Data and Measures
Social Network Data
The data comprise a network of directors and corporations collected by Harrigan (2008). It is the network of the largest 248 corporations (as measured by revenue) in Australia in February 2006. It includes publicly listed and private corporations, as well as Australian and foreign- owned corporations. There are 1,251 directors, who hold a total of 1,464 directorships. Data on corporations and directors were obtained from IBISWorld (2006), the company that compiles the yearly Business Review Weekly’s “Top 2000 Enterprises.”[1]
Directors’ Social and Financial Capital
Actor-Relation Measures
Corporations have six binary attributes. Company size is a binary variable that divides the corporations into two equal groups: the largest 124 corporations and the smallest 124 corporations, as measured by revenue. Revenue data were also provided by IBISWorld. Corporations have two (mutually exclusive) binary donation variables: “donate to both major parties” (an indicator of moderate political activity) and “donate to conservatives only” (indicating donations to either the National Party or the Liberal Party - the two major conservative parties that are typically in coalition). Political donation data were downloaded from the Australian Electoral Commission (AEC; 2006) Web site. Companies that were publicly listed corporations were coded by the variable “public” (“1” indicates that the company was listed on the stock exchange). Australian ownership (=1) or foreign ownership (=0) was coded by the variable “Australian.” As an indicator of interest in and interaction with the political process of the state, firms were classified as to whether they were located in “highly regulated industries.” The regulated industries were coded using Burris’s (1987) classification and matched against the two-digit Standard Industry Codes in the IBISWorld data set. “Regulated industries” can be thought of as comprising four categories: transport (road, rail, and air), communication, utilities (water, gas, and electricity), and banking and insurance.[2]
Individual directors have three binary attributes: director wealth (or, more accurately, “superwealth” = 1), attendance at an exclusive private school (=1), and membership of an exclusive establishment gentlemen’s club (=1). Previous studies suggest that wealth, especially old wealth, may lead to political conservatism (Bond, 2003, 2004; Bond, Glouharova, & Harrigan, 2010; Burris, 2000). Directors are classified as “superwealthy” (=1) if they were listed in the Business Review Weekly’s (2005) “Rich 200” list. In Australia, data were obtained from the social directories “Who’s Who in Australia” and “Who’s Who in Australian Business” (Crown Content, 2005a, 2005b). The majority (59.1%) of the Australian directors had an entry in “Who’s Who.” This sample compares favorably to previous studies, for example, 33.7% of Useem’s (1984) UK sample and 30.3% of Bond’s (2007) UK sample were found in directories.
“School” was defined as attendance at one of 17 exclusive private schools.[3] This list was obtained by comparing the 3,000 secondary schools in Australia on a range of socioeconomic and status measures, including the socioeconomic status of parents, school fees, listing in “Who’s Who,” and the percentage of ex-students who were members of the exclusive businessmen’s clubs. In Australia, “clubs” were defined as a list of eleven prominent businessmen’s clubs.4 This list was identified through the use of secondary sources such as studies of upper-class clubs, reciprocal membership arrangements, and membership procedures.
Summary statistics for the attributes and derived attribute interaction effects are provided in Table 20.1. All variables are binary. We provide a count of corporations/directors/directorships with each attribute (column 1). To provide a baseline value (expected value) for interaction effects, we calculate the number of directorships that would occur if directorships arose from the random assortment of ties between corporations and directors. The second column (percentage of total corpo- rations/directors/expected directorships) divides the first column by the total number of corporations (248), directors (1,251), or directorships (1,464), respectively. Column 3 is a count of the total observed directorships for this attribute-attribute interaction. Column 4 is column 3 divided by the total number of directorships (1,464). Column 5 “overrepresentation” is percent directorships (column 4) divided by column 2: a value greater than one indicates that directors/corporations/director- corporation interactions have more directorships than would be expected given a random assortment of directorships. Column 6 indicates whether the over- or underrepresentation is statistically significant (using a chi- square test on the 2 x 2 matrix of the expected and realized number of directorships).
- [1] The authors want to express their gratitude to IBISWorld and Crown Content for their provision, respectively, of the “Top 2000 Enterprises” and the “Who’s Who” databases.
- [2] Industries classified as highly regulated were air and space transport, communicationservices, electricity and gas supply, finance, insurance, other transport, rail transport,road transport, services to finance and insurance, services to transport, water supply,sewerage and drainage services, and water transport.
- [3] These exclusive private schools include Anglican Church Grammar School (QLD), Brisbane Boys College (QLD), Brisbane Grammar School (QLD), Geelong Grammar School(VIC), Melbourne Grammar School (VIC), Scotch College (VIC), Wesley College (VIC),
