The COVID-19 pandemic struck societies directly and indirectly, not just challenging population health but disrupting many aspects of life. Different effects of the spreading virus—and the measures to fight it—are reported and discussed in different scientific fora, with hard-to-compare methods and metrics from different traditions. While the pandemic struck some groups more than others, it is difficult to assess the comprehensive impact on social inequalities. This paper gauges social inequalities using individual-level administrative data for Sweden’s entire population. We describe and analyze the relative risks for different social groups in four dimensions—gender, education, income, and world region of birth—to experience three types of COVID-19 incidence, as well as six additional negative life outcomes that reflect general health, access to medical care, and economic strain. During the pandemic, the overall population faced severe morbidity and mortality from COVID-19 and saw higher all-cause mortality, income losses and unemployment risks, as well as reduced access to medical care. These burdens fell more heavily on individuals with low income or education and on immigrants. Although these vulnerable groups experienced larger absolute risks of suffering the direct and indirect consequences of the pandemic, the relative risks in pandemic years (2020 and 2021) were conspicuously similar to those in prepandemic years (2016 to 2019)

The COVID-19 pandemic has demonstrated the importance of unbiased, real-time statistics of trends in disease events in order to achieve an effective response. Because of reporting delays, real-time statistics frequently underestimate the total number of infections, hospitalizations and deaths. When studied by event date, such delays also risk creating an illusion of a downward trend. Here, we describe a statistical methodology for predicting true daily quantities and their uncertainty, estimated using historical reporting delays. The methodology takes into account the observed distribution pattern of the lag. It is derived from the “removal method”—a well-established estimation framework in the field of ecology.

Family and social networks are widely believed to influence important life decisions, but causal identification of those effects is notoriously challenging. Using data from Chile, Croatia, Sweden, and the United States, we study within-family spillovers in college and major choice across a variety of national contexts. Exploiting college-specific admissions thresholds that directly affect older but not younger siblings’ college options, we show that in all four countries a meaningful portion of younger siblings follow their older sibling to the same college or college-major combination. Older siblings are followed regardless of whether their target and counterfactual options have large, small, or even negative differences in quality. Spillover effects disappear, however, if the older sibling drops out of college, suggesting that older siblings’ college experiences matter. That siblings influence important human capital investment decisions across such varied contexts suggests that our findings are not an artifact of particular institutional detail but a more generalizable description of human behavior. Causal links between the postsecondary paths of close peers may partly explain persistent college enrollment inequalities between social groups, and this suggests that interventions to improve college access may have multiplier effects.

We measure how accurately replication of experimental results can be predicted by black-box statistical models. With data from four large-scale replication projects in experimental psychology and economics, and techniques from machine learning, we train predictive models and study which variables drive predictable replication. The models predicts binary replication with a cross-validated accuracy rate of 70% (AUC of 0.77) and estimates of relative effect sizes with a Spearman ρ of 0.38. The accuracy level is similar to market-aggregated beliefs of peer scientists [1, 2]. The predictive power is validated in a pre-registered out of sample test of the outcome of [3], where 71% (AUC of 0.73) of replications are predicted correctly and effect size correlations amount to ρ = 0.25. Basic features such as the sample and effect sizes in original papers, and whether reported effects are single-variable main effects or two-variable interactions, are predictive of successful replication. The models presented in this paper are simple tools to produce cheap, prognostic replicability metrics. These models could be useful in institutionalizing the process of evaluation of new findings and guiding resources to those direct replications that are likely to be most informative.