Socioeconomic Status

Socioeconomic implies at least two dimensions to inequality—social and economic. Although these two dimensions are understood as closely associated, they nevertheless incorporate two different aspects of stratification. The economic dimension is best represented by money or wealth as reflected in employment income, home ownership, and other financial assets (e.g., pension plans, property ownership). The social dimension incorporates education, occupational prestige, authority, and community standing.

The very earliest measures of socioeconomic status in North America relied on community reputation. A family's social standing as judged by others was used to differentiate between upper, middle, and lower classes. Although the term class was used, this was a very North American usage that understood classes as loose aggregates of families who shared similar social and economic traits. However, this early measurement tradition rested mainly in community studies. As social scientists started to focus more on entire societies, a different measurement technique was essential.

In 1947, Cecil North and Paul Hatt conducted a study in the United States in which they asked people to judge the prestige of different occupations. This study marked a watershed in the measurement of socioeconomic status. Prestige studies typically ask respondents to judge the social standing of about one hundred occupations. However, working independently, Bernard Blishen in Canada and Otis Dudley Duncan in the United States devised a way to combine the prestige scores of occupations with the typical incomes and educations of occupational incumbents. For example, Duncan's Socio-economic Index (SEI) was constructed by weighting an occupation's median education and income on the metric of occupational prestige (via a regression equation that can be simplified as follows: Prestige = a + B1[Income] + B2[Education] where each variable represents an occupational average). SEI scores were developed for all of the major occupations and allowed researchers to assign a person an SEI score based on one variable, occupation.

In the United States, prestige studies done in 1947, 1963, 1971, and 1989 have been used to generate SEI scores. Although the hierarchical placement of a few specific occupations has changed over time, the relative placement of most occupations is stable. This stability in the prestige hierarchy has meant that specific scales of socioeconomic status can be used with confidence long after they are first constructed.

More recently, scholars of inequality in North America, in particular, have moved away from single scales of socioeconomic status to amalgam measures. Rather than relying on a summary SEI score, contemporary researchers are often asking a set of questions related to socioeconomic status (SES). Thus, for example, many researchers now measure SES by combining (often through factor analysis or some analogous statistical method) measures of at least three of the following: an individual's education, earnings, home ownership, occupation, and net worth. In measuring the SES of a family, a frequent approach is to combine the education, earnings, and occupation of wives and husbands or communal partners (sometimes along with home ownership or net value of a family home).

However, researchers at times want to examine the relative effects of the separate components of SES. Therefore, measures of education and income (for example) are sometimes used separately and are not combined in a scale or index. Important information may be lost in combining education, income, occupation, and residential status. Summative family scales are also not appropriate when scholars seek to compare the relative influence of the SES of spouses or partners on, for example, the educational attainment of their children or the health status of family members.

The old assumption of the male breadwinner, whatever its historical validity, is highly problematic. Family forms have changed (e.g., single parents, gay and lesbian couples). Women's labor force participation and career commitment increased dramatically in the last decades of the twentieth century. The contributions of partners who are not in the paid labor force have been increasingly recognized. For all of these reasons, the use of a family SES measure based on information about a single family member is sometimes inadequate. However, it is also important not to exaggerate the force of this claim because the SES of spouses and partners are often similar. In the British sociological literature, there is a long-running debate on this very issue, although the referent is more often to social class than to SES (see Goldthrorpe 1983).

Individuals or families have their own SES, but they also live and work in contexts that may be defined by different levels of SES. In this sense, the SES context in which a person finds him or herself may be more or less powerful than his or her own individual SES when it comes to predicting outcomes (e.g., job satisfaction, psychological stress). This is often understood as the ecological setting or context of people or groups. For example, do children in school classrooms where the SES of their classmates is greater than their own do better academically than would be predicted simply from their own families' SES?

Today, there is no consensus upon exactly how SES should be measured. In modern research, the following considerations are important in devising one or more indicators to measure SES. First, is SES the relevant conceptual approach to inequality? Second, if SES is a core variable, how many indicators should be used, and should these be combined in a scale? Third, what is the validity and reliability of SES measures in comparison to alternatives? Fourth, will measures of SES provide the necessary comparability with other research studies in the area? Fifth, is SES applicable to all members of the population being studied? Especially in this last case, the SES of students, the unemployed, recent migrants, and the retired may be problematic.

When analyzing data, different scholars may or may not treat SES as an ordinal (i.e., ranked [beauty]) or interval (i.e., equal distance between categories [age]) measure. Often, when an ordinal measurement preference is chosen, SES is collapsed into groups, frequently with labels like upper class, middle class, and lower class. The boundaries between these groups are typically relatively arbitrary, there being no natural or theoretical cutting point in deciding at exactly what SES score the boundary should be drawn. Often, for this reason, others choose to assume SES has interval measurement properties, and they use more sophisticated statistical techniques.

In contrast to Europe where the idea of social class was more influential, in North America, seen as the land of opportunity and upward mobility, social status with its hierarchical stress was more prevalent. Among the more important studies in demonstrating the utility of socioeconomic status was the work of Otis Dudley Duncan and Peter Blau on social mobility. Using SEI scores as their basic measure, they were concerned with, among other things, the relative chances of upward mobility across generations for whites and blacks in the United States. More recently this tradition has prospered in North America and elsewhere in the guise of status attainment models, addressing a variety of research questions involving who achieves status and how they do it.

Socioeconomic status has been shown to be significantly, consistently, and universally correlated with a variety of measures of life chances (e.g., occupational attainment), lifestyles (e.g., health status), sociopolitical orientations (e.g., ideological leanings), and modes of action and association (e.g., association memberships). Why these correlations exist remains one of the central questions continuing to be pursued. Simply put, what is it about SES that creates a causal effect and why does this effect occur? It is now well understood that although economic resources are important, it is not so clear exactly why (e.g., is the effect due to access to better nutrition, better information, or more powerful networks?). Similarly, education is an important SES measure, again correlated with many diverse outcomes, but its precise role is often unclear (i.e., is it the cognitive dimension, is it the credential, is it the network of contacts?).

Finally, it is worth noting that although socioeconomic status is most frequently found in the academic literature, it is a term increasingly employed in research outside the academy. For example, the marketing firm A. C. Nielsen (which measures television audience share around the world) uses socioeconomic status as a core measure for differentiating types of viewers. Other marketing firms do likewise in reporting on voter preferences or consumer product choices.