Logo: to the web site of Uppsala University

List experiments encourage survey respondents to report sensitive opinions they may prefer not to reveal. But, studies sometimes find that respondents admit more readily to sensitive opinions when asked directly. Often this over-reporting is viewed as a design failure, attributable to inattentiveness or other nonstrategic error. This paper conducts an exploratory analysis of such a ‘failed’ list experiment measuring vote buying in the 2019 Nigerian presidential election. We take this opportunity to explore our assumptions about vote buying. Although vote buying is illegal and stigmatized in many countries, a significant literature links such exchanges to patron-client networks that are imbued with trust, reciprocity and long-standing benefits, which might create incentives for individuals to claim having been offered to participate in vote buying. Submitting our data to a series of tests of design, we find that over-reporting is strategic: respondents intentionally reveal vote buying and it’s likely that those who reveal vote buying have in fact being offered to participate in vote buying. Considering reasons for over-reporting such as social desirability and network benefits, and the strategic nature of over-reporting, we suggest that “design failure" is not the only possible conclusion from unexpected list experiment results. With this paper we show that our theoretical assumptions about sensitivity bias affect the conclusions we can draw from a list experiment.

Processes that occasionally, but not always, produce extreme values are notoriously difficult to model, as a small number of extreme observations may have a large impact on the results. Existing methods for handling extreme values are often arbitrary and leave researchers without guidance regarding this problem. In this paper, we propose an extreme value and zero-inflated negative binomial (EVZINB) regression model, which allows for separate modeling of extreme and nonextreme observations to solve this problem. The EVZINB model offers an elegant solution to modeling data with extreme values and allows researchers to draw additional inferences about both extreme and nonextreme observations. We illustrate the usefulness of the EVZINB model by replicating a study on the effects of the deployment of UN peacekeepers on one-sided violence against civilians.

This contribution to the ViEWS prediction competition 2020 proposes using Markov modeling to model the change in the logarithm of battle-related deaths between two points in time in a country. The predictions are made using two ensembles of observed and hidden Markov models, where the covariate sets for the ensembles are drawn from the ViEWS country month constituent models. The weights for the individual models in the ensembles were obtained using a genetic algorithm optimizing the fit on the TADDA-score in a calibration set. The weighted ensembles of visible and hidden Markov models outperform the ViEWS prediction competition benchmark models on the TADDA score in the test period of January 2017 to December 2019 for all time steps. Forecasts until March 2021 predict increased violence primarily in Algeria, Libya, Tchad, Niger, and Angola, and decreased or unchanged levels of violence in most of the remaining countries in Africa. An analysis of the model weights in the ensembles shows that the conflict history constituent model provided by ViEWS was dominant in the ensembles.

This article presents an update to the ViEWS political Violence Early-Warning System. This update introduces (1) a new infrastructure for training, evaluating, and weighting models that allows us to more optimally combine constituent models into ensembles, and (2) a number of new forecasting models that contribute to improve overall performance, in particular with respect to effectively classifying high- and low-risk cases. Our improved evaluation procedures allow us to develop models that specialize in either the immediate or the more distant future. We also present a formal, 'retrospective' evaluation of how well ViEWS has done since we started publishing our forecasts from July 2018 up to December 2019. Our metrics show that ViEWS is performing well when compared to previous out-of-sample forecasts for the 2015-17 period. Finally, we present our new forecasts for the January 2020-December 2022 period. We continue to predict a near-constant situation of conflict in Nigeria, Somalia, and DRC, but see some signs of decreased risk in Cameroon and Mozambique.

This article presents ViEWS – a political violence early-warning system that seeks to be maximally transparent, publicly available, and have uniform coverage, and sketches the methodological innovations required to achieve these objectives. ViEWS produces monthly forecasts at the country and subnational level for 36 months into the future and all three UCDP types of organized violence: state-based conflict, non-state conflict, and one-sided violence in Africa. The article presents the methodology and data behind these forecasts, evaluates their predictive performance, provides selected forecasts for October 2018 through October 2021, and indicates future extensions. ViEWS is built as an ensemble of constituent models designed to optimize its predictions. Each of these represents a theme that the conflict research literature suggests is relevant, or implements a specific statistical/machine-learning approach. Current forecasts indicate a persistence of conflict in regions in Africa with a recent history of political violence but also alert to new conflicts such as in Southern Cameroon and Northern Mozambique. The subsequent evaluation additionally shows that ViEWS is able to accurately capture the long-term behavior of established political violence, as well as diffusion processes such as the spread of violence in Cameroon. The performance demonstrated here indicates that ViEWS can be a useful complement to non-public conflict-warning systems, and also serves as a reference against which future improvements can be evaluated.

This article uses a large-ndataset to investigate the effect of terroristattacks with American victims on the popularity of the U.S. president.The study uses two broad theoretical frameworks to analyze thiseffect, the score-keeping framework and the rally-effect framework.Thefindings of the study show that, when excluding the effect fromthe September 11, 2001 terrorist attacks, actual terrorist attacks haveno generalizable short-term impact on the popularity of the U.S.president. This indicates that even though the topics of nationalsecurity, terrorism, and the president’s ability to handle these issuesare important in the political debate in the United States, actualterrorism has little or no short-term impact on presidential approvalratings.