Publication Type: Paper/Poster Proposal
Abstract: Area probability surveys with face-to-face interviewing usually involve a cluster sampling approach. Cluster sampling entails selecting small areas (we call these “segments”) in which every housing unit is enumerated. The lists of enumerated units are then subsampled for inclusion into the sample. People who live in these clusters of sampled housing units often are similar on some of the outcome variables, especially financial variables. The similarity of the sample clusters is measured by the intraclass correlation, which is usually positive. Positive intraclass correlations indicate that the sample is not quite as good as a simple random sample of the same sample size. This inefficiency is measured by the design effect (DEFF), and results in a smaller effective sample size (n/DEFF).

Larger segments would tend to result in smaller intraclass correlations (e.g., see Hansen, Hurwitz, and Madow, 1953), but would also result in higher costs (to enumerate every housing unit). Most area probability studies make a choice on segment size, but then cannot examine the effect of different segment sizes on the intraclass correlations. Our talk will focus on data from the Making Connections project, sponsored by the Casey Foundation, which collects data from entire neighborhoods. With this data, we compiled a frame of all addresses in the neighborhood and drew an element sample. We are now able to divide neighborhoods into segments of varying sizes to study the impact on the intraclass correlations for various response variables. This research can help choose appropriate segment sizes in future surveys.

Publication Type: Conference Paper/Unpublished Manuscript
Abstract: The rise of the contingent workforce represents one of the most significant work trends in recent years. Since the 1970s, workers have seen their job security begin to erode as employers rely more and more on contingent or insecure employment arrangements in order to cut labor costs and increase flexibility in the face of increasing global competition. While researchers have examined the individual-level causes and consequences of contingent work, empirical analyses of structural factors such as occupations remains largely absent from the literature. In this paper, I examine how occupations shape workers’ wages and in particular how they influence the wage gap between contingent and non-contingent (i.e. “secure”) workers in the U.S. Using individual-level data from the CPS Contingent Work Supplements and occupation-level data from a variety of sources, I estimate multilevel wage equations in order to test several competing explanations. I find that while social closure strategies affect overall earnings, they have little or no effect on the contingent wage gap. However, race and sex segregation and wage flexibility do have significant impacts on the wage gap. Finally, while economic segmentation by itself is important, its effect is negligible once one takes into account other occupational factors.

Publication Type: Conference Paper/Unpublished Manuscript
Review Method: Peer Reviewed
Abstract: This paper explores how lack of privacy rights, in the era of increased surveillance and segmentation, has detrimental effects on the First Amendment Rights of individuals and on the well-functioning of democracy. In the absence of privacy rights, corporations’ ability to segment citizens means that individuals will increasingly be exposed to information that is designed to manipulate, rather than inform them. Corporations’ ability to segment citizens also implies that certain segments of the population will be excluded from the democratic deliberation process because they are deemed as politically unviable.

Publication Type: Conference Paper/Unpublished Manuscript
Abstract: The existence of a European public sphere, a public network of exchange of opinions and ideas on political issues, has come to be seen as a prerequisite of democratizing the European Union. This paper conceptualizes the desired Europeanization of the national public spheres as a gradual process that may occur on three different dimensions: 1. segmented Europeanization, 2. discursive Europeanization, and 3. collective identification with Europe. It then presents the results of our empirical research on the transnationalization of public spheres in Europe: What is the prevailing pattern of Europeanization that can be observed in different countries of the EU?
We have conducted a quantitative content analysis of the political discourses in quality newspapers of five EU member states (Austria, Denmark, France, Germany, Great Britain) over a period of twenty years. According to the analysis of more than 3000 articles the main pattern of transnationalization to be found in all countries is segmented Europeanization: Within each public sphere we find more and more talk about European institutions and policies but there is no indication of an increase in the debate in between the national public spheres. This notwithstanding, we find that the level of discursive interaction transcending national borders has already been considerably high in the beginning of our analysis in 1982. In addition, we find indications of a gradually developing a European “we”-perspective.

Publication Type: Conference Paper/Unpublished Manuscript
Abstract: The patterns for global governance of electronic networks, specifically the Internet, have become a topic of increasing debate over the last years. In addition to governmental and civil society organizations, private corporations seek strengthened roles in Internet governance. A number of inter-governmental organizations, as well as in non-government organizations and states are now debating goals and forms of Internet governance.

Internet governance may be approached most usefully by distinguishing among three distinct segments of Internet service provision, and the governance goals, bodies, and mechanisms most appropriate and relevant to each. This allows different problems to be identified, different actors and institutions to be mobilized, and appropriate policy mechanisms to be debated and designed for each. By separating these segments, policy concerns may be addressed more concretely, within a feasible scope, and constructive solutions unique to different sectoral activities identified and implemented.