Suchergebnisse

AbstractPrevious studies have shown that relationship sentiments in families follow a pattern wherein either all maintain positive relationships or there are two antagonistic factions. This result is consistent with the network theory of structural balance that individuals befriend their friends' friend and become enemies with their friends' enemies. Fault lines in families would then endogenously emerge through the same kinds of interactional processes that organize nations into axis and allies. We argue that observed patterns may instead exogenously come about as the result of personal characteristics or homophilous partitions of family members. Disentangling these alternate theoretical possibilities requires longitudinal data. The present study tracks the sentiment dynamics of 1,710 families in a longitudinal panel study. Results show the same static patterns suggestive of balancing processes identified in earlier research, yet dynamic analysis reveals that conflict in families is not generated or resolved in accordance with balance theory.

Abstract
Past quantitative studies have shown that most media coverage is of men. Here we ask if the scarce coverage that women get is qualitatively different from that of men. We use computer-coded sentiment scores for 14 million person names covered in 1,323 newspapers to investigate the three-way relationship between gender, fame, and sentiment. Additional large-scale data on occupational categories allow us to compare women and men within the same profession and rank. We propose that as women's fame increases their media coverage becomes negative more quickly when compared to men (a "paper cut"), because their violation of gender hierarchies and social expectations about typical feminine behavior evokes disproportionate scrutiny. We find that while overall media coverage is much more positive for women than for men, this difference disappears and even reverses at higher levels of fame. In encyclopedic sentiment data we find no biographic basis for women's disproportionate decline in media coverage sentiment at high fame, consistent with the conjectured double standard in media discourse.

AbstractReputation systems are commonplace in online markets, such as on peer-to-peer sharing platforms. These systems have been argued to be a solution to (ethnic) discrimination on such platforms. This argument is based on empirical studies showing that ethnic disadvantages are smaller for users with ratings than for users without ratings. We argue that this conclusion may be premature, because minorities have a harder time accumulating ratings. The greater benefit of ratings to minorities may be offset by their troubles acquiring any, thereby diminishing the potential for reputation systems to reduce discrimination. We tested this counterargument using a unique data set that contains information on all interactions on a peer-to-peer motorcycle rental platform. We find that the reputation system does not reduce initial inequalities between otherwise comparable renters of different ethnicity. Platforms that wish to reduce discrimination should not only make their reputation system more effective, but also help users collect ratings.

The traditional understanding of reputation systems is that they secure trust between strangers by publicly calling out cheaters. In modern, online markets, it is increasingly common for providers of a good to also act as consumers, and vice versa. We argue that in such mixed-role markets, reputation systems serve a second important function: They allow providers who lend out their possessions (such as their house, car or tools) to earn reputational credits that can be spent on future borrowing, especially when lending lacks monetary compensation. In an experiment that introduces a new game, "the Lending Game", we show that, consistent with our argument, information on past lending leads subjects to lend to those who have themselves lent before, increasing overall lending. However, when lending is financially compensated, this mechanism of reciprocal lending ceases to operate.

We examine an increasingly common political tactic: the self-binding pledge whereby lawmakers assure a constituency, social movement, or interest group that they will take a particular position on an upcoming vote or policy issue. An empirical analysis of novel data on the 2011 debt-limit crisis shows that pre-commitment by legislators to the Tea Party's "Cut, Cap, and Balance Pledge" was the joint product of grassroots pressure and ideological sympathy. Highly conser-vative early pledgers opposed fiscal compromise that fell short of the Tea Party's position, while a number of less conservative later signers reneged by voting for the Budget Control Act that resolved the debt-limit crisis. The case of the pledge sug-gests that social movements are well positioned to collaborate with true believing insiders to promote policy change but have difficulty persuading fence sitters to take their side.