Fixed Effects Regression Models
... To analyze the transition to first and later conceptions, we approximate a piecewise-constant event-history model paired with monthly climate data by following a discrete-time hazard analysis setup (Allison, 2009;Barber et al., 2000), where the unit of analysis is person/month. Instead of using a logit specification, however, we use linear probability modeling (LPM). ...
... An individual-specific intercept, µ i , represents the combined effect of all time-constant observed and unobserved variables on y. An idiosyncratic error term, ε it , represents random variation at each time point (Allison, 2009). X represents the matrix of variables. ...
... Second, we explored an alternative approach to fixed-effects modeling because using a discrete-time setup-in which the event is considered time-dependent and the covariates change over time-with LPM and including fixed effects is not standard in the literature. Following Allison's (2009) suggestion of modeling fixed effects in event-history analysis, we used a stratified Cox regression. In a stratified Cox model, the baseline hazard is estimated for each individual in the data, which accounts for unobserved heterogeneity in the model. ...
Research on how climate affects conceptions is limited and often constrained by data limitations. Additionally, scientific knowledge is typically based on research focused on very poor or wealthy settings. Here, we examine two middle-income and climate-sensitive contexts, Armenia and Tajikistan, and use finely detailed data on local weather conditions and all conceptions, including those not ending in a live birth. We fit fixed-effects linear probability models of the time until each conception using data from the Demographic and Health Surveys, stratified by educational level and urban/rural residence. No relationship appears between heat and conceptions in Armenia or between rainfall and conceptions in either context. In contrast, exposures to higher-than-usual temperatures and hot days are associated with a lower probability of conception in rural Tajikistan. This finding persists when examining all conceptions and only those resulting in a live birth. Further, the results do not vary by women’s educational attainment nor by being childless or not. As such, we do not find evidence that specific groups of women are more vulnerable to climate variability beyond those living in rural (versus urban) areas. Given the broad impact of heat on conceptions in rural Tajikistan, differences in how individuals engage with the urban and rural environments may be important in the short-term relationship between climate and conceptions.
... The list of control variables for main models (Models 1-6) is relatively short, as I employ firm-level and country-level fixed effects. Fixed-effects models exploit repeated observations within units (countries and firms) to remove time-invariant heterogeneity, including unobserved heterogeneity (Allison, 2009). In other words, we are comparing the same company in the same country before and after an enforcement action. ...
... In other words, we are comparing the same company in the same country before and after an enforcement action. Any cross-country or cross-firm variation, to the extent that it is stable over time, will be discarded by the model and is not estimated, leading to slightly less efficient but unbiased estimates (Allison, 2009). This is important because the number of tweets we observe may well be influenced by a wide range of other factors at the firm level (e.g. ...
... The second accounts for potential differences in public visibility and transparency of publicly-traded firms vis-à-vis private companies. As noted, the main models do not explicitly include these variables because the fixed effects for firm and country remove any time-invariant heterogeneity, including unobserved heterogeneity (Allison, 2009). ...
Purpose
This study examines the relationship between revelations of bribery by multinational enterprises (MNEs) in foreign markets and negative social media coverage of these firms. Drawing on neoinstitutional theory and social norms research, it predicts that negative social media reactions to corruption revelations will be moderated by host country corruption levels.
Design/methodology/approach
This study uses corruption data from 168 MNEs charged under the US Foreign Corrupt Practices Act and approximately 1 million tweets related to these firms' corrupt activities in over 80 countries.
Findings
We find significant negative social media reactions to corruption revelations in the host country, the sanctioning country and globally. Social media reactions were weaker when bribery took place in highly corrupt countries.
Originality/value
This study takes a novel approach to researching the potential risks of engaging in corrupt activities by leveraging multi-country social media data from Twitter/X.
... We discuss the assumptions undergirding traditional time-series cross-sectional models and introduce growth curve models as an extension of these traditional estimators. Although growth curve models are often presented as a form of structural equations (Chou, Bentler and Pentz 1998), they also represent an unconstrained version of hybrid (fixed-and-random effects) panel estimators (Allison 2009;Finkel 2020). Growth curve models capture the idea that initial starting conditions and cumulative conditions shape a regime's trajectory, not just its initial level of democracy and short-term fluctuations. ...
... 5 To address these limitations, other estimators combine cross-sectional and time-varying predictors. This family of techniques includes hierarchical random-effects models (Raudenbush and Bryk 2002), fixed-effects vector-decomposition models (Plümper and Troeger 2007), correlated random effects models (Mundlak 1978;Wooldridge, 2019) and hybrid effects models (Allison 2009;Finkel 2020). We focus on the latter, as the 'hybrid' estimator is an extension of fixed-effects models and a constrained version of growth curve models. ...
This paper advances the study of democratic trajectories – whether democracies deepen, stagnate, erode or break down over time. We show that econometric panel models usually neglect cumulative effects, which are implicitly central to many theories of democratic change. Some important factors, such as economic growth, have cumulative effects that shape medium- to long-term regime trajectories. To overcome the limitations of conventional statistical estimators, we propose the use of latent growth curve models, which are better able to capture cumulative processes. We demonstrate the advantages of this approach by analyzing the trajectory of 103 democratic regimes inaugurated after 1974. Conventional estimators fail to predict democratic trajectories, while latent growth curve models properly capture cumulative effects.
... The panel Poisson fixed-effects estimation provides a robust method for analyzing historical armed conflict data, particularly when many observations are zero 53 . During the Taiping Rebellion (1851-1864), many regions remained unaffected and experienced no armed conflicts, resulting in numerous zero values in the dataset. ...
... It accounts for overdispersion and effectively handles sparse data. Additionally, by incorporating fixed effects, this model controls for unobserved regional heterogeneity, ensuring that differences across regions (such as socio-economic and ecological factors) do not distort the analysis 53 . This is particularly crucial when studying a large geographic area where baseline conflict risks may differ but remain relatively stable within each region over time. ...
The Taiping Rebellion (1851–1864), the deadliest civil war in Chinese history, has been studied for its societal and climatic links, while its spatial dynamics remain underexplored. This study quantifies the spatial relationship between hydro-climatic extremes and armed conflict during the rebellion. Using panel Poisson fixed-effects estimation, we examine how droughts and floods influence the conflict occurrence both locally and in surrounding areas, applying buffer zones of 10 km, 25 km, 50 km, and 100 km. We cross-validate results using both Qing administrative boundaries and 100 km × 100 km grid-based spatial data. Our findings show that floods, more than droughts, significantly increase the likelihood of local armed conflicts, with the strongest correlation found within a 10 km radius. This spatial pattern may reflect the limited mobility of displaced populations in late Qing China. The study underscores the importance of spatial context in understanding climate-driven conflict risks in vulnerable regions.
... Other descriptive methods include for example group level mean comparisons in terms of the key dependent and independent variables examined with traditional box plot figures with t (t-test) and F (ANOVA) statistics (Articles I and III). The main analyses in each article rely, however, on OLS regression models with several adjustments, such as including country fixed effects to control for the country-specific contextual effects or to ensure the within-country variation of the independent variables (Articles I-III) (see also Allison, 2009), several interaction effects to address the links between the key independent variables (Articles I-II), and adding lags to the dependent and independent to take account for the autocorrelation issues in terms of the popularity ratings and economic variables (Articles I-II). Finally, several robustness tests are included in each article to ensure that the results are not violated for example by the choice of certain measures (e.g., specific presidential power scores or cohabitation instead of more general ideological (2018), Sedelius and Mashtaler (2013) and complemented with information from Raunio and Sedelius (2020) and the EJPR Political Data Yearbooks. ...
... All models include country dummies for fixed effects. Fixed effects regression model is used since it controls for the effect of groups and time (Allison, 2009;Gil-García and Puron-Cid, 2013). With this approach, country-specific contextual effects are controlled, and the results are not violated by varying survey questions in the five countries. ...
This doctoral dissertation focuses on public opinion towards presidents operating alongside prime ministers. The power balance between these two executives varies especially in semi-presidential regimes, where a popularly elected fixed-term president shares executive power with a prime minister and the government, who, in turn, are accountable to the parliament (Duverger, 1980; Elgie, 1999). In terms of public opinion, this dual executive structure should benefit the president over the prime minister, but only if the former is politically subordinate to the latter. This connects to questions of political accountability, i.e., on what basis people ‘reward’ and ‘sanction’ incumbent political actors. The topic is important given that presidents are usually visible and high-profile figures with a possibility to translate their public ‘mandate’ into political power. While the majority of semi-presidential studies have focused on presidents from an institutional perspective, this thesis examines their public position by providing a comprehensive picture of the aggregate-level and individual-level explanations of presidential popularity under a dual executive structure. Broadening also our conceptual understanding of presidential popularity, it suggests that public evaluations of the president reflect not only citizens’ perceptions of the president’s performance but also their perceptions of the institution as a whole and the individual occupying the office. The existing studies – building on the seminal work by Mueller (1970) on the U.S. presidency – have mainly concentrated on relatively powerful presidents in president-led systems. Four main explanatory categories are identified in the literature: the electoral cycle (referring to the ‘honeymoon effect’ and the ‘cost of ruling effect’), major events (the so-called ‘rally around the flag effect’), the economy (with mixed evidence, though), and the president’s party background. Case studies on semi-presidential countries, France and Portugal, have to some extent considered the institutional features of the regime type, including for example periods of cohabitation, i.e., when the president and the prime minister represent different parties and the president’s party is not represented in the cabinet. Otherwise, how power dynamics between the executives along with other features of the presidential office shape public opinion on presidents remain underexplored. This thesis addresses this apparent gap by drawing links between two above-mentioned broad literatures: on semi-presidentialism and on public opinion towards political executives. Empirically, it relies on various aggregate- and individual-level public opinion surveys and contextual data collected from several European semi- presidential countries over the period from 2000 to 2020 as well as data from presidential countries outside Europe. The dissertation consists of three peer-reviewed journal articles and an introductory chapter. The first article published in European Political Science Review focuses on the association between presidential popularity and the level of presidential powers. It shows that the constitutional strength of the presidency shapes both the general level of presidential popularity and to what extent people punish the presidents for changes in economy or get disappointed in them over time. Moreover, constitutionally weaker presidents are generally more popular than their stronger counterparts, and they retain better their popularity ratings under a weakening economy and following a honeymoon period. An important implication of this study is that the level of presidential powers should be considered as the primary explanatory factor in comparative studies of public opinion on presidents in semi-presidential regimes and beyond. The second article published in Political Studies introduces an important additional perspective by focusing on the more practical power dynamics between the executives and their impact on presidential popularity in relation to the popularity of the government. This includes party-political shifts and clashes between the executives that are expected to boost presidential leverage in situations where the president is politically opposed to the government. Concentrating on European semi-presidential countries – the Czech Republic, Finland, France, Poland, and Portugal – the results show that while both weak and strong presidents benefit from periods of minority governments, only stronger presidents benefit from conflicts with prime ministers and to some extent from periods of cohabitation. Weaker presidents, in turn, are expected to maintain a more neutral position. These novel findings emphasize the importance of considering the complex relationship between the two executives when studying their public popularity. The third article published in East European Politics approaches the topic from the perspective of political trust with individual-level data by examining whether the president’s alleged position ‘above’ parties or party politics explains support for the president in relation to other, arguably more ‘partisan’ political institutions. The focus is on a group of countries in Central and Eastern Europe, where the presidents tend to enjoy stronger levels of trust in relation to other political institutions, such as the government, prime minister, and political parties, but the presidents’ efforts to stay outside party-political disputes vary significantly. The explorative analysis reveals that while trust in president is influenced by party identification – meaning that citizens who support opposite party ideologies are more likely to express distrust in the president – it also reflects dissatisfaction with the government’s work, although prime ministers seem to be more consistently harmed by this factor. The results strongly indicate that presidents like prime ministers are actually evaluated through ‘partisan lenses’ and not systematically placed above party politics in semi- presidential regimes. Together, the empirical findings of the three articles show that rather than the regime type itself the president’s authority within the system – dependent on the level of the president’s formal powers but also on the more practical power dynamics between the president and the prime minister – shapes public evaluations of the president so that the impact of external factors such as the economy or the electoral cycle become more irrelevant. This is not a problem as such, especially if the responsibility attributions reflect the actual level of presidential powers, yet it is not ideal for the functioning of our representative systems if the president’s more favourable public image shields the president from critique even in situations where voters should hold the president accountable. Against this background, the thesis presented also the idea that the ‘symbolic’ features associated with the presidency, such as the president’s alleged position ‘above’ parties or party politics, may benefit the president’s public position by obscuring the lines of responsibility. The empirical evidence indicates that this is not the case, though: people seem to hold those in power accountable (referring to stronger presidents and prime ministers), but they also pay attention to the actions of those not in power (weaker presidents). This doctoral dissertation produces novel insights on the president’s public position and role in semi-presidential democracies. Overall, the theoretical and empirical input of the thesis contributes to our understanding of political responsibility and accountability in systems with a dual executive structure. This, in turn, opens new avenues for studies concentrating on the causes and consequences of presidential authority.
... We first run a Hausman test to examine whether differences in random and fixed effects coefficients were systematic. Results of this test yielded a χ 2 value of 798.46 and a p value < 0.001, suggesting a strong preference for the fixed effects model over the random effects model, a strategy often considered a more conservative approach (Allison, 2009). We then proceed to consider modeling strategies that account for both serial autocorrelation as well as cross-sectional panel dependence, as was found to be present following the administration of Pesaran CD test (Hoechle, 2007). ...
... In comparing the Driscoll-Kraay procedure with the robust clustered standard errors commonly used in research on political economy, the Driscoll-Kraay procedure yields larger standard errors and thus represents a more conservative test (Beck and Katz 1995;Tober, 2019). All models are estimated with two-way country and year fixed effects to account for unobserved heterogeneity that may affect the robustness of the results we have found (Allison, 2009). As results are consistent across the range of lag structures we investigated, up to the maximum lag length (m(T) = T-1), we have opted to present results based on the default moving average lag setting m(T) = floor[4(T/100) 2/9 ] (Hoechle, 2007, p. 10) that results in a lag of 3 years. ...
Scholars of punishment have long been interested in secondary consequences of
criminal justice contact. Recent work in this vein demonstrates that higher levels of
incarceration puts negative pressure on labor unions and unionization rates more
broadly writ, yet much of this work focuses solely on the United States—underscoring important gaps in our knowledge of how the prison operates in broader context. This article extends this research to explore the extent to which incarceration rates across 36 OECD countries affect unionization from 1961 – 2017. Results from panel data analysis support that incarceration rates diminish union density across context and time. These findings contribute to literatures on neoliberal penality, union decline, and investigations into consequences of incarceration beyond the somewhat exceptional case of US penal practice.
... In the second analysis, the association between primary care utilisation and the MMR was examined using a panel regression with two-way fixed effects (TWFE). The TWFE panel regression has the advantage of adjusting for unobservable and time-invariant confounders by adding year and unit fixed effects [36]. All the analyses were stratified by the HDI levels to examine the differential reform impacts and the association between primary care utilisation and the MMR across regions. ...
Background
China’s maternal health has substantial inequalities across regions, a similar challenge faced by many low- and middle-income countries. The Chinese government launched a comprehensive health reform since 2015 to deliver more affordable and equitable primary health care (PHC), with pregnant women being a priority group of beneficiaries. However, little is known about the impacts of this PHC reform on primary care utilisation among pregnant women or maternal health inequalities. This study aims to examine whether and how China’s PHC reform affected primary care utilisation among pregnant women and maternity deaths differently across regions.
Methods
The study employed provincial-level panel data from the China Health Statistic Yearbook and China Statistic Yearbook (2010–2019). Reform implementation by province was identified using web-scrapping of 31 provincial government websites. Firstly, difference-in-differences method examined the reform impacts on visits to PHC facilities, the utilisation of family physician services and prenatal services, and the maternal mortality ratio (MMR). Secondly, fixed-effects panel regression models estimated the association between family physician service use, prenatal care and the MMR. Analyses were stratified by province human development index (HDI) to assess inequalities.
Results
The introduction of China’s PHC reform in a province was associated with increased utilisation of family physician services (59.7 per 10,000 people per year, 95% CI 32.8–86.5) and prenatal services (3.2% points per year, 95% CI 1.8–4.6) and reduced maternal death by 9.6 per 100,000 live births per year (95% CI 0.3–19.0) in low-HDI provinces. No reform impact was found in high-HDI provinces. In panel regression models for low-HDI provinces, with a 1.0% point increase in prenatal care utilisation and one increase in family physician visit per 100 people, maternal deaths would decrease by 1.4 (95% CI 0.2–2.5) and 2.4 (95% CI 1.4–3.5) per 100,000 live births per year, respectively. This association was not found in high-HDI provinces.
Conclusion
China’s PHC reforms and primary care utilisation were associated with reduction in MMR in less developed regions, suggesting contributions to lower inequalities in maternal health between regions. Community-level family physician services are likely effective for improving maternal health in high burden areas, but further system and quality improvements are needed in areas where maternal mortality is lower.
... We also expand that model and rely upon the 'within-between' (hybrid) framework that combines the desirable aspects of both fixed effects and random effects econometric models (Mundlak, 1978). Hybrid models can be estimated by decomposing each time-varying predictor into a within-person component and a between-person component, and then estimating a random effects model with both components; the within-person component (i.e., the observation's deviation from that unit-specific mean) is equivalent to a fixed effects estimator, whereas the between-person component models unit-specific means of each variable (i.e., equivalent to time-invariant covariates) (Allison, 2009). ...
... The selection of a random-effects (RE) model is informed by both the panel structure and the temporal properties of key variables. Given the two-wave panel (2019 and 2023), within-individual variation is limited-especially for relatively stable variables such as homeownership (Allison, 2009;Bell & Jones, 2015). In such contexts, fixed-effects (FE) models often yield inefficient estimates, as they rely solely on within-unit changes and tend to inflate standard errors due to overfitting (Clark & Linzer, 2015). ...
This study examines how housing wealth influences the political incorporation of immigrants in South Korea, identifying asset accumulation as a key factor shaping redistributive preferences. Using nationally representative panel data, the analysis demonstrates that immigrants who acquire housing assets—whether through ownership or appreciation in property value—become less supportive of public provision directed at other foreign residents. While immigrants initially exhibit stronger support for state intervention, greater exposure to the housing market is associated with a growing perception that existing government support is adequate and a reduced inclination to endorse expanded immigrant-targeted spending. These shifts reflect not a uniform turn toward conservatism, but a reconfiguration of political attitudes shaped by economic security, social positioning, financial constraints, and the redrawing of intra-group boundaries. In South Korea’s asset-based welfare regime, housing functions not only as a source of private stability but also as a mechanism that mediates immigrants’ evolving relationship to the state and to other immigrant groups. By situating these dynamics within a non-Western context, this study contributes to broader debates on immigration, housing, and the stratifying effects of economic incorporation.
... First, we present the results of non-spatial regressions of the dynamic panel model using FE and System GMM estimates. The former controls for omitted variable bias (Allison 2009, Wooldridge 2010) and the latter for the endogeneity ...
This paper examines how knowledge spillovers shape economic convergence across countries in the Americas between 1950 and 2019. Using spatial econometric models, including a Spatial Dynamic Durbin Model (SDDM) and a Spatial System Generalized Method of Moments (SGMMsys) estimator, that jointly accounts for spatial spillovers, dynamic endogeneity, and cross-sectional heterogeneity within a single framework, we explore how interdependencies in physical and human capital accumulation, as well as population growth, influence per capita income dynamics across borders. Our findings show that spatial effects significantly accelerate the rate of β-convergence, contributing to the emergence of "convergence clubs", clusters of countries with shared development trajectories. The results underscore the importance of regional cooperation, cross-border investment in human capital, and knowledge diffusion as catalysts for inclusive and sustainable development. Policy frameworks that promote spatially integrated strategies can help reduce persistent income disparities in Latin America and foster stronger regional economic integration.
... This means the models estimate elasticity coefficients where the coefficient for the independent variable is the estimated net percentage change in the dependent variable associated with a 1 percent increase in the independent variable. The primary reported models are estimated with the xtreg fe command, which uses the within estimator to account for countrylevel fixed effects, and temporal fixed effects are estimated with year-specific dummy variables (Allison 2009). ...
Bridging the areas of anthropogenic drivers research in sociology on world-economic integration and militarization, and drawing from macrosociological research on foreign direct investment (FDI), we argue that capital-intensive military power facilitates and supports transnational capital outsourcing their carbon pollution to Global South nations, and this carbon load displacement occurs independent of the overall environmental impacts of the volume of inward FDI. To test these arguments, we create a new measure that quantifies the relative extent to which stocks of inward FDI are sent by nations with more powerful capital-intensive militaries. We use this new variable, along with the well-established measure of inward stocks of FDI as % GDP, as our independent variables of interest in analyses of carbon emissions for a sample of Global South nations from 2000 to 2020. The findings support our arguments. Both primary independent variables have positive short-run and long-run effects on total emissions, emissions per unit of GDP, and per capita emissions. We also find nontrivial evidence of expansion-leaning asymmetry: with few exceptions, increases in both key predictors have larger effects on increasing emissions than proportional decreases in them have in leading to reductions in emissions.
... And thereby accounts for any unobserved heterogeneity by including a fixed intercept for each unit, i.e. the respondent. They, however, are not able to provide estimates for time-constant characteristics (Allison, 2009). All these sensitivity analyses confirm the study results (results available upon request). ...
The COVID-19 pandemic was accompanied by increasing concerns about its impact on our mental health. Especially older adults were thought to be extra vulnerable to the negative consequences of the pandemic. This was quickly validated by early studies showing increasing levels of mental distress during the first stages of the pandemic. Despite the challenges posed by the pandemic, several studies also found that older adults tend to be more resilient to its negative effects compared to younger age groups. These studies, however, lack in pre-pandemic controls as well as representativity for the older population. Furthermore, they are solely focused on the short-term impact. This study addresses these gaps by covering an observation period from 2019 until 2022, using cross-national representative data from the Survey of Health, Ageing and Retirement in Europe (SHARE). The random-effects models demonstrate that older individuals indeed show lower levels of depressive symptoms and sleeping problems during the pandemic, but this was only temporary. Measures of loneliness show a steady incline, that reaches its top during the last stages of the pandemic. In sum, this study warns for unjustified optimism about the mental resilience of older adults and advocates for continued efforts into maintaining and improving the mental health of older Europeans to alter these pessimistic long-term trends.
... We apply fixed effects linear regressions to model the dynamic associations between time use, parenthood, and mental health. Fixed effects models exploit the within-individual change in time-varying variables, eliminating time-constant individual effects and unobserved heterogeneity, which produces mostly unbiased estimations (conditionally on no confounding by unobserved time-varying variables) in a modeling framework that requires few assumptions (Allison 2009). The estimated parameters reflect the average effect of a within-individual change in the variable of interest on the outcome, that is, the parenthood variable reflects the effect of the transition into parenthood and the progression into subsequent years on individuals' mental health, relative to their baseline level before becoming parents (i.e., −3 years before), everything else equal. ...
Objective
This study examines the role of changes in paid work and housework time on first‐time mothers' and fathers' mental health trajectories.
Background
The transition into parenthood is a key life course event with important consequences for individuals' activity patterns and couples' division of labor. Yet, whether gendered shifts in paid and domestic work time are linked to men's and women's mental health during the transition to parenthood remains unclear.
Method
Using large‐scale panel data from the Australian HILDA survey for men and women in different‐sex couples (2002–2022; N = 5932), we apply a longitudinal mediation framework with fixed effects models to determine the extent to which mental health trajectories are affected by changes in paid work and housework hours across the first transition into parenthood, considering both individual and partner‐relative contributions.
Results
Individual and partner‐relative paid work hours are positively associated with mental health for both men and women, while individual housework hours negatively impact only women. Following parenthood, women experience substantial reductions in paid work and increases in housework hours, but men's time use stays unchanged. Accordingly, despite overall improvements in women's mental health trajectories, findings show that these parenthood‐related changes in time allocations suppress some of the positive effects of childbearing for women, whereas men remain unaffected.
Conclusion
The transition to parenthood markedly reinforces gendered time use patterns in paid work and housework within couples, with disadvantageous shifts for women that result in small reductions in first‐time mothers' mental health trajectories. The potential factors underlying these findings are discussed.
... The FE model is our primary analytical method to control unobserved time-invariant heterogeneity across organizations, such as institutional culture and geographic constraints that may affect flexible labor adoption. However, FE models eliminate time-invariant variables through within-unit transformation (Allison, 2005(Allison, , 2009, making Hypothesis 6, which is proxied by timeinvariant dummies (i.e., the type of arts nonprofits), untestable in this model. To address this limitation, we supplement our analysis with an OLS model. ...
As nonprofits operate in a competitive environment with limited resources, they constantly seek new ways to optimize their resources. This study investigates factors influencing nonprofits’ decision to integrate flexible labor, such as independent contractors, into their workforce. Using longitudinal data from 2008 to 2018 in the arts and cultural sector in the United States, this study tests hypotheses related to the impact of an organization’s financial health, cost of permanent employment, reliance on government funding and donations, organizational size, and service demand variations on flexible labor use. The findings confirm that nonprofits offering higher fringe benefits and facing greater service demand fluctuations rely more on flexible labor. However, contrary to our expectations, this study also finds that nonprofits with stronger long-term financial health are more inclined to adopt flexible labor, while larger nonprofits use less flexible labor than their smaller counterparts. This study advances our understanding of the organizational and sector-level factors behind flexible labor adoption in nonprofits and offers practical implications for managing it.
... Fixed Effect (FE) estimate is widely utilized in many disciplines since its justification is clear-cut and compelling. All higher-level variation and any between-effects variance are eliminated by employing the higher-level entities themselves (Allison, 2009), which are included in the model as dummy variables Dj to prevent the issue of heterogeneity bias. The formula is expressed as Equation 17: ...
In the pursuit of sustainable development, green productivity has emerged as a crucial concept, bridging the gap between ecological conservation and economic prosperity to foster both environmental protection and human welfare. This research undertakes an in-depth exploration of the intricate associations among natural resource utilization, the forces of globalization, advancements in financial technology (fintech), human development, and green productivity within the ASEAN region during the period from 2000 to 2021. Employing advanced econometric techniques such as the Augmented Mean Group (AMG), Common Correlated Effects Mean Group (CCEMG), Fixed Effects modeling, and the Moments Quantile Regression (MMQR) to dissect panel data, the study aims to decipher the multifaceted impacts of these variables on environmental sustainability. The results reveal that overreliance on natural resources and the process of globalization pose challenges to green production, while economic expansion and human development initiatives act as catalysts for sustainability. The symbiotic relationship between human development and green business practices further accentuates the importance of holistic policy formulation. Notably, the findings highlight that the implementation of stringent environmental regulations and the adoption of sustainable resource management strategies not only enhance green production but also stimulate economic growth without compromising environmental integrity. This study not only enriches the theoretical framework of ecological economics but also uncovers the critical interdependencies of sustainability components within the ASEAN context. By spotlighting the necessity for integrated policies that harmonize economic and environmental objectives, this research offers valuable insights to policymakers striving to achieve sustainable development in the region. Moreover, it emphasizes the urgency of a comprehensive approach to address the sustainability conundrums faced by ASEAN nations, paving the way for future investigations into the complex web of multiple sustainability issues prevalent in developing economies.
... Specifically, fixed effects regression removes the influence of time-invariant factors by differencing out or absorbing counselor-specific characteristics, ensuring that only within-counselor variations over time contribute to the estimated effects. This approach reduces omitted variable bias by ensuring that any unobserved but constant caller-related factors associated with specific counselors do not confound our estimates (Allison, 2009). We utilized robust standard errors adjusted for the 281 clusters in the panel to account for heteroskedasticity and autocorrelation, ensuring that our estimates remain statistically valid even in the presence of within-cluster correlation. ...
Emotional support helplines are a longstanding strategy for supporting parents, but few studies have measured how much these programs work to address negative feelings in parents. Drawing on a sample of 281 helpline calls from parents to the California Parent and Youth Helpline®, the authors analyzed changes in emotions expressed through language, while parents spoke to a counselor. The authors used a sequential explanatory mixed methods design. They worked with a Natural Language Inference (NLI) model, BART, to measure and quantitatively display changes in feelings parents express. They found that 85% of the parents made more positive statements as the call proceeded. The authors tested BART by measuring its rate of agreement with qualitative coding by the authors and against a codex constructed by service users. The authors discuss the implications of this study for addressing parental isolation and how the research methods could be adopted by other helplines.
... Yet, rather than jumping ship for a fixed effects model at the first sign of a leak, this exogeneity assumption can be relaxed within the framework of a mixed effects model through a parameterization where some time-varying covariates are centered by their group specific means ( x ij − x j ) , and these group mean-centered variables are added into the regression equation along with the group-specific means. This "between-within" (Bell and Jones 2015) or "hybrid" parameterization (Allison 2009) is represented as: ...
Purpose
To provide guidance on the use of Bayesian mixed-effect modeling for inference in criminological applications involving complex longitudinal data, and a demonstration involving police firearm seizures and gun violence.
Methods
We outline a Bayesian strategy for specifying and fitting mixed-effects models using the brms package in R, including procedures for prior predictive simulation, model selection, and posterior comparison during sensitivity analyses.
Results
To demonstrate the utility of the Bayesian approach, we apply it to complex longitudinal data from a police directed patrol intervention in Flint Michigan, estimating the impact of gun seizures on firearm violence using a three-level mixed effects zero-inflated negative binomial regression model. The demonstration suggested a small and fragile reduction in reported gun violence associated with firearm seizures, which was not sufficiently distinguishable from placebo-based sensitivity analyses.
Conclusions
Bayesian methods are underutilized compared to frequentist approaches in criminological applications. This manuscript outlines an approach to expanding the criminological “toolkit” for mixed-effects modeling that leverages recent advances in software for Bayesian inference.
... Allison [25], and specifically regarding CRTs in previous work [26,27,28]. However, the main purpose of our article is not to discuss the philosophical differences between specifying cluster or cluster-period effects via random or fixed effects. ...
In the analysis of cluster randomized trials (CRTs), previous work has defined two meaningful estimands: the individual-average treatment effect (iATE) and cluster-average treatment effect (cATE) estimand, to address individual and cluster-level hypotheses. In multi-period CRT designs, such as the cluster randomized crossover (CRXO) trial, additional weighted average treatment effect estimands help fully reflect the longitudinal nature of these trial designs, namely the cluster-period-average treatment effect (cpATE) and period-average treatment effect (pATE). We define different forms of informative sizes, where the treatment effects vary according to cluster, period, and/or cluster-period sizes, which subsequently cause these estimands to differ in magnitude. Under such conditions, we demonstrate which of the unweighted, inverse cluster-period size weighted, inverse cluster size weighted, and inverse period size weighted: (i.) independence estimating equation, (ii.) fixed effects model, (iii.) exchangeable mixed effects model, and (iv.) nested exchangeable mixed effects model treatment effect estimators are consistent for the aforementioned estimands in 2-period cross-sectional CRXO designs with continuous outcomes. We report a simulation study and conclude with a reanalysis of a CRXO trial testing different treatments on hospital length of stay among patients receiving invasive mechanical ventilation. Notably, with informative sizes, the unweighted and weighted nested exchangeable mixed effects model estimators are not consistent for any meaningful estimand and can yield biased results. In contrast, the unweighted and weighted independence estimating equation, and under specific scenarios, the fixed effects model and exchangeable mixed effects model, can yield consistent and empirically unbiased estimators for meaningful estimands in 2-period CRXO trials.
... All higher-level Frontiers in Environmental Science frontiersin.org variation and any between-effects variance are eliminated by employing the higher-level entities themselves (Allison, 2009), which are included in the model as dummy variables Dj to prevent the issue of heterogeneity bias. ...
The rapid decline in environmental quality and the expanding ecological footprint (EFP) have become critical challenges, particularly for G20 nations that play a central role in global economic growth. This study investigates the determinants of the ecological footprint and its sub-components across 17 G²⁰ countries over the period 1996 to 2021. Using advanced econometric methods such as cross-sectional dependence tests, slope homogeneity tests, unit root tests, cointegration tests, GMM, fixed effect models, and Granger causality analysis, this research provides a comprehensive analysis of key drivers. The findings highlight that technological advancements significantly reduce the ecological footprint, especially by enhancing environmental regulations and fostering sustainable practices. Human capital (HC) and institutional quality (IQ) emerge as critical contributors to sustainability, while globalization (GB) demonstrates mixed effects on ecological outcomes. Moreover, stringent environmental policies (EPS) exhibit robust bidirectional causal relationships with EFP, underscoring their vital role in mitigating environmental degradation. The study underscores the importance of targeted governmental interventions to promote technological innovation, strengthen institutional frameworks, and enforce rigorous environmental regulations. These insights provide actionable guidance for G20 nations to balance economic growth with environmental sustainability, aligning with global sustainability goals.
... Journal of Emerging Markets effects model to address unobserved characteristics in the context of a longitudinal dataset (Allison, 2009). We also obtain robust standard errors to produce a consistent variancecovariance matrix estimator. ...
Purpose
This research examines the connection between intellectual capital (IC) and risk profile (credit risk, insolvency risk and liquidity risk) in the Vietnamese banking sector.
Design/methodology/approach
The data were adopted from the quarterly financial statements of 27 commercial banks in Vietnam from 2008 to 2024. This paper used the value-added intellectual coefficient (VAIC) methodology to measure IC.
Findings
The findings of the main analysis reveal that capital employed efficiency (CEE) and human capital efficiency (HCE) are associated with all three types of risk measured. The disaggregated ownership results indicate that only HCE is related to risks in state-owned commercial banks, whereas both HCE and CEE are linked to risks in private commercial banks. The analysis findings also suggest that the relationship between CEE and risk is moderated by income diversification, highlighting the importance of diverse management strategies for more effective risk management.
Practical implications
By focusing on improving human capital, integrating risk management into their core strategies and leveraging technology, Vietnamese banks can effectively mitigate risks and enhance their financial stability.
Originality/value
This paper is the first to examine the correlations between IC and banking risks in Vietnam. This study divided the ownership structure of Vietnamese banks into public and private banks. It also examined the moderating effect of income diversification on IC and bank risks in Vietnam.
... These types of models are ideal for count data and were more appropriate than their Poisson counterpart given the means and variances of our dependent variables were not equivalent (Fisher & Lee, 2019). The addition of fixed effects also allows for the researcher to investigate changes in the independent variables (here, pandemic rhetoric and restrictions) on changes in the dependent variables (here, hate crime) independent of time-stable unobserved heterogeneity across states (Allison, 2009;Wolfowicz et al., 2023). The unit of analysis for the fixed-effects regression models is the state-month, which is used to account for unmeasured time stable variation across U.S. states. ...
Xenophobic extremism directed towards Asian persons, as opera-tionalized in a multitude of ways, was found to be especially heightened during the COVID-19 pandemic. The emotion-threat and emboldenment frameworks suggest that such extremism may have been the result of pandemic-related restrictions and rhetoric, which led to an increase in negative emotions directed toward the "outgroup, " as further defined by political leaders. Informed by these perspectives, this study examines whether COVID-19 pandemic restrictions and rhetoric affected hate crime in the U.S. Using a multi-definition and multi-method approach, this research finds that both COVID-19 restrictions and rhetoric were related to state-level increases in hate crimes against Asian persons. However, subsequent analyses revealed that individual cities saw both reductions in hate crimes and null impacts, while quarterly analyses were unable to replicate findings that were observable using monthly data. This suite of results highlights the need for a trian-gulation strategy when examining questions related to hate crime.
... We tested the association between time-saving purchases and relationship satisfaction using multilevel mixed-effects models with disaggregated within-person and between-person effects to address confounding factors and time-invariant alternative explanations (Curran & Bauer, 2011;Wang & Maxwell, 2015). This approach allows us to isolate the effect of daily fluctuations in time-saving purchases on changes in relationship satisfaction, controlling for stable individual differences that may influence this association, such as household income (Allison, 2009;Hoffman & Stawski, 2009). ...
Seven studies examine the association between time-saving purchases (e.g., housecleaning and meal delivery services) and relationship satisfaction. Study 1 uses an 11-year longitudinal panel survey to show that increases in time-saving purchases predict long-term increases in relationship satisfaction. Study 2 replicates these findings with a 6-week daily diary study, demonstrating that time-saving purchases predict daily increases in relationship satisfaction, particularly for members of dual-income couples who are experiencing higher levels of stress. Studies 3 through 4b reveal that time-saving purchases are most beneficial when couples translate this influx of temporal resources into quality time spent together. Study 5 identifies two key aspects of quality time—positive mood when together and perceived support—that uniquely predict relationship satisfaction. Study 6, a preregistered study, provides evidence for our conceptual model: Members of committed relationships who make time-saving purchases more effectively manage daily stressors (i.e., household chores) and spend more quality time together, which predicts increased relationship satisfaction. Once again, these benefits are strongest for individuals experiencing higher levels of stress. These findings develop a nuanced framework connecting time-saving purchases to relationship satisfaction.
... Additionally, is the constant term, is the coefficient of x it , u i represents the characteristics of the panel unit, and e it is the error term. The FEM treats u i as the parameter to be estimated (Allison 2009). The appropriateness of the FEM was determined using an F-test, where the null hypothesis H 0 ∶ u i = 0 was tested. ...
This study applies the fixed-effects threshold model (FETM) to examine nonlinear dynamics and threshold effects in Korean finfish aquaculture using panel data from 2010 to 2023. The analysis identifies a critical threshold in current biomass (log-transformed 13.9483, approximately 1,142,010 fish), beyond which the marginal productivity of stocking density declines due to resource competition, oxygen depletion, and metabolic stress. Among key production inputs, feed input exhibits the highest elasticity (0.4717), confirming its dominant role in production efficiency, whereas water temperature shows a significant negative elasticity (-1.3301), highlighting the adverse impact of rising water temperatures on fish growth and survival. Additionally, many finfish aquaculture regions in Korea face severe constraints due to overuse of water surface areas, leading to diminishing marginal returns and heightened ecological stress. These findings highlight the urgent need for comprehensive environmental assessments, real-time monitoring systems, and adaptive management strategies to enhance sustainable aquaculture practices. Particularly, optimizing stocking densities based on the identified current biomass threshold and implementing spatial planning policies can balance ecological sustainability and production efficiency. This study demonstrates the effectiveness of FETM in detecting nonlinear relationships and production thresholds in aquaculture systems, providing insights for evidence-based policymaking and sustainable resource management aligned with global sustainability goals.
... We chose a fixed-effects model since this technique nets out all unobserved time-invariant heterogeneity of individuals. In other words, each individual is treated as their own control in a longitudinal analysis (Allison 2009). In our case, this meant comparing the trajectory of researchers' Hindices in years before and after participating in WGs, and then averaging those differences across researchers. ...
Working groups are recognized as a highly effective method for synthesizing science. It is less clear if participating in working groups benefits individual researchers, or if benefits differ between men and women. This is a critical question, for the working group method is not sustainable if the benefit to science comes at a cost to academic careers or gender equity. Here we analyze the publications of Canadian university faculty specialized in ecology and evolution (N = 1244), a field that has embraced the working group method. Researchers were more likely to have participated in a working group as their academic age and prior H-index increased, but controlling for these factors there was no effect of gender. Using a longitudinal analysis, we find that researcher H-indices accrue 14% faster following their first working group publication, regardless of gender. Part of this acceleration may be the 3- to 5-fold higher citation rate of working group synthesis publications. In a survey (N = 169), researchers also report indirect benefits of working groups, at similar rates for men and women. Working groups are therefore good not just for science but also for scientists. Formalized mechanisms for collaborations such as working groups may also offset gender inequities in science.
... These additional factors are known to provide significant explanatory power for publication success [17]. Finally, we add binary dummy variables to account for fixed effects of publication year b y and journal of publication b j [21] (particularistic factors). The regression models are summarized by following, logarithmic link equation: ...
Most transportation publications to date lack sufficient transparency in terms of open source and data, in order to enable reproducibility. This work sets out to highlight the need for reproducibility of computational simulation studies, and to develop the level of professionality in the transportation domain when it comes to open source. Moreover, this work aims to equip the readers with a guide on best practices inspired by the computer science domain, templates for documentation of software and data, and guidelines on repository management and open source. The empirical analysis of more than 11,879 simulation studies and 672 repositories in the field reveals that only few works (less than 2%) share supplementary material alongside the scientific publication, on a mostly mediocre repository quality, but with a strong trend towards more transparency in the last years. Besides, simulation studies with enhanced transparency are found to not receive significantly more citations when compared with studies that do not provide additional repositories. What’s more, a survey of more than 87 transportation researchers conveys that reproducibility is perceived as a significant issue by the community, and that time and legal constraints are the largest impediments to more reproducible research.
... Next, we replicate the main models using alternative specifications to the mixed-effects models specified in Equation (1), including random-effects models, with and without Mundlak's (1978) correction, as well as hybrid models (Allison, 2009). Appendix 9 shows that our main findings remain consistent. ...
Research Summary
Research on organizational climate change adaptation has focused on high‐income nations, neglecting vulnerable regions already facing severe climate impacts. This study addresses this gap by examining how decision‐makers in vulnerable contexts respond to climate change, drawing on a unique panel dataset of over 3000 agricultural producers in Brazil. Using an abductive mixed‐methods approach, we develop a framework that explains how, contrary to findings from high‐income contexts, the experience of climate change and the resulting heightened perception of climate change risk are more likely to increase maladaptive rather than adaptive responses. These maladaptive responses create “climate traps”—vicious cycles driven by the interplay between decision‐makers' perceptions and ecological feedback loops, which further exacerbate vulnerability. Our framework offers a foundation for understanding adaptation patterns in vulnerable contexts.
Managerial Summary
Climate change poses significant challenges for decision‐makers, particularly in vulnerable regions where adaptation is critical for survival. This study examines how small agricultural producers in Brazil respond to climate change, revealing that, rather than adopting adaptive strategies, many implement maladaptive responses that exacerbate their future vulnerability. Using both quantitative and qualitative data, we find that the experience of climate change and a heightened risk perception drive short‐term survival strategies, creating “climate traps” that reinforce ecological and economic distress. Emotions such as fear and hopelessness play a crucial role in fostering decisions that prioritize immediate relief over long‐term resilience. This research highlights the need for interventions that account for the psychological barriers to adaptation in these contexts.
... Utilizing individual fixed effects for controlling time-univarite confounders the analysis still left space for possible alternative timespecific confounders. Due to its observational nature this study cannot determine that job autonomy causes mental health changes (Allison, 2009), so future research should address this issue. This study demonstrates perceived discrimination and well-being relationships merely through correlation while preventing strong causal deduction. ...
The research investigates how psychological well-being relates to discrimination levels among Pakistani female populations who are both married and unmarried. Psychological well-being plays an essential role in mental health but discrimination functions as a primary source to create negative impacts on one's well-being. This study obtained data through survey questionnaires from 206 women located in Multan as part of its cross-sectional research design. The research found a negative association between discrimination and psychological well-being which demonstrated significance at a p value of .013 (r = -.174). Married women showed higher average discrimination levels according to the results of t-test evaluation when compared to unmarried women. Psychological health scores among married women stood at lower levels than those of unmarried women according to research findings. Psychological well-being of women decreases because of discrimination yet marital status influences the experience of these negative effects. Researchers must study various social factors affecting discrimination alongside psychological well-being so proper programs for gender equality and mental health promotion can be developed.
... We standardized Black poverty (subtracting the grand mean and dividing by the grand standard deviation) and broke race/ethnicity down into two dummy-coded variables (comparing White to Black participants and White to Hispanic participants). We included year-fixed effects (Allison, 2009) and county-level random slopes for race/ethnicity dummies (Heisig & Schaeffer, 2019). Given the complexity of the analysis, we used restrictive iterative generalized least squares (RIGLS) rather than maximum likelihood (ML) estimation (Rasbash et al., 2004) (for the multilevel regression equation, see Supplementary Materials, p. 4). ...
In this article, we argue that White Americans exposed to Black poverty may rationalize racial inequality in ways that deflect blame from their own racial group. In Study 1 (11,855 participants; 537 counties), White Americans in counties with higher Black poverty rates were ironically more likely to believe in racial equality of opportunity, while Black residents were less likely. In Study 2 (4,297 participants; 227 counties), White Americans living in counties with higher Black poverty attributed racial inequality to more internal causes (lack of effort), which predicted reduced support for racial equity policies. In Study 3 (1,036 participants), White Americans experimentally exposed to Black poverty made more internal attributions, an effect mediated by increased interracial anxiety and identity threat. Effects on external attributions were inconsistent across studies. We discuss how internal attributions function as a psychological tool for White Americans to deflect negative emotions and maintain a positive group identity.
... To ensure that fixed-effects models were more appropriate than random-effects models, Hausman tests were conducted. 38 These tests yielded statistically significant results, indicating that random-effects models were potentially biased due to correlations between the independent variable and unobserved time-invariant characteristics. Based on these findings, fixed-effects models were more suitable for this analysis. ...
Objectives: This study explores the association between financial strain and loneliness over time, investigates loneliness as a mediator in the relationship between financial strain and mental health, and examines the influence of gender. Study design: Longitudinal study. Methods: Using data from six waves (2017-2023) of the UK Household Longitudinal Study (UKHLS), which includes 34,535 participants (154,316 person-years), this study employs fixed-effects regression models. Results: Financial strain is significantly associated with increased loneliness over time (b = .155, p < .001). Loneliness partially mediates the relationship between financial strain and mental health, accounting for approximately 15 % of the effect. Notably, the relationship between loneliness and mental health is moderated by gender (b = .120, p < .001), with women experiencing more severe negative effects. Conclusions: These findings suggest that alleviating financial strain could reduce loneliness and its detrimental effects on mental health. Addressing financial and social stressors is essential for public health strategies. Implementing gender-sensitive approaches is critical for addressing specific vulnerabilities, particularly among women.
This study investigates how the transition to parenthood is associated with shifts in personal attitudes and perceived public opinion towards formal childcare for children under three years. Using German panel data, we analyse those changes during the family formation phase using panel regressions with linear between-within (hybrid) models. The results for within-components suggest that cultural perceptions are strongly influenced by lifecourse events. After the birth of the first child, parents are less sceptical about daycare for children under three; however, we did not find any systematic changes in perceived public opinion. Between-components reveal that women, though less sceptical themselves, are more likely to perceive societal scepticism. There is an increasing divergence between personal attitudes and the perception of public opinion regarding formal childcare. This points to a conflict between the desire to use formal childcare and societal expectations, which has strong implications for gender equality.
Objective
The aim of this paper is to analyze (a) how highly educated women and men differ in their parental leave‐taking behavior and (b) how parental leave‐taking affects their subsequent wages.
Background
Labor market theories suggest that taking parental leave can have negative effects on career progress and wages. Consequently, the fact that women are much more likely than men to take parental leave is likely to contribute to long‐term career‐related gender inequalities. At the same time, some studies have shown that wage losses resulting from parental leave are greater for men than for women and are especially pronounced among highly qualified individuals.
Method
We analyzed data from a nationally representative panel study (7 waves from 2015 to 2021) with doctoral graduates in Germany from the cohort 2013/2014. We used fixed‐effects regressions to estimate intra‐individual changes in hourly wages due to parental leave‐taking.
Results
Highly educated women took parental leave more often and for much longer periods than highly educated men did. Taking extended parental leave was associated with a reduction in hourly wages. However, this general finding was slightly insignificant for first‐time parents. Against expectations, our analyses did not confirm higher wage penalties for men following a period of parental leave.
Conclusion
Our findings suggest that the gender‐specific use of parental leave is an important factor in the gender pay gap.
This study examines the relationship between work-to-family conflict (WTFC) and family-to-work conflict (FTWC) and drinking frequency and quantity by gender. Using longitudinal data from the Household, Income, and Labour Dynamics in Australia (HILDA) survey and analyzing around 40,000 observations of employed parents, we employ Tension Reduction Theory, Affect Regulation Theory, and Conservation of Resources Theory to propose an increase or decrease in alcohol consumption. Further, we expect gender differences due to gender roles and drinking motives. The results of our ordered logit regression models with fixed effects show that an increase in WTFC is positively and significantly associated with an increase in drinking frequency. However, this effect is only found among mothers but not fathers. In contrast, while there is no significant association with drinking quantity overall, gender differences are observed: higher levels of FTWC are significantly associated with an increase in drinking quantity, but only among fathers and not mothers. These findings highlight the importance of further research on conflict directions, the role of coping motives, and gender-specific responses to work-family conflict concerning alcohol consumption.
Although prevalent network theories posit that job seekers are motivated to utilize networks for instrumental benefits, we argue that sociopsychological barriers can discourage network use even when access is readily available. We examine this overlooked dynamic by exploring how sociopsychological factors influence network activation among mothers reentering the workforce after child-related career breaks. Analyzing panel data from a representative sample of Korean job seekers, we find that these mothers are markedly less inclined to use network-based search methods than other groups of women. This inclination remains when we account for individual fixed effects, employment gap length, and potential network disturbances. To uncover the reasons behind this distinct job-search behavior, we conducted in-depth interviews with 60 mothers who experienced child-related employment breaks and 38 other individuals, including mothers without such breaks, nonmothers, and fathers. Whereas existing theories attribute limited network use to restricted access, our qualitative evidence reveals deeper sociopsychological hurdles: mothers often anticipate challenges in reciprocating job-related assistance and the potential revelation of their inferior status when seeking help. These multifaceted internal conflicts not only hamper mothers’ swift reintegration into the labor force but also nudge them toward roles for which they are overqualified. As we bridge these findings, we identify novel theoretical avenues for future research, underscoring how sociopsychological hurdles to network activation can produce gender inequalities in the labor market.
Funding: Support from the Wharton Dean’s Research Fund and Wharton’s Center for Human Resources.
Objectives
Our goal was to examine the role of perceptions of mothers' cognitive impairment (CI) in shaping the impact of maternal differential treatment on adult children's psychological well‐being.
Background
The detrimental impact of mothers' differential treatment on adult children's psychological well‐being has been well‐documented; however, little is known about whether this association is moderated by mothers' CI.
Methods
Mixed‐methods data were collected from 287 adult children (mean age = 59 years) nested within 142 families as part of the Within‐Family Differences Study. Analytic approaches were multilevel regression and in‐depth qualitative analyses.
Results
Perceiving oneself as the child to whom one's mother was most emotionally close or with whom she had the most conflict was associated with lower psychological well‐being among adult children who did not report that their mothers had symptoms of CI; however, such perceptions were not associated with well‐being when children perceived that mothers showed symptoms of CI. Qualitative analyses revealed that when children reported that mothers did not have CI, children attributed their greater closeness and conflict to unique characteristics of themselves or their mother–child ties. However, when children perceived that mothers showed symptoms of CI, attributions for mothers' differentiation focused on contextual factors surrounding mothers' cognitive health, and thus did not reflect on the children themselves. Thus, the negative impact of perceptions of mothers' differential treatment on well‐being was muted.
Conclusion
This study sheds light on conditions under which perceptions of maternal differential treatment affect adult children's psychological well‐being, and the role of cognitive health in these processes.
Objectives
This study investigates the role of criminal justice intervention practices, i.e., opioid arrests, in effectively preventing or increasing opioid overdoses, paying particular attention to whether arrests in spatially proximate or socially connected communities lead to the displacement or prevention of opioid overdoses in a local community.
Methods
Combining data from the Cook County medical examiner, emergency medical services information, arrest reports, and commuting network statistics for Chicago’s 77 community areas between 2016 and 2019, this study uses fixed effects spatial autoregressive models with spatial lags to explain community-level opioid overdose rates.
Results
We find evidence for the diffusion and displacement of overdose risk as well as the diffusion of overdose-reducing benefits. Findings suggest complex spatial and social spillover mechanisms that both diffuse and prevent opioid overdoses, dependent on the type of opioid-related crime and overdose rate investigated.
Conclusions
These results have important implications for understanding the effectiveness of criminal justice policies in their goal of preventing opioid-related crime and overdoses and provide insights for designing more appropriate and effective policy responses to address substance use and illicit drug markets.
Background
Cannabis use is linked to treatment non-adherence and relapses in psychotic disorders. Antipsychotic medication is effective for relapse prevention in primary psychoses, but its effectiveness after cannabis-induced psychosis (CIP) remains unclear.
Aims
To examine the effectiveness of antipsychotic medication for relapse prevention following the first clinically diagnosed CIP.
Method
A cohort of 1772 patients (84.1% men) with incident CIP was identified from the Swedish National Patient and Micro Data for Analyses of Social Insurance registers. The primary outcome was hospitalisation due to any psychotic episode. Drug use data were collected from the Prescribed Drug Register and modelled into drug use periods using the PRE2DUP method. A within-individual Cox regression model was used to study the risk of outcomes during the use of different oral or long-acting injectable (LAI) antipsychotics compared with non-use.
Results
The mean age at first diagnosis was 26.6 years (s.d. = 8.3). Of the cohort, 1343 (75.8%) used antipsychotics and 914 (51.3%) experienced psychosis hospitalisation during the follow-up. Any antipsychotic use was associated with a decreased risk of psychosis hospitalisation (adjusted hazard ratio (aHR) 0.75; 95% CI 0.67–0.84). Specific antipsychotics associated with decreased risk included aripiprazole LAI (aHR 0.27; 95% CI 0.14–0.51), olanzapine LAI (aHR 0.28; 95% CI 0.15–0.53), clozapine (aHR 0.55; 95% CI 0.34–0.90), oral aripiprazole (aHR 0.64; 95% CI 0.45–0.91), antipsychotic polytherapy (aHR 0.74; 95% CI 0.63–0.87) and oral olanzapine (aHR 0.81; 95% CI 0.69–0.94).
Conclusions
In particular, LAIs, clozapine and oral aripiprazole were associated with a decreased risk of psychosis relapse following CIP. Prescribers should consider using more LAIs for better treatment outcomes after CIP.
The rapid increase in the elderly population presents a significant challenge in many developing countries, including ASEAN members. This study investigates the factors influencing population aging in Indonesia, Malaysia, the Myanmar, Thailand, and Vietnam over the period of 2000 to 2023. Employing a Fixed Effects Model (FEM), this research examines the relationship between the proportion of the elderly population and the independent variables: fertility rate, life expectancy, Human Development Index (HDI), and female labor force participation rate. The analysis reveals that fertility rate and female labor force participation rate exhibit a statistically significant negative relationship with the proportion of the elderly population. Conversely, life expectancy and HDI demonstrate a statistically significant positive influence on the proportion of the elderly population. These findings highlight the importance of policies that consider trends in fertility, improvements in healthcare and longevity, advancements in human development, and the evolving role of women in the workforce when addressing the implications of population aging in the region. In response to this demographic shift, ASEAN countries should prioritize comprehensive strategies encompassing social security reforms, enhanced elderly care systems, initiatives for workforce adaptation, and policies that support sustainable human development.
This study explores the role of financial technology (FinTech) in promoting financial inclusion across 28 emerging and developing economies (EMDEs) from 2011 to 2021. While financial inclusion has been widely studied, limited research focuses on the concurrent impact of FinTech on financial access in multiple EMDEs. Using Principal Component Analysis (PCA), a new financial inclusion index was constructed, incorporating two dimensions: access and usage. A System Generalized Method of Moments (GMM) model was applied to examine the relationship between FinTech and financial inclusion, while the Ordinary Least Squares (OLS) technique was used to assess factors influencing barriers to inclusion. The results reveal that a 1% increase in FinTech leads to a 0.1772 unit rise in the financial inclusion index. Additionally, education, GNI per capita, and broad money to GDP were found to affect barriers to financial inclusion. This study underscores the importance of FinTech in achieving the United Nations’ Sustainable Development Goals (SDGs) and G20 principles for digital financial inclusion, recommending investment in digital infrastructure and stronger regulatory frameworks for FinTech development.
Governors and state legislatures in the United States have a complex relationship. While we often consider them smaller versions of the federal government, this is not accurate. Like at the federal level, there is significant political competition between state legislatures and the executive branch. However, unlike the federal government, state legislatures often face a serious disadvantage because they are part-time in 46 out of 50 states and must consider how to balance power even in their absence. Our work focuses on how legislatures use tools such as Sunset Laws and Administrative Procedure Acts (APAs) to maintain this balance of power and ensure government agencies are responsive to legislative priorities. The central argument is that these mechanisms are not primarily about efficiency or cost savings but rather about power dynamics and preference alignment.
This study explores the complex and often overlooked connection between climate change and mental health, specifically examining how this relationship manifests across different income groups (high, middle, and low) from 1990 to 2020. A fixed‐effects model is used to analyze the impact of two key climate change indicators—annual surface temperature changes (TEMP) and greenhouse gas emissions—on a range of mental health outcomes, including mental disorders (MD), anxiety disorders (AD), depressive disorders (DD), bipolar disorder (BD), and schizophrenia (S). The findings reveal a statistically significant association between rising temperatures, increased greenhouse gas emissions, and declining mental health outcomes. While the negative effects of climate change on mental well‐being are observed across all countries, disproportionate impacts between income groups, suggesting that some populations are significantly more vulnerable than others. This study underscores the urgency of implementing comprehensive and equitable measures to mitigate climate change and protect mental health globally.
Background
Knowledge of airborne pollen seasons is essential for physicians to accurately diagnose and treat patients with allergic respiratory diseases. Although the Dallas-Fort Worth metropolitan area in North Texas is home to more than 8 million residents, it lacks a published pollen calendar.
Objective
Our objectives were 3-fold: (1) determine the most common allergenic pollens in North Texas and when they are present, (2) identify how pollen concentrations have changed over time, and (3) assess how weather affects the daily pollen concentrations.
Methods
We obtained 15 years of daily pollen concentration and weather data for North Texas. Data were analyzed in R by using the AeRobiology package. The AeRobiology package was used to interpolate missing data, create heatmaps of daily pollen concentrations, and calculate the pollen seasons. We use regressions accounting for seasonal effects to determine changes over time and the effect of weather factors.
Results
In North Texas, pollen is present throughout the year. Spring-dominant Quercus and winter-dominant Juniperus constitute more than half of the total annual pollen production. Ulmus and Ambrosia make up the bulk of the fall pollen. Consistent with climate change, daily pollen concentrations have been increasing over time. Moreover, higher pollen concentrations are associated with higher maximum daily temperature and average daily wind speed.
Conclusions
These pollen calendars will help physicians in the region care for patients with allergic respiratory disease, who may present with more severe disease as pollen concentrations are increasing over time.
This study examined how users’ self-views (i.e., identity coherence, identity confusion, self-esteem) and their self-presentation through avatars (i.e., the number of avatars used, self-avatar dissimilarity, the frequency of and the expenditure on avatar customization) in the metaverse mutually influence each other, based on a three-wave panel survey of Zepeto users in South Korea (N = 640). Dynamic fixed-effects models yielded no significant effects of self-views on avatar-based self-presentation, but identified some significant reverse relationships: users who spent more money on customizing their avatars’ appearance experienced greater identity confusion and lower self-esteem two months later. By contrast, contemporaneous fixed-effects models found no significant effects in either direction. Overall, the results indicate limited unidirectional effects of avatar-based identity experimentation on users’ self-views that manifest over time, while casting doubt on the permeability between virtual and physical self-identities.
The fertility expectations of older women and men are becoming increasingly important for understanding fertility dynamics, given the increasing share of births after age 30. Because most health conditions deteriorate with age, understanding the relationship between health and fertility expectations is essential. We investigate whether changes in self-assessed general, physical, and mental health are linked to revised fertility expectations and how these associations vary over the life course. Drawing on a large longitudinal dataset for Australia, we demonstrate that across each health indicator, self-assessed poor health corresponds to lower fertility expectations and that a deterioration (or improvement) in self-assessed health coincides with a decrease (or increase) in men's and women's expectations of having a child. Individuals adapt their expectations more in response to physical health changes if they are older, and mental health conditions at younger ages appear relevant to men's fertility intentions. The results highlight that general, physical, and mental health are crucial drivers of changes in fertility plans, emphasizing the importance of integrating health considerations into future theoretical frameworks and empirical analyses of fertility.
Anxiety is often seen as a driver of far-right politics in British political culture that is strategically irrational insofar as the consequences of the policies pursued by such parties contribute to an increase in poverty and inequality, which are drivers of anxiety. This article shows that anxiety can also drive voter support for strategically rational, progressive policies, which hold out a real prospect for addressing the threats that voters face. Moreover, once established, those preferences appear to outlast occurrent anxiety experiences. Previous studies have found cross-sectional associations between socioeconomic status and anxiety and support for Basic Income, as an archetypal redistributive social security measure, on the other. In this article, we present the results of a novel longitudinal study of red wall voters’ (n = 304) policy preferences over two waves between 2022 and 2024 in the run-up to the UK General Election, using a fixed effects model examining associations between socioeconomic status, anxiety, and changes in public policy preferences. We also examine cross-sectional associations between various socioeconomic and demographic variables and Basic Income support in the 2024 wave (total n = 1988 [red wall n = 913]) in comparison to the 2022 wave (red wall n = 805). While overall longitudinal levels of anxiety have reduced, real income levels—taking into account inflation—and support for Basic Income remain remarkably stable, with a slightly significant positive association between increased faith in politicians and support for Basic Income. The findings suggest a relationship between socioeconomic status, anxiety, and support for social security-enhancing measures. This suggests that progressive parties can satisfy psychosocial needs via redistributive reform and gain popular support as a consequence.
ResearchGate has not been able to resolve any references for this publication.