Andrew S. Fullerton

A Conceptual Framework for Ordered Logistic Regression Models

Ordinal-level measures are very common in social science research. Researchers often analyze ordinal dependent variables using the proportional odds logistic regression model. However, this ‘‘traditional’’ method is one of many different types of logistic regression models available for the analysis of ordered response variables. In this article, the author identifies 12 distinct models that rely on logistic regression and fit within a framework of three major approaches with variations within each approach based on the application of the proportional odds assumption. This typology provides a degree of conceptual clarity that is missing in the extant literature on logistic regression models for ordinal outcomes. The author illustrates the similarities and differences among the different models with examples from the General Social Survey and the American National Election Study.

Putting Poverty in Political Context: A Multi-Level Analysis of Adult Poverty across 18 Affluent Democracies

Our study analyzes how political context, embodied by the welfare state and Leftist political actors, shapes individual poverty. Using the Luxembourg Income Study, we conduct a multi-level analysis of working-aged adult poverty across 18 affluent Western democracies. Our index of welfare generosity has a negative effect on poverty net of individual characteristics and structural context. For each standard deviation increase in welfare generosity, the odds of poverty decline by a factor of 2.3. The odds of poverty in the United States (the least generous welfare state) are greater by a factor of 16.6 than a person with identical characteristics in Denmark (the most generous welfare state). Significant interaction effects suggest that welfare generosity reduces the extent to which low education and the number of children increase poverty. Also, welfare generosity reduces poverty among those with low education, single-mother households and young households. We show that Leftist parties and union density reduce the odds of poverty, however their effects channel through the welfare state. Ultimately, poverty is shaped both by individual characteristics and the political context in which the individual resides.

Reexamining the Relationship Between Flexibility and Insecurity

Purpose – In this chapter, we examine individual- and country-level differences in perceived job insecurity in the 27 European Union countries (EU27) within a multilevel framework. Design/methodology/approach – We primarily focus on cross-national differences in perceived job insecurity in the EU27 and consider several possible explanations of it, including flexible employment practices, economic conditions, labor market structure, and political institutions. We examine both individual- and country-level determinants using multilevel partial proportional odds models based on individual-level data from the 2006 Eurobarometer 65.3 and country-level data from a variety of sources. Findings – We find that European workers feel most insecure in countries with high unemployment, low union density, low levels of part-time and temporary employment, relatively little social spending on unemployment benefits, and in post-socialist countries. Research limitations/implications – The findings from this study suggest that flexible employment practices do not necessarily cause workers to feel insecure in their jobs. This is likely due to the different nature of part-time and temporary employment in different institutional contexts. Originality/value – This study is one of the most comprehensive accounts of perceived job insecurity in Europe given the focus on a larger number of countries and macro-level explanations for perceived job insecurity.

Explaining the Persistence and Eventual Decline of the Gender Gap in Voter Registration and Turnout in the American South, 1956–1980

Recent studies of the gender gap in politics tend to focus on candidate choice rather than registration and turnout. This shift in focus away from gender inequality in political participation may be due to the finding in several studies of U.S. voting behavior since 1980 that differences in rates of registration and voting between men and women are modest and not statistically significant after controlling for traditional predictors of participation. However, we argue that researchers have overlooked the substantial gender gap in registration and voting in the South. While the gender gap in participation virtually disappeared outside the South by the 1950s, substantial gender differences in rates of voter registration and turnout remained in the South throughout the 1950s and 1960s. We test several explanations for the persistence of the gender gap in registration and voting in the South in the 1950s and 1960s and why it began to decline in the 1970s. These explanations include female labor force participation, resources, mobilization, and political engagement. Using American National Election Studies data for every presidential election year from 1956 to 1980, we employ heteroscedastic probit models within a cross-classified multilevel age-period-cohort framework to examine the declining gender gap in voter registration and turnout in the South. The results indicate that the decline of the gender gap is due to converging rates of political engagement and employment for women and men in the South during this time period. We conclude with a discussion of the theoretical implications.

For and against European Union expansion: Examining mixed opinion on enlargement and specific countries’ entries

European Union (EU) residents hold complex attitudes toward expanding the union not completely captured in extant models of enlargement opinion. We develop two-by-two matrices of support and opposition toward (1) enlargement and (2) a potential member’s entry. Descriptive analyses of 2006 Eurobarometer data reveal that, across two-by-two matrices for enlargement and 13 potential members’ entries, EU-25 residents often support enlargement and a potential member’s entry or oppose both. Yet, what we call ‘qualified support of enlargement’ – support for enlargement but opposition to a potential member’s entry – varies by potential and current members. Synthesizing literature on similar attitudinal combinations with research on enlargement opinion, we develop hypotheses to account for the individual- and country-level variation in qualified support of enlargement, relative to supporting enlargement and a potential member’s entry or opposing both. According to our multilevel multinomial logit results, individual-level factors such as education, political awareness, and political ideology are generally related to qualified support of enlargement; economic distance between current and potential EU members is related to the levels of qualified support of enlargement in member countries. However, attitudes toward enlargement and the entries of countries, such as Turkey, Albania, and Iceland are exceptions to some patterns. Our findings highlight the multidimensionality of EU enlargement attitudes and suggest that literature on similar attitudinal combinations should incorporate group dynamics into their models.

Spatial Agglomeration and Wages in the U.S. Biotechnology Sector

Recent studies show that firms and workers benefit from the geographic concentration or spatial agglomeration of employment in a local labor market. While economic geographers have devoted considerable attention to the benefits firms receive from spatial agglomeration, the reasons workers may benefit in the form of higher wages are less clear. I draw from several theoretical perspectives in order to explain the relationship between spatial agglomeration and wages for U.S. biotech workers, including the new economic geography, new structuralism, organizational ecology, and the new economic sociology. Using the 1990 PUMS-L, I employ a hierarchical linear model in order to test hypotheses from these perspectives. The results indicate that biotech workers benefit more from industrial than urban agglomeration, and these benefits are primarily due to the greater degree of organizational diversity in large industry clusters and urban labor markets.

Who Is a Part-Time Worker Around the World and Why Does It Matter? Examining the Quality of Employment Measures and Workers’ Perceived Job Quality

ABSTRACT Theory and research do not fully account for the cross-national variation in part-time work definitions and measures, which may affect conclusions. Using the 2004 to 2009 International Social Survey Programme (ISSP) as a special case, this study analyzes the reliability of different part-time work measures for 47 countries, finding high levels of consistency between what resembles a country-specific measure and measures based on 30- and 35-hr thresholds. Bivariate analyses reveal that sex is consistently related to—and suggestive of the construct validity of—part-time measures, except in some non–Organisation for Economic Co-operation and Development (OECD) countries. Analyses of selected countries in 1 year indicate that the choice of part-time work measure matters for understanding workers’ perceptions of their earnings, advancement opportunities, and job security in a few countries. Our results yield good news for ISSP users but highlight the need for the more exhaustive conceptualization of part-time work we offer, research outside of the OECD, and care in interpreting data in some countries.

Staying Close or Going Away: How Distance to College Impacts the Educational Attainment and Academic Performance of First-generation College Students

It is widely documented that first-generation college students attain bachelor’s degrees at lower rates than their peers. First-generation students also consistently prioritize distance to college in their school decision-making process. How distance impacts their educational performance, however, is an issue that has not received sufficient research attention. This study uses the Beginning Postsecondary Students Longitudinal Study (BPS:04/09) to investigate whether the distance between the permanent residence of first-generation students enrolled in four-year degree programs and their attending college impacts their educational attainment and grade point average (GPA). We find that first-generation students who attend colleges at a greater distance from home are more likely to graduate from college with a bachelor’s degree. We do not find strong support for the relationship between distance and a student’s GPA in most years of enrollment. We discuss the way college accessibility reinforces inequality within higher education along with the theoretical implications of our findings.

Constrained and Unconstrained Partial Adjacent Category Logit Models for Ordinal Response Variables.

Adjacent category logit models are ordered regression models that focus on comparisons of adjacent categories. These models are particularly useful for ordinal response variables with categories that are of substantive interest. In this article, we consider unconstrained and constrained versions of the partial adjacent category logit model, which is an extension of the traditional model that relaxes the proportional odds assumption for a subset of independent variables. In the unconstrained partial model, the variables without proportional odds have coefficients that freely vary across cutpoint equations, whereas in the constrained partial model two or more of these variables have coefficients that vary by common factors. We improve upon an earlier formulation of the constrained partial adjacent category model by introducing a new estimation method and conceptual justification for the model. Additionally, we discuss the connections between partial adjacent category models and other models within the adjacent approach, including stereotype logit and multinomial logit. We show that the constrained and unconstrained partial models differ only in terms of the number of dimensions required to describe the effects of variables with nonproportional odds. Finally, we illustrate the partial adjacent category logit models with empirical examples using data from the international social survey program and the general social survey.

Job Insecurity and Substance Use in the United States: Stress, Strain, and the Gendering of Precarious Employment

Abstract Research on the social determinants of health demonstrates that workers who feel insecure in their jobs suffer poorer health as a result. However, relatively few studies have examined the relationship between job insecurity and illegal substance use, which is closely related to health. In this study, we develop a theoretical model focusing on two intervening mechanisms: health and life satisfaction. Additionally, we examine differences in this relationship between women and men. We test this model using logistic regression models of substance use for women and men based on longitudinal data from the National Survey of Midlife Development in the United States. The results indicate that job insecurity is associated with a significantly higher probability of illegal substance use among women but not men. We interpret this as further evidence of the gendering of precarious employment. This relationship is not channeled through health or life satisfaction, but there is evidence that job insecurity has a stronger association with illegal substance use for women with poorer overall health.