SNAPSHOTS OF SOCIAL CHANGE AUGUST 29, 2006:
BIRTH YEAR AND BELIEFS
Since 1972, the General Social Survey (Davis et al. 2005) has tracked U.S. public opinion through a series of annual or biennial polls, and made the data available for research. Each poll contacts 1,500 or more people. As measured through these polls, opinions on some controversial issues have remained surprisingly stable over the years, but others have shown definite trends. For example, the percentage of respondents who believe that same-sex sexual relations are “always wrong” has declined since the 1980s, from around 85% to less than 60%. Figure 1 graphs this trend in surveys from 1973 to 2004. Error bars indicate the uncertainty each year in generalizing from the survey to a larger population.
Figure 1
The statistically significant decline in Figure 1 represents real social change, but not necessarily just a simple process of individuals changing their opinions. The change also reflects a sociological dynamic called cohort replacement: there exist differences between generations of people born in different years or decades, which can cause social change as younger generations rise and gradually supplant the older. We see evidence of such differences if we separate the opinions of people born after World War II from those born earlier, within each year of the General Social Survey, as done in Figure 2. The percentages responding that same-sex relations are “always wrong” follow somewhat similar paths for both groups, but among respondents from postwar generations these percentages are consistently lower.
Figure 2
In his study of cohort differences in socioeconomic progress, Myers (2004) classified the adult U.S. population into a set of 10-year birth cohorts. He found substantial differences in socioeconomic mobility among these cohorts, as the shifting historical context affected each generation’s economic and educational opportunities.
Birth years |
Cohort |
Defining events during coming-of-age years |
1906–15 |
Grandparents of Baby Boom |
The Great Depression |
1916–25 |
Parents of Baby Boom |
World War II |
1926–35 |
Parents of Baby Boom |
Postwar economic boom |
1936–45 |
World War II |
Postwar economic boom |
1946–55 |
Early Baby Boom |
Civil rights and sexual revolutions |
1956–65 |
Late Baby Boom |
Employment restructuring |
1966–75 |
Generation X |
Economic polarization, AIDS/HIV |
1976–85 |
Generation Y |
Era of digital dominance |
If we apply Myers’ cohort scheme to the GSS respondents, we see a pattern in their beliefs about same-sex relations (Figure 3). More recently-born generations express less disapproval. A red line at 69% in Figure 3 marks the overall rate for all the General Social Surveys — a large majority, but the postwar cohorts fall progressively farther below this. As time passes and more recently-born cohorts make up a larger fraction of the total population, overall disapproval could decline without even one person changing his or her mind (although of course, many do).
Figure 3
Beliefs regarding same-sex relations vary from place to place. For example, disapproval tends to be higher in the South than in other regions, and higher in rural areas than in cities. The cohort effect or relationship between birth cohort and beliefs about same-sex relations appears remarkably stable, however. Figure 4 shows that a similar pattern exists both in city/suburbs and outside them, within each U.S. region. Regional differences become visible through the graphical details. In Northeastern cities, only the earliest cohorts fall above 69% disapproval, and later cohorts drop off steeply. In the rural South, all but the latest cohort fall above 69%, and the cohort effect — that is, the difference in beliefs across cohorts — is noticeably less steep.
Figure 4
A More Technical Note
Statistically-minded readers might wonder how respondent age, birth cohort, and year of the survey perform if we enter all three as predictors in one model. Is birth cohort the strongest predictor, even after we control for age and year? This seemingly straightforward question proves intractable due to multicollinearity — less than 3% of the variance of birth cohort is independent of both age and year (because if we know a person’s age and what year it is, we can guess their birth cohort). When birth cohort is entered together with either age or year, however, cohort alone is a significant predictor of belief about same-sex relations. Although cohort replacement might not be the only process driving change over time, this evidence suggests it is most important.
References
Davis, J.A. T.W. Smith, and P.V. Marsden. 2005. General Social Surveys, 1972–2004 Cumulative File [computer data file]. Chicago: National Opinion Research Center [producer]. Ann Arbor, MI: Inter-University Consortium for Political and Social Research [distributor].
Myers, D. 2004. Cohorts and Socioeconomic Progress. New York: Russell Sage Foundation.
Author
Snapshots of Social Change is written by Lawrence Hamilton, a Senior Fellow with the Carsey Institute and Professor of Sociology at the University of New Hampshire ( http://pubpages.unh.edu/~lch ).