Current Sociology

While ongoing debates in sociology exist over whether the family or the individual is the correct conceptual `unit' for stratification research, both sides to this debate have generally relied on individual-level occupation to operationalize social position. However, because industrial nations differ in the extent of labour force participation by women, stability of working hours, the stability of households and in state tax and transfer policies, comparative analysis of occupational mobility may not accurately describe cross-national differences in household mobility in living standards over the life course. Studies of standard of living mobility in the USA, Germany and Sweden show greater similarity between the countries than do studies of male occupational mobility.

Following the seminal work of Esping-Andersen, many studies have identified a variety of welfare regimes in Western Europe and North America. This study examines a set of quantitative social indicators, using hierarchical cluster analysis, in order to identify such regimes, which display specific arrangements between markets, the state and families in the production and distribution of the resources required for the well-being of people. Indeed, these empirical analyses reveal the existence of the three regimes originally identified by Esping-Andersen - social-democratic, liberal, and conservative - to which one must add, as many authors had pointed out, a fourth, distinct from the latter, the Latin regime. These results pertain whether one turns to data from the 1980s or the 1990s. The data also reveal strong and durable relations of presumably mutual causality between the configuration of social programmes in the various societies under analysis, the social situations which largely result from these social programmes and, lastly, the level of civic participation which leads (or not) people to collective mobilization, which in turn shapes social programmes. The authors' comparative analysis allows them to identify Canada's place in the worlds of welfare capitalism.

This article explains how large-scale random-assignment social experiments such as the Canadian Self-Sufficiency Project can address important sociological and developmental issues. The authors begin with an explanation of how random assignment solves the bias problem that plagues most research based on sample surveys. They then review methods and results from two recent sets of random-assignment experiments, one manipulating family economic conditions and the other manipulating neighborhood conditions. The authors conclude with a discussion of the strengths and weaknesses of the experimental approach.

The fundamental point of this article is that methods change what we are able to see. Using a series of examples, the article suggests that substantive literatures are regularly set off course by the limitations built into widely used methodological frameworks to analyse data. In each example, simpler and/or better known methods are compared to methods which either incorporate a broader range of possible influences or clarify the issue at hand, and consequently essential findings change. These findings suggest either that the inertial state of findings in research literatures could be redefined or redirected by the application of methods which take into account more clearly the effects of time and place, or that given theories may be transformed by methods which resolve conflicting or limiting features of theoretical debate. Examples cover a wide range of issues, from longitudinal vs cross-sectional data, to the specification of models to replace equations, to the influence of social context on individual behaviour, to the importance of time in the modelling of events, to the nuances of capturing the complexity behind interdependent processes over the life course.