Robert Andersen

The Political Conditioning of Economic Perceptions

Economic theories of voting argue that party popularity and vote are heavily influenced by the performance of the economy. Inferences about the direction of causality between perceptions of the economy and party support remain questionable, however. This article evaluates the microfoundations of economic theories of voting and party popularity using multiwave panel data. We model the dynamic relationships between party support and retrospective economic perceptions—both sociotropic and egocentric—through the 1992–97 British electoral cycle. Our findings indicate that sociotropic perceptions are strongly conditioned by prior opinions of the incumbent Conservative Party and once this temporal relationship is taken into account, they have little affect on incumbent party popularity. Throughout the electoral cycle, lagged political support has a substantially stronger effect on sociotropic perceptions than the latter have on concurrent party support. Moreover, egocentric perceptions appear to be neither strongly affected nor influenced by party support. The findings of these dynamic individual-level analyses indicate that conventional wisdom is likely to considerably overstate the importance of retrospective economic considerations for political preferences.

Political Knowledge and Enlightened Preferences: Party Choice Through the Electoral Cycle

This article adapts and tests the theory of enlightened preferences on two British electoral cycles: 1992–97 and 1997–2001. Using individual-level panel data, it extends previous work by explicitly incorporating the role of political knowledge. Its findings are generally very supportive of the theory. It is shown that knowledge of party platforms varies through both electoral cycles in a manner predicted by the theory; that is, it is highest immediately following election campaigns; these changes in political knowledge are closely mirrored by changes in the explanatory power of a model of party choice containing so-called ‘fundamental variables’ (i.e. socio-demographic and issue-related variables) as predictors. More specifically, fundamental variables do a much better job of accounting for party choice during election years than in mid-cycle. Finally, for all years of both panels a positive interaction is found between political knowledge and the ability of voters to match their issue preferences to party platforms.

Trends in Civic Association Activity in Four Democracies: The Special Case of Women in the United States

This study assesses whether civic association activity has declined in four Western democracies: Canada, the Netherlands, the United Kingdom, and the United States. Influential accounts of decreasing civic engagement in the United States lead to the expectation of similar patterns in the other three nations. The authors test this hypothesis using data from time-use surveys of adult national samples for the mid-1960s to the late 1990s. One major finding is a clear decline in association activity in the United States, especially after 1975, but relative stability in the other three countries. Equally important are further results indicating that the American decline pertains only to women. Findings are sustained even after controlling for social background characteristics and four other activities (television watching, paid work, childcare, and physical activity). The analysis casts doubt on the theory that declining civic association activity in the United States reflects generational differences. Possible explanations for the reduced activity among American women, including lower levels of state support, are considered.

Social Background, Credential Inflation and Educational Strategies

The primary goal of this article is to examine the impact of credential inflation on educational attainment in twentieth-century United States. To do so, we create a measure of ‘intergenerational credential inflation’ (intergeneration inflation factor) and include it in regression models predicting educational transitions. Using the General Social Surveys of 1972–2000, we find that people are generally less likely to invest in schooling if its value is relatively low. An exception is the final transition to a postgraduate degree, where we find that when its value is low children of parents with postgraduate education are more likely to take it. This finding supports relative risk aversion theory, which assumes that the main goal of children is to avoid downward social class mobility. Perhaps most important, we find that credential inflation is particularly influential on transition probabilities if parents had made the same transition. This pattern is consistent with the information differential thesis that children are more informed about the value of education if their parents acquired it.

Education and occupational status in 14 countries: the role of educational institutions and labour market coordination

This article explores the role of national institutional factors--more specifically, the level of skill transparency of the education system and labour market coordination--in accounting for cross-national differences in the relationship between education and occupational status. Consistent with previous research, our findings suggest that skill transparency is the primary moderator. Countries with a highly transparent educational system (i.e., extensive tracking, strong vocational orientation, limited tertiary enrolment) tend to be characterized by a strong relationship between education and occupational status. These findings hold even after controlling for the level of labour market coordination. Nevertheless, we also find that labour market coordination plays an independent role by dampening the effect of education on occupational status. Taken together, these results suggest two quite different policy implications: (1) strengthening the skill transparency of the education system by increasing secondary and tertiary-level differentiation may strengthen the relationship between education and occupation, regardless of the level of coordination, and (2) increasing labour market coordination could lead to improved social inclusion and a reduction in inequalities related to educational attainment.

EFFECT DISPLAYS FOR MULTINOMIAL AND PROPORTIONAL-ODDS LOGIT MODELS

An “effect display” is a graphical or tabular summary of a statistical model based on high-order terms in the model. Effect displays have previously been defined by Fox (1987, 2003) for generalized linear models (including linear models). Such displays are especially compelling for complicated models—for example, those including interactions or polynomial terms. This paper extends effect displays to models commonly used for polytomous categorical response variables: the multinomial logit model and the proportional-odds logit model. Determining point estimates of effects for these models is a straightforward extension of results for the generalized linear model. Estimating sampling variation for effects on the probability scale in the multinomial and proportional-odds logit models is more challenging, however, and we use the delta method to derive approximate standard errors. Finally, we provide software for effect displays in the R statistical computing environment.

Social identities and political cleavages: the role of political context

Using a novel method, the paper investigates the influence of social group identities on attitudes and on voting in a variety of political contexts. Examining the major regions of Britain, Canada and the USA, we find considerable national and regional diversity in the nature of social cleavages. For example, social class and race had widely different effects across societies, but within societies their effects on attitudes and on voting were very similar. However, despite that, age and religion had a similar effect on attitudes across societies; the effects on voting varied considerably. The significant within-country differences underline the importance of using region, rather than country, as the unit of analysis. More importantly, these results highlight the role of political context, especially competing cleavages and the structure of party competition, in the establishment of politically relevant social cleavages. Copyright 2003 Royal Statistical Society.

By Popular Demand: The Effect of Public Opinion on Income Inequality

This paper examines the relationship between public support for wage differentials and actual income inequality using data from the World Values Surveys. The distribution of income is more equal in nations where public opinion is more egalitarian. There is some evidence that the opinions of people with higher incomes are more influential than those of people with low incomes. Although the estimated relationship is stronger in democracies, it is present even under non-democratic governments, and the hypothesis that effects are equal cannot be rejected. We consider the possibility of reciprocal causation by means of an instrumental variables analysis, which yields no evidence that income distribution affects opinion.

Social cleavages, electoral reform and party choice: Israel's 'natural' experiment

Abstract This paper examines party preferences in Israel from 1993 to 1999. This period provides a unique ‘natural’ experiment in that the middle was marked by a change from a single-ballot party list to a two-ballot system that included direct election of the Prime Minister. The paper has two goals: (1) to explore the patterns of social cleavage voting in Israel during this period; and (2) to determine whether the change in the electoral system in 1996 affected these patterns. Contrary to the findings of previous research — most of which used ad hoc measures of social class — our results indicate that party choice was significantly related to social class. Moreover, the effects of ethnicity, class and religiosity on party preferences were strong and fairly constant throughout the period, suggesting that the change in electoral system had little impact on the party/cleavage relationship.

Bridging and Bonding: Ethnic Background and Voluntary Association Activity in Canada

This paper is the first Canadian analysis to consider voluntary association activity among a large array of ethnic groups, using national sample survey data and applying multivariate controls. Findings from the Economy, Security, and Community Survey of 2000 indicate that most of Canada’s major ethnic or racial minority groups differ little in their average number of voluntary association involvements. The main exceptions are relatively low levels of association membership for Latin/Central/South Americans and East Asians, especially before controls, and quite high levels for the Jewish minority. Further analyses reveal that minorities are more likely than other groups to join associations related to their own particular ethnic group and that this partly accounts for the relatively small differences in overall voluntary activity between ethnic minorities, on the one hand, and non-minorities, such as North/West European and British respondents, on the other hand. The implications of these results for understanding the “bridging” and “bonding” roles of voluntary associations in Canadian society are briefly discussed. Cet article présente la première analyse canadienne sur les activités d’associations de bénévolat au sein d’un grand nombre de groupes ethniques, à partir d’une enquête de 2000 sur l’économie, la sécurité et les communautés, portant sur un échantillon national et y appliquant des contrôles multivariables. Selon les résultats obtenus, au Canada l’engagement dans ces associations n’est en moyenne guère différent pour la plupart des groupes, qu’ils appartiennent à des majorités ethniques ou à des minorités raciales. Les principales exceptions sont soit des niveaux relativement bas de membres, quand ils viennent d’Amérique latine / centrale / du Sud et d’Asie de l’Est, en particulier avant les contrôles, soit des niveaux nettement plus élevés dans la minorité juive. Des analyses plus poussées révèlent que dans les minorités, on se joindra plus qu’ailleurs à des associations reliées à son propre groupe ethnique, d’où en partie les différences, somme toute mineures, dans les acti-vités de bénévolat en général entre les deux : minorités ethniques d’une part, et non-minorités d’autre part, telles qu’Européens du Nord et de l’Ouest, ainsi que britanniques. Nous examinons brièvement ce qu’impliquent ces résultats qui permettent de comprendre les rôles de “relais” et de “liaison” que jouent les associations bénévoles dans la société canadienne.