Digital Politics in Western Democracies: A Comparative Study

Notes

CHAPTER 1: Introduction 1. Gerald Sussman (2005) has suggested that Americanization of contemporary Western politics may be understood as the result of imposition not by the United States but by transnational financial capital. From this perspective, the development of highly professionalized, capital-intensive, television-centered electioneering has allowed global economic elites to better control the public sphere through the financing of political speech in campaigns. As a result, neoliberal ideas have become dominant because politicians cannot run counter to the interests that feed their electoral coffers, thus relieving democracy from excessive popular demands. This argument, however, was developed with respect to broadcast rather than digital media, and it has not yet been reassessed in light of the changes brought about by the internet. It is, nonetheless, a thought-provoking perspective that should not be overlooked by scholars in the field. 2. Norris (2000: 149–159), for instance, suggests that the developments in campaigning that she describes depend on a set of mediating conditions: the regulatory environment, the party system, the media system, and the electorate. However, she does not go beyond stating these principles and does not apply these considerations to her empirical analysis of changes in electioneering methods. 3. The goal of relationship marketing is defined as ensuring that “organizational objectives are achieved by mutual exchange and fulfillment of promises” (BowersBrown 2003: 100) between citizens and political actors. This approach contrasts with the temporary convergence of interests that transaction-based marketing focuses on and that, as applied to politics, corresponds to trying to affect short-term voting behavior only. CHAPTER 2: Understanding Digital Politics in Western Democracies 1. Data on internet diffusion were provided by the International Telecommunication Union, see http://www.itu.int/ITU-D/ict/statistics/index.html. 2. The data come from the 2010 Australian Election Study (Australia), various issues of the Eurobarometer (France, Germany, Italy, Spain, and the United Kingdom), and the 2006 World Values Survey (the United States). In the latter survey, the question asked was, “How democratically is this country being governed today?” 3. For instance, the German Pirate Party, which Schweitzer (2011) found to have a very robust online presence, and which has recently been quite successful in local elections, was not included in the analysis of the 2009 German elections. Notes 244 Notes to Pages 38–55 4. The weights were calculated based on the algorithm: wj = proportion in population in cell j/proportion in sample in cell j. First, country weights were calculated, then within-country weights were calculated, and finally country weights were multiplied by within-country weights to calculate the combined weight for each website. As an example, let us consider the case of the Australian Labor Party, which was analyzed three times. The total number of Australian websites that were analyzed was 46, so each Australian website had a .074 probability of being part of the total corpus. To equalize the weight of Australian websites to that of websites from other countries, I assumed that Australian websites were one-seventh of the websites in the population. Thus, the .14 (1 in 7) proportion in the population was divided by the .074 proportion in the sample, resulting in a 1.92 weight for all Australian websites. Country weight was then multiplied by an individual weight, which, for the Australian Labor Party, was calculated as follows. This website was analyzed three times out of 46 total observations of Australian websites, so its proportion in the sample was .06. Because 28 Australian parties and candidates were included in the analysis, the proportion of the ALP in the population was .03 (1 in 28). This proportion was then divided by the proportion in the sample (.06), resulting in a within-country weight of .54. Finally, I multiplied the country and within-country weights, so that the ALP had a combined weight of 1.92 * .54 = 1.05. 5. In discussing how to analyze comparative survey data, John Curtice (2007: 905– 6) suggests that sample sizes may be either weighted to be proportionate to countries’ populations or that each national sample can be considered “as a separate reading of the phenomenon,” which would imply that the weighting algorithm should “equalize the sample sizes for each country.” The latter is the choice that I have made here. 6. To ensure reliability in measurement, an instructed research assistant coded all the 31 websites included in the November and December 2007 US observations, which constitute 5% of the corpus. The results were compared...