The idea of “generations” is a fruitful if vexing one, attracting a vast array of scholars – most notably in sociology and political science – whose writings cover conceptual, theoretical, empirical, and methodological grounds. The topic also draws journalists in hordes, with innumerable pieces written about this or that generation.1 This essay will survey some of this terrain, highlighting the promise as well as the pitfalls of using the lens of generations to study politics.
The Concept(s) of “Generation”
To start, there are two generation concepts, not one, that are of interest to political scientists (Braungart and Braungart 1986; Alwin and McCammon 2003; Biggs 2007). Alwin and McCammon (2003) distinguish them by calling them “generation” and “Generation,” whereas I will use the terms “lineage generation” and “generation” (or “political generation”), respectively. The former is a genealogical concept: “a term denoting kinship relations, a generation consists of a single stage or degree in the natural line of descent” (Alwin and McCammon 2003, p. 25). Grandparents, parents, and children are, in this sense, distinct generations. Much of the work in political socialization studies lineage generations, exemplified by the landmark books The Political Character of Adolescence (1974) and Generations and Politics (1981) by M. Kent Jennings and Richard G Niemi. The well-known study on which those books are based was initiated by M. Kent Jennings in 1965 and continued with data collections in 1973, 1982, and 1997, yielding data on three lineage generations across a span of 32 years.
The concept of “generation” in the second sense refers to a group that is distinctive in any number of respects by virtue of having experienced a specific set of social, economic, technological, and/or political circumstances at a formative period in their lives. The classic citation for this concept of generation is Karl Mannheim’s 1926 essay “The Sociological Problem of Generations,” although the idea has a longer pedigree (Braungart and Braungart 1986; Delli Carpini 1989). In this usage, we would call the group a political generation if it is distinctive politically. This variant of the generation concept rests on a two-part argument about how people develop their political and other orientations to the world.
The first claim is that there is a period in people’s lives that is especially formative, a period during which basic beliefs, attitudes, identities, habits, and other predispositions tend to develop and crystallize. This formative period is commonly called the “impressionable years” (IYs) by political socialization researchers and is thought to range from adolescence through early adulthood. Scholars have studied the possibility that there is no such formative period, that people are instead always open to change as they move through their lives. The weight of the theoretical expectations, as well as the evidence, suggests otherwise (Sears 1983, 1990; Sears and Brown 2013). Political identities and attitudes are quite unstable from adolescence to early adulthood but stabilize subsequently (e.g., Jennings and Markus 1984; Alwin, Cohen, and Newcomb 1991; Sears and Funk 1999). Political engagement or disengagement during the impressionable years builds habits that tend to be followed later in life (Plutzer 2002). Socio-political events are more likely to be remembered and judged important (Schuman and Rodgers 2004), to be cited as influential to one’s political development (Jennings and Stoker 2006), and to provoke attitude change (Campbell 2002; Stoker and Jennings 2008; Dinas 2013) if experienced in young adulthood.
The second part of the argument is that people can be grouped into generations on the basis of the historical period during which they went through their impressionable years, with or without also considering the person’s “social location” at the time. In its simplest form, the idea is that the classification of generations is based merely on the intersection of age and history. One belongs to a given generation if one’s IYs occurred during an historical period the researcher singles out – the 1930s or Great Depression, post-911, the years of Ronald Reagan’s presidency, and so on. Some feature or set of features of the historical period is thought to be responsible for the generation’s distinctiveness.
As Mannheim (1952 ) emphasized, however, how a young person is influenced by that historical milieu will often depend on the particulars of the person’s circumstances, which he summarized using the term “social location.” This highlights the possibility that there can be different generations – or generation units, as Mannheim called them – within the group of citizens experiencing their IYs at any specific moment or period in history. Thus, for example, experiencing the Civil Rights movement during one’s IYs would have different consequences for African Americans and Whites, yielding two generation units within the cohort instead of a single generation.
It is sometimes argued that to be worthy of the “generation” label, the group members must also be aware of their distinctiveness and identify themselves in generational terms. This consciousness was a feature of Mannheim’s (1952 ) theorizing about generations. It is what distinguishes a “cohort” from a “generation” according to others (Braungart and Braungart 1986; Alwin and McCammon 2003). As Braungart and Braungart (1986, p. 217) put it, “A cohort becomes transformed into a political generation when many of its members become aware that they are bound together by a shared age-group consciousness and mobilize as an active force for political change.” This idea has tended to be absent from work on political generations by political scientists, but is common in the work of sociologists studying social movements (see Braungart and Braungart 1986; Alwin and McCammon 2003).
A large and varied literature has sought to understand the roots of political behavior in generational terms. For example, African Americans going through their impressionable years during the Warren Court era still held highly favorable attitudes toward the Supreme Court many years later, more so than did those who came of age earlier or later (Gibson and Caldeira 1992). College students who joined civil rights and anti-war protests during the 1960s and early 1970s emerged from that experience with a set of left-leaning political attitudes and participatory tendencies that continued to distinguish them as they aged (Jennings 1987, 2002). Young adults today are more likely to form partisan attitudes tied to ideology than those socialized in previous periods, and to bring their party and ideology into even closer alignment as they age, both of which reflect the growing polarization of party elites (Stoker and Jennings 2008). The partisanship of young people is also more tied to religion and religiosity (Campbell 2002). Because youth tend to develop voting or non-voting habits in early adulthood, high-stimulus elections leave a “footprint” in the form of higher turnout subsequently (Franklin, Lyons, and Marsh 2004). The party identification and voting choices of young adults are typically most responsive to prevailing national tides. Young voters were more Democratic during the New Deal era, more Republican during the Reagan era, and lined up strongly behind Obama in 2008 (Green, Palmquist, and Schickler 2002; Dalton 2008).
As these examples illustrate, the ways in which generations can be thought to be distinctive are varied. Most common is the claim that generations differ in their modal beliefs, attitudes, and/or behavior – differences that can be attributed to events and experiences in each generation’s IYs. Also common is the idea that the relationships among variables vary across generations. Although argued less frequently, generations can also be thought to differ in how they develop across the life cycle and in how they react to events that occur well past the IY life stage.
Generations and Macro-Political Stability and Change
Both generational concepts become especially fruitful once considered along with the passage of time. With lineage generations, a key issue is whether macro-political stability is sustained by the intergenerational transmission of values and political orientations. The traditional perspective on this matter holds that social learning and the dynamics of social influence within families prompt children to acquire political orientations that resemble those of the parents – in a relative if not an absolute sense (more on that below). Jennings, Stoker, and Bowers (2009) offer a recent take on the matter, while reviews of the socialization literature provide a broader perspective (Niemi and Sobieszek 1977; Niemi and Hepburn 1995; Sears and Brown 2013; Sapiro 2004). Studies challenging that traditional perspective hold that parent-child similarity is at least partially genetic in origin (for reviews, see Carmen 2007; Alford and Hibbing 2008; Funk 2013; Hatemi and McDermott 2012). Either way, the processes that cause children to resemble their parents, politically, have a major bearing on how the polity changes (or does not change) as the filial generation comes to “replace” the parental generation every 30 or so years – to step into the roles that the parents occupied some 30 years before.
Work on lineage generations has also been directed toward mapping and explaining the familial reproduction of socio-economic stratification (e.g., Buchmann and Hannum 2001; McLanahan and Percheski 2008) and political inequality (e.g., Verba, Burns, and Schlozman 2003; Jennings and Stoker 2014) over time. Jennings and Stoker (2014), for example, demonstrate that parents who are at best episodically involved in politics while their children are growing up tend to produce children who are at best episodically involved in politics while their children are growing up, who then go on to produce children who……
The lineage generation framework is also capable of providing a fruitful lens for thinking about cycles of macro-level change. A rare example of this is Beck’s (1974) socialization theory of party realignment. Beck argued that partisan attachments, imperfectly transmitted from parent to child, would tend to weaken lineage generation after lineage generation, leaving America ripe for partisan realignment every third generation or so – a pattern consistent with historical evidence on the timing of partisan realignments in the US.
Far more attention has been given to studying the over-time implications of generation formation in the Mannhemian sense. The thinking here ties ideas about the formation of (political) generations to the ongoing process of population replacement. The logic runs like this:
At any starting point in time, the population can be characterized in terms of its generational composition. For example, 60% of the population could be in Generation A and 40% could be in Generation B.
As time passes, new members enter the polity and others exit through death, which produces slow-moving shifts in the generational composition of the population. Generation B, for example, could grow from 40% to 100%, while Generation A dwindles from 60% to 0%. With the passage of time, however, there is always the possibility of a new generation forming. Thus, while Generation A is dwindling from 60% to 0%, Generation B could be holding steady at 40% while a new generation, Generation C, is growing from 0% to 60%.
Finally, the changing composition of the population in generation terms will bring about aggregate-level change on any attribute in which the generations are distinctive. The population is slowly but inexorably losing people who have one trait and replacing them with people who have a different trait. With time, the character of the polity can change markedly. When macro-political changes are wrought by shifts in the generational composition of the population they are said to be the result of “generation effects” or “generation replacement effects.”2
As an example, consider the case of women categorized into one of two generations on the basis of whether their IYs came before or after the passage of the 19th amendment to the U.S. Constitution, which extended the right to vote to women. One group, the “pre-19th amendment generation” of women would have been unlikely to develop the habit of voting because their political predispositions were formed before the right to vote was extended to them. They should have been significantly less likely to vote at every point in their lives than their similarly situated counterparts coming of age later, the “post-19th generation.”
By extension, as the percentage of women in the pre-19th amendment generation dwindled from 100% in the early 1900s to nearly 0% by the late 1980s (and the percentage of women in the post-19th amendment generation grew from 0% to 100%), the turnout of women should have climbed. Indeed, survey research working with data from the 1950s onward shows steady increases in women’s voting rates, with the gap between the turnout of women and men closing in the 1980s, patterns fueled by generational replacement (Firebaugh and Chen 1995; Dinas and Stoker 2014).
Each variant of the generation concept is, thus, a multi-level concept (Weatherford 1992), i.e., helpful in organizing research on both individual-level and macro-level phenomena. The focus on parent-child transmission of work on lineage generations helps us understand the origins of political predispositions among individual citizens as well as how social and political stability and stratification is sustained within a society over time. The work on political generations, in calling attention to the socio-political circumstances encountered during the impressionable years of young adulthood, helps us understand the political character of any one generation and how society evolves as population replacement carries on.
Putting Lineage Generations and Political Generations Together
A scattering of research has considered lineage generations and political generations together. A first insight is that family transmission dynamics can produce relative continuity across lineage generation even when historical circumstances and events are leading the younger generation to break from the older generation in aggregate terms. Thus, for example, the 1965 high school seniors in the Jennings’ Political Socialization Study held more liberal attitudes on racial integration in the public schools than did their parents, as would be expected given the political circumstances typifying the two generations’ IYs. At the same time, liberal views were more common among those with liberal parents, just as conservative views were more common among those with conservative parents (Jennings, Stoker, and Bowers 2009). Jennings and Niemi’s Generations and Politics contains many examples of this dual focus on lineage and political generations.
A second insight is that intergenerational transmission dynamics may bear on who ends up being generationally distinctive in the Mannheimian sense. This is the central thesis of Dinas (2014), which argues that children from highly political families are most likely to mirror their parents’ views in childhood but also most likely to abandon them as they go through their IYs. In line with this, Jennings, Stoker, and Bowers (2009) found that 1965 high school seniors from apolitical homes were more similar to their parents on the issue of school integration than were those from politicized homes, precisely because the latter were more likely to embrace the new racially egalitarian norms. The classic Bennington Study found that the women studied were less likely to embrace liberal political views the more tied they were to their relatively conservative families and home towns (Alwin, Cohen, and Newcomb 1991).
A third way of thinking about the intersection of the two generations involves intergenerational conflict, rivalry, or misunderstanding. These are popular themes in the mass media3 and the focal point for some research on generations (e.g., Arber and Attias-Donfut 2000). This topic joins the two generational concepts because differences in political generation can be a source of conflict and misunderstanding between grandparents, parents, and children, a topic of some interest to sociologists and psychologists. At the same time, the bonds that tie family members to one another may work to diminish the distinctiveness of successive political generations (Biggs 2007).
Challenges of Studying Political Generations
Of the challenges of studying political generations and generational effects, two seem to be most troublesome. The first and most storied is the so-called “identification problem.” The second is the difficulty of understanding just what, about any historical period, is driving generational differences, a problem that is itself tied to how generations end up being identified. I take these up in turn.4
Identifying Generational Differences
The central problem with identifying generational differences is revealed by thinking through two ways that cohorts can be compared. In the first, two or more cohorts are compared at one point in time. In the second, they are compared at different points in time but at the same stage of the life cycle. To elaborate, it will be helpful to have some specific cohorts in mind. I will use the example of the Baby Boomers, often defined as those born between 1946 and 1964 (though sometimes that range is broken down into early boomers and late boomers); Generation X, born between 1965 and 1980; and the Millenials, born since 1981.
When comparing the three generations5 with cross-sectional data, the problem is that age and generation are fully confounded. Millenials in 2014 are aged 18–33 years, Gen Xers are aged 34–49 years, and Baby Boomers are aged 50–68 years. If the generations are distinctive on any attribute, we should see it in a cross-sectional comparison. However, any differences that exist could just as well be due to age or anything age-related. For example, a comparison of the three would show turnout to be lowest among Millenials, but that may have nothing to do with generations, since turnout tends to increase with age (Jennings 1979; Wolfinger and Rosenstone 1980). 6 Media commentators are particularly likely to blithely interpret age differences in generational terms, but the problem also slips into scholarly work.
It might seem more sensible to compare the generations at the same moment in the life cycle. For example, we could compare the voting turnout levels of the generations when they faced their first, second, and third presidential elections, or the civic engagement levels of the groups when they were in their 1920s. The problem with such comparisons is that year or time-period is fully confounded with generation. For the Boomers, we would be looking at whether they voted in the 1960s, whereas for the Gen Xers, the focus would be on the 1980s or 1990s, and for the Millenials it would be 2000 or later. Any differences that are evident holding life-stage constant could be differences attributable to time or period instead of generation.
Figure 1 presents hypothetical data to further illustrate the problem of distinguishing age, period, and cohort or generation effects. The figure shows a trend in some dependent variable of interest over a period of 24 years. Imagine that the dependent variable is the percentage of strong party identifiers in the US. The curves indicate the average trend across six cohorts, numbered from the oldest to the youngest. One can think of them as six of the many birth-cohort groupings that could have been represented.
First of all, the figure suggests that some event occurred between time 19 and time 20 that generated a short-term period effect – an abrupt decline in the level of partisanship among all cohorts, lasting only a year. Notice that the event is depicted as having a greater effect among the young. As mentioned earlier, the political orientations of young people are thought to be especially vulnerable to events taking place in the political environment due to their lower levels of political information and attitude crystallization.
Figure 1 also shows a general upward trend across time for each cohort. This could reflect the strengthening of partisanship that typically comes with age – i.e., an aging effect. When an aging effect is operating, some process affecting the outcome variable unfolds in a systematic fashion over the life cycle. Time or age is not, itself, the causal variable driving the change, but simply indexes the causal variable operating. With respect to strength of partisanship, the age-dependent causal variable could be experience with the political system (Converse 1976). Another interpretation of the over-time growth within each cohort, however, is that everyone is responding to period forces prompting strong partisanship. Growing polarization of party elites could be strengthening partisanship across the board.
Finally, Figure 1 also shows a gap between each cohort at each moment in time. These gaps could be due to aging effects or cohort effects. If the cause of the gap between cohorts illustrated in Figure 1 is age-related, then the gap should disappear if one compares cohorts at the same moment of the life-cycle (visually, imagine sliding the trend-lines for the younger cohorts leftward until they are overlaid to hold age constant). If a gap persists, this could be taken as evidence of cohort effects. However, period effects could also be responsible for any gap, since time is no longer controlled.
These problems have been exhaustively addressed in the literature on Age-Period-Cohort (APC) analysis. APC analysis was introduced in the 1950s as an approach for modeling population change (Markus 1983). To use the approach, one needs longitudinal data on an age-varied sample of people (or any other entity of interest). Typically, the longitudinal component comes from repeated cross-sections (i.e., a fresh sample of people each time) rather than panels (i.e., repeated measures on the original sample). Generational differences are confounded with age and period effects in either case.
The goal is to estimate how some attribute varies across birth cohorts or generations (cohort or generational effects), by age (age or life-cycle effects), and over time (period effects). The “identification problem” is that once one knows two of the three, the third is determined: year (period) – year of birth (cohort) = age. Thus, there is no simple way to estimate the effect of all three simultaneously. Many solutions to the APC identification problem have been proposed and implemented. Two recent special journal issues on APC analysis provide perspective on this terrain, one in Sociological Methods & Research (introduced by Smith 2008) and one in Electoral Studies (introduced by Neundorf and Niemi 2014).
Figures 2 and 3 further illustrate these quandaries, this time working with panel data from the Political Socialization Study initiated by M. Kent Jennings in 1965. The figures depict over-time data on two measures – social trust and trust in government, respectively – for three generations: high school seniors from the class of 1965, interviewed in 1965, 1973, 1982, and 1997, as they aged from 18 to 50 (Gen 2); their parents, interviewed in 1965, 1973, and 1982, as they aged from 47 to 64, on average (Gen 1); and their children, interviewed in 1997 when they were, on average, 23 years old (Gen 3).7 Although these are lineage generations, they can also be treated as representatives of three political generations. Gen 1 went through their impressionable years during the Great Depression and World War II, while Gen 2 is a Baby Boomer cohort that came of age during the turmoil surrounding the civil rights movement and the Vietnam War. The youngest generation came of age during the relatively prosperous and placid years of the Clinton presidency.
Figure 2 depicts the percentage of people who chose “you can trust other people to do what is right” instead of choosing either “you can’t be too careful in dealing with other people” or “it depends” in response to the standard social trust question also carried in the General Social Surveys. Figure 3 shows the percentage of people who provided the least trusting response when asked: “How much of the time do you think you can trust the government in Washington to do what is right? Would you say just about always, most of the time, or only some of the time?”
Consider the social trust results first (Figure 2). What jumps off the page is the massive gap (30+ percent) between the trust levels of the Gen 1 and Gen 2 on the one hand, and Gen 3 on the other. Members of the youngest generation are only about half as likely to say that they trust others as are those of their parent and grandparent generations. This holds true when one compares the generations at the same point in time (Gen 2 vs. Gen 3 in 1997) or at comparable points in the life cycle (Gen 3 in 1997 vs. Gen 2 in 1973). The possibility that this is a period effect is lessened by the fact that there is almost no downward trend in the responses for Gen 2, although Gen 1’s responses do trend downward slightly. The problem here, nevertheless, is that age-related changes and period effects could be working in opposite directions, with trust tending to rise with age yet period forces pushing in the opposite direction.
In a more thorough analysis of this question, Kent Jennings and I argue that the gap between Gen 3 and their parents that is shown in Figure 2 is a mix of generational differences and aging effects (Jennings and Stoker 2004). Social trust, we argue, tends to drop as people move into and through early adulthood before rebounding in the mid-thirties. A hint of this pattern is seen in the Gen 2 trust levels which bottomed out in 1973, when that generation was 26 years old. The 1997 survey caught members of Gen 3 at their trusting nadir.
In the trust in government results (Figure 3), we see that both Gen 1 and Gen 2 became more distrusting over time. Although this is likely a period effect, age-related change cannot be ruled out. Of the two, it is the older generation that is more distrusting, consistently so, though the gap between them diminishes somewhat as time passes. This gap is consistent with the idea that cynicism increases with age, but could also be attributable to generation differences. Notice also that Gen 2 at age 50 years (1997) is much more distrusting of government than was Gen 1 at a comparable age (1965). Although this gap could be interpreted in generational terms, a period effect interpretation is more plausible in light of the other evidence on trends in trust in government (see Levi and Stoker 2000 for a review). Members of the youngest generation were also more distrusting of government at (average) age 23 than were their parents at a comparable age, but there are no differences between the two when their views are compared at the same point in time. These patterns, too, could result from a combination of age, period, and generational effects.
Explaining Generational Differences
These figures illustrate the complexities of identifying generational differences and generational change. They also help introduce the second problem mentioned above, that of figuring out what, if anything, is provoking generational distinctiveness. The distinctiveness of Gen 3 on social trust is a perfect case in point. These representatives of Gen X are much less trusting of others than are their elders, no matter whether the comparison holds time or life-stage constant. But why? Although the massive literature on social capital and its decline has taken up this question, an answer remains elusive. The fact of the matter is that we are much better at mapping and tracking what look to be generational differences than understanding what has produced them in the first place.
The extent of this problem varies depending upon how the question of generational differences arises, which in turn bears on how generational boundaries are delineated. One finds three major approaches to this in scholarly writings and in the mass media. The first uses a generational lens to understand the consequences of a discrete historical event or varying historical conditions. Examples would include studies of generational differences tied to passage of the 19th amendment (e.g., Firebaugh and Chen 1995), the terrorist attacks of 911 (e.g., Sander and Putnam 2010), and the rise of party competition in the US South (e.g., Carmines and Stanley 1990). The delineation of generations flows from the researcher’s stipulation of the historical independent variable.
When the focus is on the effects of a discrete event, it is relatively straightforward to demarcate potential generations using a before and after logic – e.g., pre-911 vs. post-911 – though some definition of the IY period is required, which may be contested (age 15–25? 18–30?). If the focus is on historical conditions that are gradually changing, like the slow growth of party competition in the South, it is more productive to work with birth-year cohorts, since the expectation is that generational differences are emerging continually. Either way, the fact that researchers start with an explicit independent variable means that the question of why any generational difference would arise is addressed carefully in advance.
In a second approach, the focus is on understanding macro-level change in one or more dependent variables. A classic example here is the “Strange Disappearance of Social Capital in America” (Putnam 1995). Why is social capital disappearing? Why are more and more people bowling alone? (Putnam 2000). In cases like this, generational change is often one of a panoply of possible explanations, and can end up being little more than a black-box that is retained once other explanations have been discarded. Of course, once generational change is implicated, that sets up the challenge of explaining why generations differ. In the social capital case, many studies have followed up on Putnam’s idea that the rise of television is partly to blame for the generational differences (e.g., Norris 1996; Olken 2009).
In the third and most problematic approach as far as the “why” question is concerned, people are classified into groups by virtue of having been born within some time-frame, given generational labels, and then examined to see how they might differ from one another. This is commonplace in the mass media, where it seems as if journalists are constantly pointing out new ways in which young people (“Millenials,” “Generation Nice,” “the Internet Generation,” ….) differ from their elders. But it is also common in the work of serious scholars, who are drawn to understanding today’s youth in generational terms because of the possibility that, because of population replacement, what is true of youth today will be true of America down the road. The generational designations and boundaries that are used are often variable and somewhat arbitrary, which adds to the problems. But the larger issue is that we end up with a pile of ostensible generation differences and little clue as to why they all came about.
A Closing Example
Consider what researchers have discovered about Gen Xers and their younger brethren, the Millenials, compared to the Baby Boomer and Depression generations (to use a set of common labels, though others are certainly found) in the US. The younger cohorts are not only less trusting of others. They are also significantly less likely to vote, to view voting as an important civic duty, to express interest in politics, and to be knowledgeable about public affairs. While they do participate in politics, they tend to do so in more individuated and non-traditional ways. They are less likely to read newspapers at all, let alone follow politics in the medium, and to view news broadcasts on television. They are less likely to join civic organizations as adults and were less likely to do so as adolescents. When they do get involved in their communities, they participate in a more sporadic fashion. They are more materialist in their aspirations – increasingly focused on attaining their own economic goals – and yet also more post-materialistic in their values – increasingly likely to value self-expression and quality of life over physical and economic security.8
It is easy to draw these developments together into a bleak portrayal of the character of the youngest generations. But the youngest cohorts are also different from older cohorts in a number of more salutary ways. They are at least as likely or even more likely to engage in volunteer work (see, e.g., Levine 2007). New cohorts of college freshman are more likely than older cohorts to cite “becoming a community leader” as very important and to express an intention to participate in community service (Higher Education Research Brief 2008). Voting rates may be comparatively low for the young cohorts, but the age gap diminished in the most recent presidential elections as youth turnout rebounded from its 2000 low. Presumably because of candidate Obama’s targeting of youth through grass-roots mobilization, participation gains among young adults were even more pronounced in the 2008 caucuses and primaries (Dalton 2008). In light of these disparate trends, scholars have suggested that the norms and styles of citizenship are changing, with younger generations still concerned about the American society and polity, but expressing their citizenship in new ways (Bennett 2008; Dalton 2008).
Generation X and, especially, the Millenials are also distinctive in their political attitudes. Compared to earlier generations/older Americans, they hold more liberal attitudes on cultural issues involving gender roles, homosexuality, and gay rights, though not abortion. They are stronger advocates of civil liberties and hold more egalitarian or progressive attitudes on questions involving immigration and racial equality. Environmental conservation and clean energy are a higher priority. Their views on foreign policy are more anti-war and more supportive of cooperative, multilateral efforts. And, while distrusting of politicians and the government, they are nevertheless even more supportive than their elders of a strong governmental role in solving economic and social problems, including national health care. Not surprisingly in light of this configuration, they are more Democratic (and Independent) in their party identification and voting.9
Because the composition of the population in generational terms is shifting, with members of the older generations dying out, all of these findings – if truly generational differences – are important to America’s future. Will apathy, distrust, and disengagement be the norm among citizens in the coming decades? With the help of aging Gen Xers and Millenials, will the Democratic Party’s partisanship advantage persist or even grow in the future? Questions like these are important, and underscore why scholars are drawn to understanding what is distinctive about the young people of today.
Yet, the question remains: What is driving these patterns? I think it is fair to say that we are nowhere close to understanding what has produced these developments, although potential explanations abound. Compared to their elders, the younger generations are demographically different – more educated, less likely to be married, less religious, and more ethnically diverse. They were more likely to have grown up with a mother in the workforce, as children of immigrants, and in families marked by divorce. They are children of the digital age, with ready access to video games and the Internet. And, of course, they have grown up in a social and political milieu that in many ways is unlike what previous generations experienced. All of these characteristics and more are likely relevant to understanding the political distinctiveness of today’s younger generations. What we need now is clever research capable of going well beyond the mapping and tracking of differences to pin down the reasons that they have emerged.
Alwin, Duane F., and Ryan J. McCammon. 2003. “Generations, Cohorts, and Social Change.” In Handbook of the Life Course, edited by Jeylan T. Mortimer and Michael J. Shanahan, 23–49. New York: Kluwer Academic/Plenum Publishers.Google Scholar
Alwin, Duane F., Ronald L. Cohen, and Theodore M. Newcomb. 1991. Political Attitudes over the Life Span: The Bennington Women after Fifty Years. Madison, WI: University of Wisconsin Press.Google Scholar
Arber, Sara, and Claudine Attias-Donfut, eds. 2000. The Myth of Generational Conflict: The Family and State in Ageing Societies. London: Routledge.Google Scholar
Beck, Paul Allen. 1974. “A Socialization Theory of Partisan Realignment.” In The Politics of Future Citizens, edited by Richard G. Niemi, 199–219. San Francisco: Jossey-Bass.Google Scholar
Bennett, W. Lance. 2008. “Changing Citizenship in the Digital Age.” In Civic Life Online: Learning How Digital Media Can Engage Youth, edited by W. Lance Bennett, 1–24. Cambridge, MA: The MIT Press.Google Scholar
Carmines, Edward G., and Harold Stanley. 1990. “Ideological Realignment in the Contemporary South: Where Have All The Conservatives Gone?” In The Disappearing South? Studies in Regional Change and Continuity, edited by Robert P. Steed, Laurence W. Moreland, and Tod A. Baker, 21–33. Tuscaloosa, AL: University of Alabama Press.Google Scholar
Converse, Philip E. 1976. The Dynamics of Party Support: Cohort-Analyzing Party Identification. Beverly Hills, CA: Sage Publications.Google Scholar
Dalton, Russell J. 2008. The Good Citizen: How a Younger Generation Is Reshaping American Politics, Revised Edition. Washington, DC: CQ Press.Google Scholar
Delli Carpini, Michael X. 1989. “Age and History: Generations and Sociopolitical Change.” In Political Learning in Adulthood: A Sourcebook of Theory and Research, edited by Roberta S. Sigel. Chicago: University of Chicago Press.Google Scholar
Delli Carpini, Michael X., and Scott Keeter. 1996. What Americans Know about Politics and Why it Matters. New Haven: Yale University Press.Google Scholar
Dinas, Elias. 2014. “Why Does the Apple Fall Far from the Tree? How Early Political Socialization Prompts Parent-Child Dissimilarity.” British Journal of Political Science 44: 827–852.CrossrefGoogle Scholar
Funk, Carolyn L. 2013. “Genetic Foundations of Political Behavior.” In The Oxford Handbook of Political Psychology, edited by Leonie Huddy, David O. Sears, and Jack S. Levy. New York: Oxford University Press.Google Scholar
Green, Donald, Bradley Palmquist, and Eric Schickler. 2002. Partisan Hearts and Minds: Political Parties and the Social Identities of Voters. New Haven: Yale University Press.Google Scholar
Higher Education Research Brief. 2008. “The American Freshman: Forty-Year Trends: 1966–2006.” Accessed October 4, 2014. http://www.heri.ucla.edu/PDFs/pubs/briefs/40yrTrendsResearchBrief.pdf.
Inglehart, Ronald. 2007. “Postmaterialist Values and the Shift from Survival to Self-Expresive Values.” In The Oxford Handbook of Political Behavior, edited by Russell J. Dalton and Hans-Dieter Klingemann, 223–239. New York: Oxford University Press.Google Scholar
Jennings, M. Kent, and Richard G. Niemi. 1974. The Political Character of Adolescence. Princeton, NJ: Princeton University Press.Google Scholar
Jennings, M. Kent, and Richard G. Niemi. 1981. Generations and Politics. Princeton, NJ: Princeton University Press.Google Scholar
Jennings, M. Kent and Gregory B. Markus. 1984. “Partisan Orientations over the Long Haul: Results from the Three-Wave Political Socialization Panel Study.” American Political Science Review 78: 1000–1018.CrossrefGoogle Scholar
Jennings, M. Kent, and Laura Stoker. 2006. “Evaluating Dramatic Events during the Formative Years: A Longitudinal Approach.” Paper presented at the Midwest Political Science Association Convention, Chicago.Google Scholar
Jennings, M. Kent, and Laura Stoker. 2014. “Continuities in Political Participation across Generations.” Working Paper.Google Scholar
Levine, Peter. 2007. The Future of Democracy: Developing the Next Generation of American Citizens. Medford, MA: Tufts University Press.Google Scholar
Madland, D., and R. Teixeira. 2009. “New Progressive America: The Millennial Generation.” Center for American Progress. Accessed 10/4, 2014. http://www.americanprogress.org/issues/progressive-movement/report/2009/05/13/6133/new-progressive-america-the-millennial-generation/.
Mannheim, Karl. 1952 . Essays on the Sociology of Knowledge. London: Routledge and Kegan Paul.Google Scholar
Mindich, David T. Z. 2004. Tuned Out. New York: Oxford University Press.Google Scholar
Niemi, Richard G., and Mary A. Hepburn. 1995. “The Rebirth of Political Socialization.” Perspectives on Political Science 24: 7–17.Google Scholar
Pew Research Center. March 7, 2014. “Millenials in Adulthood.” http://www.pewsocialtrends.org/files/2014/03/2014-03-07_generations-report-version-for-web.pdf.
Plutzer, Eric. 2002. “Becoming a Habitual Voter, Inertia, Resources, and Growth in Young Adulthood.” American Political Science Review 96: 41–56.Google Scholar
Putnam, Robert D. 1995. “Tuning In, Tuning Out: The Strange Disappearance of Social Capital in America.” P.S.: Political Science and Politics 28: 1–20.Google Scholar
Putnam, Robert D. 2000. Bowling Alone. New York: Simon & Schuster.Google Scholar
Rahn, Wendy M. and John E. Truansue. 1998. “Social Trust and Value Change: The Decline of Social Capital in American Youth, 1976–1995.” Political Psychology 19: 545–565.Google Scholar
Sander, Thomas H. and Robert D. Putnam. 2010. “Still Bowling Alone? The Post-9/11 Split.” Journal of Democracy 21: 9–16.Google Scholar
Sears, David O. 1983. “The Persistence of Early Political Predispositions: the Roles of Attitude Object and Life Stage.” In Review of Personality and Social Psychology, edited by Ladd Wheeler and Phillip R. Shaver, 79–116, Vol. 4. Beverly Hills, CA: Sage Publications.Google Scholar
Sears, David O. 1990. “Whither Political Socialization Research? The Question of Persistence.” In Political Socialization, Citizenship Education, and Democracy, edited by Orit Ichilov, 69–97. New York: Teachers College Press.Google Scholar
Sears, David O., and Christia Brown. 2013. “Childhood and Adult Political Development.” In Oxford Handbook of Political Psychology. 2nd ed., edited by Leonie Huddy, David O. Sears, and Jack S. Levy, 59–95. New York: Oxford University Press.Google Scholar
Smith, Tom W. 2005. “Generation Gaps in Attitudes and Values from the 1970s to the 1990s.” In On the Frontier of Adulthood: Theory, Research, and Public Policy, edited by Richard A. Settersten, Frank F. Furstenberg and Rubén G. Rumbaut, 177–221. Chicago: University of Chicago Press.Google Scholar
Stoker, Laura, and M. Kent Jennings. 2008. “Of Time and the Development of Partisan Polarization.” American Journal of Political Science 52 (July): 619–635.Google Scholar
Verba, Sidney, Nancy Burns, and Kay L. Schlozman. 2003. “Unequal at the Starting Line: Creating Participatory Inequalities across Generations and among Groups.” The American Sociologist 34: 45–69.CrossrefGoogle Scholar
Wolfinger, Raymond E., and Steven J. Rosenstone. 1980. Who Votes? New Haven, CT: Yale University Press.Google Scholar
Wuthnow, Robert. 1998. Loose Connections: Joining Together in America’s Fragmented Communities. Cambridge, MA: Harvard University Press.Google Scholar
Zukin, Cliff, Scott Keeter, Molly Andolina, Krista Jenkins and Michael X. Delli Carpini. 2006. A New Engagement? Political Participation, Civic Life, and the Changing American Citizen. New York: Oxford University Press.Google Scholar
For example, recent articles or blog posts from the New York Times discussed the “Self(ie) Generation” (March 7, 2014), the “Antidepressant Generation” (April 17, 2014), the “Online Generation” (March 21, 2014), and “Generation Nice” (August 15, 2014).
Some scholars prefer to refer to cohort effects (or cohort replacement effects). Cohort or generational effects are often contrasted with period effects and age or life cycle effects, as discussed below.
For example, two recent articles in the Huffington Post claim to identify “Fifty Things Only People Over 50 Understand.” http://www.huffingtonpost.com/2013/11/01/baby-boomer-memories_n_4173687.html, and “5 Things Boomers Don’t Get About Young People.” http://www.huffingtonpost.com/2014/03/07/generation-gap_n_4876159.html.
The terminology to use here is tricky. Strictly speaking, one should talk about studying cohorts to see if they represent distinct generations, i.e., to see if there are lasting differences between or among them tied to the historical period in which they went through their IYs. At the same time, cohorts are often given generational names (e.g., Generation X or Y, Baby Boom Generation) and are studied so as to identify the way that they are different from one another. I use “generation” here in that context.
Compositional differences across the groups can also be a problem, for example, the fact that older Americans tend to have less education than younger Americans. I consider compositional confounds in a different context, below.
Support for most of these statements can be found in Bennett (1998), Dalton (2008), Delli Carpini (2000), Levine (2007), Putnam (2000), and Zukin et al. (2006). See also Rahn and Transue (1998) on social trust and materialism, Mindich (2004) on media usage, Wuthnow (1998) on sporadic civic involvement, Inglehart (2007) on post-materrialism, and Delli Carpini and Keeter (1996, ch. 4) on political knowledge. Not all of the evidence extends to both Gen X and the Millenials. Some of the findings taken as evidence of “generation gaps” are based on cross-sectional analyses although most research tries to supplement that comparison with one that holds life stage constant.
About the article
Laura Stoker is Associate Professor of Political Science at the University of California, Berkeley.
Published Online: 2014-10-31
Published in Print: 2014-10-01