Pepper D. Culpepper

Around the world, the growing power of giant technology firms has outpaced regulatory oversight, raising urgent questions about how public opinion can be mobilized in support of meaningful reform. When do broad coalitions emerge in support of stronger regulation? This study investigates corporate scandals as windows of opportunity—focusing events—that can galvanize public support for new regulations and amplify issue salience. To assess this possibility, we conducted a two-wave survey experiment that we fielded simultaneously in France, Germany, the United Kingdom, and the United States. After collecting pre-treatment data in a separate, dedicated survey wave, our subjects were later assigned to read media coverage describing the 2018 Facebook/Cambridge Analytica scandal. The results confirm our preregistered hypothesis: in all four countries, exposure to scandal coverage significantly increases support for technology regulation and elevates the issue’s salience. Anger mediates this shift, and increased issue salience correlates with a greater willingness to engage in political action around technology policy. At a moment of heightened concern about the concentration of economic and political power in a handful of mammoth technology firms, these findings illuminate the microfoundations of how focusing events can break political inertia and catalyze regulatory reform.

This article advances the literature on media effects by examining how contrasting partisan narratives influence support for regulation after a real-world corporate scandal. Using both multi-wave observational and randomized experimental data, we show that self-selected media exposure and experimentally assigned information shape public opinion in distinct ways. While scandals are narratives of regulatory failure, partisan media environments differently attribute blame for that failure. In two separate observational waves, only Democrats exposed to news about the FTX bankruptcy increased their support for crypto regulation. In the experiment, only Republicans shifted in favor of regulation. Research on media effects needs to take into account not only media content, but also the partisan information environments that expose citizens to that content.

With the availability of large volumes of electronic communications data and the increasing sophistication of computational techniques, the development of automated approaches for different kinds of framing analysis is an important goal of researchers. There is as yet no standard method for the “unsupervised” inductive identification of frames based upon the content of articles. Three groups of core approaches underlie a wide range of work in this area, and we compare three techniques based on these approaches against each other and against manual human analysts. The three techniques are a k-means clustering algorithm together with a sophisticated natural language processing (NLP)-based feature selection process; evolutionary factor analysis (EFA), a factor analysis approach; and the structural topic model (STM). We use two datasets – one very focused and one extremely broad – as examples of the kind of frame analysis problems readers may wish to attempt. Even in a highly targeted dataset, we find some distance between the frames generated by computational analysis and those manually produced by our human analysts. The details of each method have a substantial impact on frame quality and interpretation. We find that the STM approach is the most effective for our narrow-scope dataset, but that it returns definite topics, and not frames, when working on our very broad dataset. We also show that we can get results emphasizing different parts of the framing problem by combining parts of different methods as a multi-stage process, rather than viewing available methods as simple plug-and-play models.

Do narratives about the causes of inequality influence support for redistribution? Scholarship suggests that information about levels of inequality does not easily shift redistributive attitudes. We embed information about inequality within a commentary article depicting the economy as being rigged to advantage elites, a common populist narrative of both the left and right. Drawing on the media effects and political economy literature, we expect articles employing narratives that portray inequality as the consequence of systemic unfairness to increase demands for redistribution. We test this proposition via an online survey experiment with 7426 respondents in Australia, France, Germany, Switzerland, the United Kingdom, and the United States. Our narrative treatment significantly increases attitudes favoring redistribution in five of the countries. In the US the treatment has no effect. We consider several reasons for the non-result in the US – highlighting beliefs about government inefficiency – and conclude by discussing general implications of our findings.

In 2017 the Australian government appointed a Royal Commission of inquiry into malfeasance in the banking sector. This article reports findings from a 2018 survey on attitudes to financial regulation and a survey experiment testing different media treatments. Attitudes on financial regulation are distinct from left-right positions on redistributive issues; we find no significant relationship between partisan identification and preferences for financial regulation. In the experimental treatment, all three frames catalysed anger and disgust from readers. However, neither of the two strong partisan frames moved policy preferences. The non-partisan frame – which included messages associated with both left and right, and which linked both parties to systemic capture by the banks – was the only article that had any effect on policy preferences, but only with non-partisan identifiers. Our results suggest that persuasive frames focused on the capture of politics by banking interests can move opinions of swing voters on financial regulation.

The Dodd–Frank Act of 2010 is the most comprehensive reform of American finance since the Great Depression and an ideal case to study how public opinion can counter the political power of finance. This article shows how pivotal congressional hearings created a clear story line for American media, one built around the way in which the investment bank Goldman Sachs made money during the crisis. We demonstrate that Goldman and CEO Lloyd Blankfein became the face of finance during these hearings. Results from a 2016 online survey experiment enable us to examine whether media portrayals highlighting the personal attributes of Blankfein and Goldman’s deal-making activate public opinion differently than articles foregrounding conflict of interest regulation and Goldman. Compared to the control condition, focus on Blankfein as the face of banking triggers negative affective response, greater appetite for regulating markets and greater attribution of blame toward banks for the financial crisis.

With the availability of large volumes of electronic communications data and the increasing sophistication of computational techniques, the development of automated approaches for different kinds of framing analysis is an important goal of researchers. There is as yet no standard method for the “unsupervised” inductive identification of frames based upon the content of articles. Three groups of core approaches underlie a wide range of work in this area, and we compare three techniques based on these approaches against each other and against manual human analysts. The three techniques are a k-means clustering algorithm together with a sophisticated natural language processing (NLP)-based feature selection process; evolutionary factor analysis (EFA), a factor analysis approach; and the structural topic model (STM). We use two datasets – one very focused and one extremely broad – as examples of the kind of frame analysis problems readers may wish to attempt. Even in a highly targeted dataset, we find some distance between the frames generated by computational analysis and those manually produced by our human analysts. The details of each method have a substantial impact on frame quality and interpretation. We find that the STM approach is the most effective for our narrow-scope dataset, but that it returns definite topics, and not frames, when working on our very broad dataset. We also show that we can get results emphasizing different parts of the framing problem by combining parts of different methods as a multi-stage process, rather than viewing available methods as simple plug-and-play models.

With the availability of large volumes of electronic communications data and the increasing sophistication of computational techniques, the development of automated approaches for different kinds of framing analysis is an important goal of researchers. There is as yet no standard method for the ‘unsupervised’ inductive identification of frames based upon the content of articles. Three groups of core approaches underlie a wide range of work in this area, and we compare three techniques based on these approaches against each other and against manual human analysts. The three techniques are a k-means clustering algorithm together with a sophisticated natural language processing (NLP)-based feature selection process; evolutionary factor analysis (EFA), a factor analysis approach; and the structural topic model (STM). We use two datasets – one very focused and one extremely broad – as examples of the kind of frame analysis problems readers may wish to attempt. Even in a highly targeted dataset, we find some distance between the frames generated by computational analysis and those manually produced by our human analysts. The details of each method have a substantial impact on frame quality and interpretation. We find that the STM approach is the most effective for our narrow-scope dataset, but that it returns definite topics, and not frames, when working on our very broad dataset. We also show that we can get results emphasizing different parts of the framing problem by combining parts of different methods as a multi-stage process, rather than viewing available methods as simple plug-and-play models.

We argue that the evolving preferences and power resources of large cross-border banks help explain the crucial political moves to European banking union. As they became larger and more European, these banks benefited from the supranationalization of supervision through reduced compliance costs and the effective opening of European markets. The political divergence in the interests of large international banks and small national ones eventually caused the German and the French governments’ change of position in intergovernmental bargaining. Once in place, banking union accelerated balance sheet consolidation to the benefit of large banks that took over their weaker competitors.