Party System Institutionalization and Stability in Competitive Authoritarian Regimes

ICYMI (In Case You Missed It), the following work was presented at the 2020 Annual Meeting of the American Political Science Association (APSA). The presentation, titled “Electoral Volatility in Competitive Authoritarian Regimes” was a part of the session “Elections Under Autocracy” on Sunday, September 13, 2020.

Until recently, there has been little need to measure the electoral volatility, changes in vote shares between parties, in authoritarian regimes because most conventional authoritarian regimes were either one-party or no-party systems. In general, high levels of volatility are considered to be a sign of instability in the party system and show that the existing parties are unable to build connections with their constituencies. 

New research by Wooseok Kim, Allen Hicken, and Michael Bernhard examines the ways that electoral volatility in democratic regimes may be useful for understanding competitive authoritarian regimes. 

As a greater number of authoritarian regimes have permitted electoral competition and greater party autonomy, electoral volatility has become more salient. Multiparty elections in competitive authoritarian regimes are different from those in democracies, in that competition is more constrained and incumbents have the ability to manipulate the outcomes. 

Electoral volatility can provide clues about the level institutionalization in the ruling and opposition parties, as well as the level of support for the authoritarian incumbent. Low volatility suggests a high level of stability and control in the ruling party institutionalization; high volatility as associated with weak party organizations, weak societal roots, and low levels of cohesion. 

The authors tested the relationship between electoral volatility, which is the most commonly used measure of party system institutionalization, and the survival of competitive authoritarian regimes. To do this, they used a dataset that included authoritarian regimes in the post-WWII period that hold minimally competitive multiparty elections with basic suffrage, which are determined using indicators from the Varieties of Democracy (V-Dem).


Specifically, the authors measure two types of electoral volatility in competitive authoritarian regimes: type-A volatility and type-B volatility. Type-A is volatility measures the exit and entries of parties from the system. Type-B volatility measures the reallocation of votes or seats from one party to a competitor.

Type-A is volatility measures the exit and entries of parties from the system. Type-B volatility measures the reallocation of votes or seats from one party to a competitor.

Electoral authoritarian regimes are more stable when they tightly control the party system and the opposition is disorganized. The authors conclude that type-B volatility promotes authoritarian replacement, while type-A volatility is associated with a greater likelihood of a democratic transition. In addition to considering measures of party system institutionalization in authoritarian regimes, future case studies may shed more light on the link between electoral dynamics and outcomes.

 

The Role of Economic Decline & Malaise in the Rise of Extreme-Nationalist Populism 

ICYMI (In Case You Missed It), the following work was presented at the 2020 Annual Meeting of the American Political Science Association (APSA). The presentation, titled “Local Economic Malaise and the Rise of Anti-Everything Extremism” was a part of the session “Extreme Parties and Positions” on Saturday, September 12, 2020. Post developed by Hayden Jackson and Katherine Pearson. 

Is it economic downturns or threats to cultural identity that lead some individuals to respond to populist and extreme-nationalist appeals? These explanations complement, rather than compete with one another, according to new research by Diogo Ferrari, Rob Franzese, Hayden Jackson, ByungKoo Kim, Wooseok Kim, and Patrick Wu

Some people experiencing a decline in standard of living may react by supporting populist movements, including those that place blame for economic and social deterioration on out-groups. While economic downturns can spur support for nationalism, these factors are also deeply entwined with feelings of being left behind in a social and cultural context. This is especially true in hard-hit rural communities that feel neglected and misunderstood by policy makers and elites. 

Whereas previous literature presents economic malaise and cultural or status threat as competing explanations for the rise of populist attitudes, the authors of this paper argue that these effects are not competing, but complementary. When the community experiences economic decline, some individuals will feel that their identity is under threat, and that they are looked down upon by elites. The feeling that their way of life is under attack leaves some individuals susceptible to extremist appeals. However, these appeals do not work on all members of the community equally; important differences may be explained by life experiences, education, personal income, and demographics, especially race. 

One’s views and behaviors grow as a result of complex economic and cultural experiences. Some people will have experiences or personalities that predispose them to respond differently to economic and social shocks. For some, economic decline may trigger xenophobic, anti-elite reactions that will not be experienced by all members of the community. 

To test the relationship between economic malaise and the perception of social threat, the authors conducted two empirical explorations. The first study reanalyzed data from Mutz 2018 to identify the effects of features like neighborhood decline or individual characteristics in subgroups with different responses to economic decline. 

A second study focuses on structural differences that appear in data from Twitter data before and after automotive plant shutdowns in southeast Michigan and northeast Ohio. Data suggests that neighborhood economic shocks, like the closing of a factory, triggered rising extremist expression in at least some contexts. The increase in extremist-engaging Tweet activity was largest in the community around Lordstown, Ohio, which is predominantly white and rural/exurban. By comparison, the data showed slightly negative trends in extremist Tweets in the predominantly Black, urban community around Hamtramck, Michigan, which was also hit by a plant closing. 

The authors hypothesize that the response to an economic shock, such as a plant closing, is likely to depend on the size of the closure “shock”, or how much impact it has and on the community, as well as the social and demographic characteristics of the local workers, particularly the community’s urban or rural nature and the racial makeup. In the analysis, these factors were most relevant when determining an extremist-engagement response. The bigger the economic shock to the community, and the more white and rural the community, the more likely it is to see an extreme response. 

Joint Image-Text Representations Using Deep Learning 

ICYMI (In Case You Missed It), the following work was presented at the 2020 Annual Meeting of the American Political Science Association (APSA). The presentation, titled “Joint Image-Text Classification Using an Attention-Based LSTM Architecture” was a part of the session “Image Processing for Political Research” on Thursday, September 10, 2020Post developed by Patrick Wu and Katherine Pearson. 

Political science has been enriched by the use of social media data. However, automated text-based classification systems often do not capture image content. Since images provide rich context and information in many tweets, these classifiers do not capture the full meaning of the tweet. In a new paper presented at the 2020 Annual Meeting of the American Political Science Association (APSA), Patrick Wu, Alejandro Pineda, and Walter Mebane propose a new approach for analyzing Twitter data using a joint image-text classifier. 

Human coders of social media data are able to observe both the text of a tweet and an attached image to determine the full meaning of an election incident being described. For example, the authors show the image and tweet below. 

Photo of people waiting to vote and text of tweet reading “Early voting lines in Palm Beach County, Florida #iReport #vote #Florida @CNN”

If only the text is considered, “Early voting lines in Palm Beach County, Florida #iReport #vote #Florida @CNN”, a reader would not be able to tell that the line was long. Conversely, if the image is considered separately from the text, the viewer would not know that it pictured a polling place. It’s only when the text and image are combined that the message becomes clear. 

MARMOT

A new framework called Multimodal Representations Using Modality Translation (MARMOT) is designed to improve data labeling for research on social media content. MARMOT uses modality translation to generate captions of the images in the data, then uses a model to learn the patterns between the text features, the image caption features, and the image features. This is an important methodological contribution because modality translation replaces more resource-intensive processes and allows the model to learn directly from the data, rather than on a separate dataset. MARMOT is also able to process observations that are missing either images or text. 

Applications

MARMOT was applied to two datasets. The first dataset contained tweets reporting election incidents during the 2016 U.S. general election, originally published in “Observing Election Incidents in the United States via Twitter: Does Who Observes Matter?” The tweets in this dataset report some kind of election incident. All of the tweets contain text, and about a third of them contain images. MARMOT performed better at classifying the tweets than the text-only classifier used in the original study. 

In order to test MARMOT against a dataset containing images for every observation, the authors used the Hateful Memes dataset released by Facebook to assess whether a meme is hateful or not. In this case, a multimodal model is useful because it is possible for neither the text nor the image to be hateful, but the combination of the two may create a hateful message. In this application, MARMOT outperformed other multimodal classifiers in terms of accuracy. 

Future Directions 

As more and more political scientists use data from social media in their research, classifiers will have to become more sophisticated to capture all of the nuance and meaning that can be packed into small parcels of text and images. The authors plan to continue refining MARMOT, and expand the models to accommodate additional elements such as video, geographical information, and time of posting.