and 200 non-financial firms, categorized according to the dominant sectoral activity of each firm. In order to compile this data, we relied partly on existing datasets, 2 as well as various archival sources, including stock exchange yearbooks, lists of largest companies provided by other scholars, commercial providers and annual reports. For certain years, it was not possible to locate sources that would have allowed us to compile a full sample of 250 firms. As a result, we only have 187 firms in 1958, 218 in 1983 and 247 in 2010. For 1993, we have 251 companies; for all other years, 250. By and large, however, the samples are of comparable size across our nine census dates. We have used the Pajek software package in order to compute a series of indicators of network structure and coherence, as well as actor centrality measures.

According to the general principles defined by Thomas David and Gerharda Westerhuis in order to insure comparability between countries, we have selected the ca. 50 biggest French financial and the ca. 200 biggest French non-financial firms for each benchmark date. Due to constraints in our sources, we however departed from studies on other countries by always choosing share capital as our sampling criterion. Other minor peculiarities of our sampling scheme can be found in the Sources and Methods file. The total number of sample firms is as follows: 1911: 245 1928: 236 1937: 241 1956: 255 1979: 247 1990: 252 2000: 250 While the exact names of positions varied a lot between firms, especially before World War II, we generally included all board members (members of the conseil d’administration, that had a quite flexible role, more or less executive or control-like, depending on firms), plus the two or three top executives, even when the source did not clearly include them in the board. We systematically gathered and coded data on many attributes of firms. Those that we used in our chapter attributes can be found in the Sources and Methods file. Some first results can also be found in the Sources and Methods file.

The source we used in this work for the benchmark years from 1913 to 1983 is Notizie statistiche sulle principali società italiane per azioni, edited by the Associazione fra le Società Italiane per Azioni (Assonime). The Imita.db database is an electronic version of this source. This dataset contains information on companies, boards of directors, and balance sheets of a large sample of Italian joint-stock companies for several benchmark years. On the whole, the dataset contains data on more than 38,000 companies, almost 300,000 directors, and more than 100,000 balance sheets. Representativeness, in terms of capital, is very high as the sample covers well over 90 percent of the total universe in all but the first two benchmark years (1911 and 1913) and the last one (1983), for which the proportion is around 85 percent. For the benchmark year 2001 we selected the top 250 companies from Le principali società italiane, the annual report on Italian joint-stock companies edited by R&S-Mediobanca. As this source does not report the names of the board members, we extracted them from Infocamere, a large dataset of Unioncamere, the association of the Italian chambers of commerce. This paper focuses on seven benchmark years: 1913, 1927, 1936, 1960, 1972, 1983, and 2001. In compliance with the White Paper rules, for each benchmark year we have selected the top 250 companies by total assets, with the exclusion of subsidiaries. The top 250 companies have been selected according to the following repartition: 50 financials and 200 non financials. As for the directors, we used only data for members of a board of directors in the strict sense, leaving out the members of Collegi sindacali. We have carefully standardized the names of the directors to make them as homogeneous as possible. However, we estimate that the information on boards of directors contained in Imita.db has a margin of error of about one percent, as is the case with other similar databases (Mintz and Schwartz 1985). These errors are mainly due to cases of homonymy, misprints, or shortcomings in the source. Data on board members of the top 250 Italian joint-stock companies for the year 1991 are not available so we had to drop this benchmark.

To create this dataset we collected the directorship data of the 200 largest nonfinancial firms and the 50 largest financial firms, in terms of total assets, for the years 1911, 1928, 1937, 1957, 1973, 1982, 1992, 1998 and 2009. In order to obtain the asset data and information on directors and auditors for pre-war Japan, we used a series of company directories compiled by a major securities company, Osakaya Shoten (Kabushiki Nenkan ). Because this source focuses on those companies with shares actively traded on the stock market, we supplemented the data for companies with closed ownership, such as the companies affiliated to major zaibatsu, with other sources, including Ginko Kaisha Yoroku by Tokyo Koshinjo, a credit bureau. With respect to the post-war period, the principal data source is a series of handbooks on the so-called keiretsu business groups by Toyo Keizai Shinposha (Kigyo Keiretsu Soran). As this source only covers the period from 1972 to 1999, we obtained the data for 1957 from Jojogaisha Soran by the Tokyo Stock Exchange, while the data for 2009 were collected from the annual reports of individual companies. Using the dataset constructed from those sources, we identify the 200 largest non-financial firms and the 50 largest financial firms, in terms of total assets, for the aforementioned benchmark years. Then, for these firms, we construct a database of directors and auditors in order to explore the features and evolutions of Japanese corporate networks.

Our dataset integrates several of the leading research projects on corporate networks in the postwar United States. The data from the 1960s were derived from the Mathematical Analysis of Corporate Networks (MACNET) project, directed by Michael Schwartz (see Mintz and Schwartz, 1985, for a description). Data from the 1970s were drawn from a project directed by Gerald Davis, Eric Neuman, and Mark Mizruchi. The data from the 1980s through 2003 were collected by Gerald Davis (see Davis, Yoo, and Baker, 2003). And the data from 2005 and 2010 were collected by Johan Chu (see Chu, 2012). Although we drew on several data sets, the directors of these projects used similar data collection strategies. All of the efforts were designed to document the networks of leading American corporations using the best available information sources for the time period studied. In each case, the researchers began by identifying the set of leading corporations, usually by means of the lists in Fortune, although Chu used the largest firms ranked by Standard & Poor. For the networks prior to 1990, the researchers identified director names and affiliations primarily from Standard & Poor’s Register of Corporations, Directors, and Executives and Moody’s Manuals, with supplemental reference to Dun and Bradstreet’s Million Dollar Directory, Directory of Directors of New York, and records of the Securities and Exchange Commission (SEC). From 1990 to 2003, the primary data source is SEC data on board composition distributed through Compact Disclosure and Global Access. For 2005 and 2010, the information portals were RiskMetrics and BoardEx. For all of the data, the researchers matched names across boards, and then checked to ensure that the identical names across firms actually represented the same individuals. Further details are available in the references cited above. After piecing together these datasets, we then selected subsets of the data files to conform to the requirements for this project. In particular, we selected the largest public firms by assets for each time point according to the quota scheme for large countries: 150 manufacturers, 50 financial institutions, and 50 other firms (utilities, mining, transportation, and service). All of the datasets on which we drew for our analysis collected data on more firms in each category than was necessary for our purposes. This ensures that we do indeed have the largest public firms in each subgroup for each year.

The data is collected by including all major active Taiwanese firms in pre-war colonial period, and all listed companies in post-war period for totally eight time points since 1941 to 2003 (i.e., 1941, 1962, 1969, 1982, 1988, 1992, 1998, and 2003). Data from 1941 is based on the Taiwanese Company Yearbook published by the Taiwan Economic Research Association. The Yearbook includes the major 168 companies that were active in Taiwan during 1941 when the Taiwan was still under the rule of Japanese colonial government. The list encompasses firms with a head office in Taiwan but also those headquartered in Japan. Data from the post WWII era includes all listed companies in Taiwan Stock Exchange Corporation (TSEC). As the development of stock exchange market in Taiwan was limited in the early period, the numbers of our sample are very small, and the numbers only begin to conform to a comparable benchmarked sample since 1990s. Prior to the 1980s, data is collected from the Listed Company Directory published by the Taiwan Stock Exchange Corporation. For the years after 1980, private firm data is retrieved from the electronic database of listed companies compiled by the Taiwan Economic Journal. Data on state-owned listed companies for the same periods is also comprehensively covered, and was obtained through the Annual Report on State-owned Enterprise published by the Commission of National Corporations, a branch of the Ministry of Economic Affairs. For all years, statistics like the number of firms in the main component, the percentage of isolated firms, the average number of ties, and the percentage of multiple ties are calculated based on the entire network of firms and directors. The data on big linkers and central firms, that are part of this project, are not part of this data package.

To create this dataset we used a sample of the 100 largest Swiss companies for eight benchmark years in the twentieth and twenty-first centuries (1910, 1929, 1937, 1957, 1980, 1990, 2000 and 2010). We included the 25 most important firms in the financial sector (banks, insurance companies and financial companies) and the 75 largest industrial firms. Moreover, we derived the composition of the board of directors (BoD) and the main executive director(s) (1– 4 people, depending on the year and the company) from different publications such as stock exchange manuals, financial yearbooks, annual reports of the firms and enterprises’ monographs. The data on big linkers and central firms, that are part of this project, are not part of this data package.

This data set consists of samples of the largest German firms for the years 1896, 1914, 1928, 1933, 1938, 1992 and 2010 (for sources see the sources and methods file). Sample size for each year: 1896: N= 211 1914: N= 346 1928: N= 377 1933: N= 405 1938: N= 363 1992: N= 614 2010: N= 252 For the years 1914, 1928, 1933, 1938 and 1992 a standardized firm sample of size N=250 has been selected. Firms have been sorted according to the size of their share capital. 1896 has been chosen as the first sample year. 1896 is the first year for which complete data for the largest public German corporations are available (Publisher: Hoppenstedt). For the year 1896, it was not possible to collect a sample of 250 firms because of the small size of many German firms. The names and titles (education, nobility, etc.) of all members of the management and supervisory board are derived from various handbooks and other sources (see the sources and methods file). Also part of this data set are samples of the largest US-firms for the years 1900, 1914, 1928, 1938, 1992 (for sources see the sources and methods file). Sample size for each year: 1900: N= 250 1914: N= 242 1928: N= 369 1938: N= 408 1992: N= 500 For the years 1928, 1938 and 1992 a standardized firm sample of size N=250 has been selected. Firms have been sorted according to the size of their share capital. Information about the members of Board of Directors are derived from various sources (see the sources and methods file).

This dataset contains data on the top 10 central firms, the top 10 big linkers and the corresponding network indicators in Austria for the years 1937, 1949, 1958, 1967, 1983, 1992, 1999 and 2008. The data are part of the ´The power of corporate networks: A comparative and historical perspective' project, which can be found via the following link: http://cgeh.nl/power-corporate-networks-comparative-and-historical-perspective

This dataset covers 10 benchmark years in the period 1903– 2008. For each year, we picked the top 100 exchange-listed Dutch firms and the top 25 listed and non-listed banks (except for 1998: 22 banks, and 2003: 15 banks). The selection was made on the basis of total assets, except for the banks in the period up to 1958, in which the selection was based on the nominal value of equity capital. Dutch firms are characterized by a two-tier board system, which consists of a managing board and a supervisory board. The supervisory board involves non-executive directors. For each firm, we collected the names of the managing and supervisory directors in 10 benchmark years. We gathered interlock data using Van Oss Effectenboek for the period 1903– 1973, and annual reports for the period 1978– 2008. Information of the banks’ boards was selected by the Financieel Adresboek for the period 1903– 1958, and annual reports for the period 1963– 2008. The data on big linkers and central firms, that are part of this project, are not part of this data package.

We collected information on the CEOs and boards of directors of 100 of the largest companies in Finland for specific benchmark years: namely, 1912, 1927, 1938, 1974, 1984, 1994 and 2004. We only included the boards of directors, but not the (non-compulsory ) supervisory boards (hallintoneuvosto/förvaltningsråd ). We selected 100 companies from the manufacturing industry, financial industry, trade and services and the utilities. We collected information about managing directors/ CEOs and boards of directors for these companies. The data for the period before World War II were collected from archival material in the National Board of Patents and Registration and several printed publications (Hjerppe 1979; Banker och Aktiebolag i Finland). After World War II, the sample was based on the list of biggest firms in Finland that is published yearly in the economic weekly magazine, Talouselämä, which has collected and published lists of the biggest companies in Finland since the early 1970s. The information on board members and CEOs was collected from a variety of printed and primary sources. Unfortunately, in the benchmark years 1974, 1984 and 1994, some companies had been omitted because of missing information. The total dataset consisted of 329 firms. We also look at ‘big linkers’. We focus primarily on the number of positions a person holds. The data from each of the seven sample years were first coded into Pajek (software for network analysis) as a 2-mode network consisting of companies and their board members (De Nooy and Batagelj 2011). The 2-mode network was then transformed into two 1-mode networks, one for the companies and the other for the members of the board. In the 1-mode networks, the businesses are connected to each other by common board members, whereas the directors are connected by having a seat on a board or being a CEO in a specific company.