Aged Communities and Political Knowledge

Abstract

Past work emphasizes the decline of cognition into older age. Recent work suggests that living in an aged community provides ample opportunity for social interaction with peers and that these older residents perform better cognitively than more isolated seniors. I test whether this relationship is evident for the political cognition of older residents with NAES data from 2000 and 2004. Findings indicate higher levels of political knowledge among seniors living in aged communities compared with their peers living in places without the same social context.

Keywords

political knowledge aging social context community

Knowledge of political processes, institutions, and participants is important for sustaining democratic government, by the people and for the people. Citizens who are politically knowledgeable are also likely to be interested in and more likely to participate in their government (Delli Carpini & Keeter, 1997), and these qualities ensure that well-designed institutions continue to work well for society (Galston, 1991, 2001). Political knowledge helps citizens identify their interests, helps them form coherent opinions on a range of issues in line with those interests, and then gives people resources to link their attitudes with activities that serve their interests (Delli Carpini & Keeter, 1997, p. 219).

Political knowledge should increase with age (Strate, Parrish, Elder, & Ford, 1989; Wolfinger & Rosenstone, 1980). As people gain political experience over the course of their lives, they have the opportunity to accumulate and retain knowledge of political processes (Plutzer, 2002). People also become more interested in local and community politics with age, as they become enmeshed in their communities (Alford & Scoble, 1968; Blum & Kingston, 1984; Nie, Verba, & Kim, 1974; Sykes, 1950) and accrue more detailed and specific political knowledge on candidates and campaign platforms in addition to gaining an increasing familiarity with political institutions at all levels of government.

Many aspects of the unfolding life cycle encourage the accumulation of political knowledge even as people enter retirement. However, some forms of political knowledge may decrease for those reaching advanced age due to declining cognitive abilities that commonly accompany old age (Lau & Redlawsk, 2006, 2008; Riggle & Johnson, 1996). Decreased political knowledge among some of the most politically engaged citizens means that older people may be less likely to express their political preferences accurately. Republican government is based on the premise that elected officials represent the preferences of their constituents. When preferences are unclear or mistaken, the goals of representative government are undermined.

These consequences may be especially worrisome for the United States as the unusually large and historically active Baby Boomer Generation moves into older adulthood. On the other hand, recent research indicates that neighborhoods with large older adult populations preserve cognitive abilities among their older residents through increased opportunities for social interaction and information exchange with peers (Clarke et al., 2012). Communities with substantial older populations are on the rise in the United States as the population ages.

Americans who were age 65 and older¹ in 2010 comprised approximately 13% of the country’s population that year. In comparison, the percentage of older adults is significantly greater for the aged communities studied here. For example, older adults (65 and above) made up about 31% of Florida’s Sarasota County in 2010, and 33% of the local population in La Paz County, Arizona. While a great number of aged locales can be found in the traditional retirement hotspots like Florida and Arizona, neighborhoods with overwhelming numbers of older adults are popping up all over the country as people increasingly age in place (Frey, 2011; Wolf, 2001). Aged locales now include a diversity of communities like Garfield County in Nebraska, Northumberland County in Virginia, and Towns County in Georgia.

We know that political knowledge increases with age, then peaks into old age, and finally declines with very advanced years. However, as aged communities become more commonplace, older adults may increasingly find themselves socially integrated with their peers. Although demographically and politically diverse, the aged social context will present opportunities for social interaction with peers in the older communities and thus prevent, curb, or delay the natural losses to political cognition experienced by the oldest Americans. In this article, I examine locations with varying age distributions across the United States and test whether the places with older-leaning age distributions support increased political knowledge among their elderly residents.

Aging and Cognition

People obtain knowledge about the world around them as they age and accumulate new experiences, which become familiar and routine as they are encountered again and again. Wisdom and perspective often comes with age, but advanced age may also bring senility, dementia, and a host of other ailments and impediments to maintaining and expressing a lifetime of accumulated knowledge (Schaie, 2005; Zarit & Zarit, 2007). These processes associated with later years have consequences for many aspects of an aging person’s life, including political cognition and participation (Lau & Redlawsk, 2006, 2008; Riggle & Johnson, 1996). Understanding voters’ considerations when making political decisions is critical for political behavior research. Paying attention to these factors, particularly among the oldest generations may be of the utmost importance for research in American politics as the aged population surges in the coming decades.

Older people generally take longer to process information than younger adults (Hartley, 2006) and have been found to employ simpler decision-making strategies (Mata, Schooler, & Rieskamp, 2007). The aging cognition effects vary by the type of memory considered and may not always indicate poorer decision-making (Mata et al., 2007). Recent research on aging and political cognition has been consistent with the more general findings above. Experimental research has shown that older people consider less information for longer periods of time when making political decisions, compared with younger participants (Lau & Redlawsk, 2006; Mata et al., 2007; Riggle & Johnson, 1996).

Declines in semantic memory, the concepts and meanings developed over time (Quillian, 1968; Smith, Shoben, & Rips, 1974), tend to be small or insignificant in older adulthood (Hasher & Zacks, 1988; Hess, 2005; Hoyer & Verhaeghen, 2006).

Lau and Redlawsk (2008) make the comparison between the resilient semantic memory and crystallized political knowledge. This knowledge includes memories related to government processes, parties, and better known candidates—the recurring and persistent aspects of politics. Lau and Redlawsk (2008) find that 78-year-olds answer 6 more correct answers (out of a possible score of 20) to factual-based political knowledge questions than 18-year-olds.

On the other hand, older adults have more difficulty recalling detailed information related to periodic events, otherwise considered to be part of short-term memory (Hoyer & Verhaeghen, 2006). Age may be a stronger predictor of accurate political party knowledge than of candidate knowledge because candidates come and go, but the two major political parties are entrenched in American history (Lau & Redlawsk, 2008). The ability to engage in more active cognitive processing regresses with age (Hasher & Zacks, 1988; Hess, 2005). Such processes might include remembering new information generated from current political campaigns, relating the information to personal preferences, and choosing the candidate that most accurately represents those preferences. Lau and Redlawsk (2008) show that the likelihood of making correct vote choices (matching one’s preferences to the candidate representing those preferences) decreases once people reach their mid-60s.

Some aspects of political cognition are retained and even peak well into older adulthood while others experience decline with the onset of retirement age (Lau & Redlawsk, 2006, 2008; Riggle & Johnson, 1996). However, even some of the most crystallized political memories can be forgotten with very advanced age. Losses to cognition are evident in most people reaching their mid-1970s and 1980s (Schaie, 2005). Lau and Redlawsk (2008) find that decline in political cognition begins after age 69 for political knowledge, 67 for memories related to party politics, a much later 81 for memories of candidates, and around age 70 for correct voting.

Community Context and Cognition

The process of getting older influences cognition later in life, but whether a person ages in community or alone may have an additional impact on their cognitive abilities, including their political knowledge. To begin with, older people belonging to a social network of peers may be less likely to develop dementia than older adults living in solitude. Fratiglioni et al. (2000) followed a community-based sample of people without dementia for 3 years and collected social network information for the participants. At the end of the 3-year study, the residents who lived alone and had very few social ties were found to be at higher risk for developing dementia, compared with the residents living with and among others.

The cognitive reserve hypothesis proposes that living in community may directly influence brain development of seniors and guard against dementia and maintain cognitive abilities into older adulthood (Stern, 2006). Living in a community of other older adults may offer seniors emotional support (Seeman, Lusignolo, Albert, & Berkman, 2001) and promote opportunities for social interaction and intellectual stimulation (Fratiglioni, Paillard-Borg, & Winblad, 2004; Wang, Karp, Winblad, & Fratiglioni, 2002).

Different types (and measures) of community are linked with varying levels of cognitive capacity. An entire literature has been devoted to examining the relationship between particular characteristics of geographic locales (often measured as census tracts, counties, or zip codes) and the well-being of area residents. For example, the local context of socioeconomic deprivation has been shown to be associated with diminished cognitive capacity among older residents (Lang et al., 2008). Related work suggests a link between the average education level in a community and the cognitive abilities of residents. Older people living in highly educated locations show higher levels of cognition compared with their peers living in places with less-educated residents, even when accounting for individual education (Wight et al., 2006). Socioeconomic conditions may influence individual cognition because impoverished communities are less likely to have the concentrations of physical, social, and institutional resources (parks, libraries, recreation and community centers, etc.) often found in well-off communities (Clarke et al., 2012).

Another community characteristic that likely has implications for the cognitive reserve of older residents, and the one considered for the present study, is the local age structure. Given the findings related to community socioeconomic conditions, Clarke et al. (2012, p. 2) speculate that seniors living in places with greater percentages of older adults will benefit from “increased opportunities for social interaction with peers or a greater density of age-specific resources” like senior centers and social clubs for retirees.

Using data collected in Chicago neighborhoods, Clarke et al. (2012) find support for their hypotheses. Living in parts of the city with concentrations of older adults was related to greater levels of cognitive function for older residents, however, their finding is qualified by years of residency. The effect is not as evident for longer term residents. Given the differences in cognitive decline among older adults in their 60s and early 70s compared with those reaching very advanced ages (Lau & Redlawsk, 2008; Schaie, 2005), the difference in effect for residency may simply reflect age differences between shorter and longer term residents. Shorter term residents are likely the younger, more active older adults, primed and able to absorb what the aged community has to offer.

Clarke et al. (2012) are not alone in considering implications of a skewed local age structure for individual cognitive engagement. Past research connects aged communities with concentrations of community resources (Cagney, 2006) and greater opportunities for social networks and engagement (Cagney, 2006; Longino, McClelland, & Peterson, 1980; Ward, LaGory, & Sherman, 1985). An aged context may also have implications for local and national politics. Older age distributions create environments that on one hand promote attitudes supportive of safety net programs, which often benefit senior citizens, and on the other hand, encourage politically inefficacious attitudes among residents. The present research builds from this work by examining the impact of the community age structure on political cognition. The cognitive reserve benefits of living in aged communities may include increased political knowledge and cognition into older adulthood.

Theorizing About the Aged Context and Political Cognition

The above research provides two explanations for the contextual influence of the aged communities.

Older residents retain more political knowledge (compared with their peers in other locations) through more frequent opportunities for social interaction with other older adults provided by the older adult context.

Older residents retain more political knowledge (compared with their peers in other locations) through a greater density of older-age-related resources provided by the older adult context.

Both of the mechanisms are likely influencing mental stimulation and thus, cognitive reserves of political knowledge—and they are likely working together. The two explanations are interconnected and may be impossible to completely separate with the available data. Participating in the social networks of aged communities may not be possible without also taking advantage of the age-related resources and vice versa.

Past research suggests that the age context may influence different types of political knowledge to different degrees, and the effects for older adults may not be the same for those of very advanced age. In this article, I examine three types of political memories or knowledge: factual-based knowledge of the political system, candidate issue positions, and candidate positions on senior-related issues. The two older-adult age categories of particular interest are emergent older adults, ages 65 to 74, and those of more advanced age, 75 and older. Consideration of research on aged contextual effects and on cognitive decline lead to a number of hypotheses:

Hypothesis 1: Emergent older adults (those age 65-74) living in a context of older adults will know more about the political system than their peers living elsewhere. The older adult context will have a negligible effect on the oldest residents.

Hypothesis 2: Emergent older adults (those age 65-74) living in a context of older adults will know more about candidate issue positions than their peers living elsewhere; however, this effect will be smaller than the effect for the political knowledge of well-established institutions. The older adult context will have a negligible effect on the oldest residents.

Hypothesis 3: Emergent older adults (those age 65-74) living in a context of older adults will know more about the candidate positions on senior-citizen-related issues than their peers living elsewhere. This effect will be greater than the effect for candidate issue positions in general but still less than the effect for the most concrete and lasting knowledge of the political system. The older adult context will have a negligible effect on the oldest residents.

Data and Method

The National Annenberg Election Survey (NAES) data from 2000 and 2004 provide a variety of knowledge questions posed to individual respondents. United States Census population data from 2000 and 2004 (estimated) are used for the contextual measures. Choosing a unit of analysis at the community level should be done with care but, ultimately, there may be no perfect fit (Taylor, Gorard, & Fitz, 2003; Williams, 1999). Using more than one measure is often preferred to ease concerns about the modifiable areal unit problem.² I measure the contextual community as both county and zip code in the following analyses and compare the results. These geographic contextual measures have been used often in social science research, with the county as a useful political jurisdiction and the zip code as a convenient neighborhood record provided by a great number of surveys (Williams, 1999).

The independent variable of primary interest measures the aged population within a community context. As with past work on age contexts, I divide the local population of people of a certain age by the total local population. This measure results in a proportion.³ I am interested in the proportion of aged people living in counties and zip codes, so I create two aged community measures. For the county measure, I divide the county population of people age 65 and older by the total county population. For the zip code measure, I divide the zip code population of people age 65 and older by the total zip code population.

The resulting county-level aged community measure ranges from 3% to 35% while the zip code–level measure ranges from 0% to 83%. To clarify, older adults make up 35% of the population in the county with the highest proportion of older adults and only 3% in the county with the lowest proportion in the sample. Counties are larger land areas and often contain more diverse populations, including a greater diversity of ages, compared with zip codes. The smaller zip codes, used for the purposes of the U.S. Postal Service, may be made up entirely of age-restricted communities and be home to very few younger residents. On the other end of the aged distribution spectrum, many universities have their own zip codes with zero older adults in residence.

Hierarchical general linear modeling is used to model the relationship between the older age distributions at the county- and zip code levels and the political knowledge of individual community residents. Since the dependent variables are count variables of equal exposure, an Overdispersed Poisson model with a log link is used. Survey items for political system knowledge, candidate issue positions, and candidate positions on the elderly related issues (social security and prescription drug coverage)⁴ were added together to make three dependent variables, which serve as knowledge indexes (see Table A1 in the online appendix for the list of questions and wording). Two of the measures rely on 4 items while the index for candidate issue positions relies on 10 items⁵ (see Figures A1-3 in the appendix for survey respondent distributions for the dependent variables).

Additional items will be accounted for (at both levels of analysis) when predicting responses to the knowledge questions. At the individual level, I control for partisan strength, income, race, gender, and level of education. Many of these characteristics are highly predictive of political knowledge and have been widely used in previous research (Delli Carpini & Keeter, 1997; Lau & Redlawsk, 2008).

At Level 2, the community level, I control for population density and median household income by modeling the age category slopes. Seniors differ fiscally and make different choices (with some not having any choice) for their retirement migration plans. Some choose to move to and live in retirement havens like Sumter County, Florida or Maricopa County, Arizona, while others remain in their lifetime homes in small town locations such as Woods County, Oklahoma, or Izard County, Arkansas. The active retired locations advertise their senior-living amenities and draw older residents from all over the country. Other places with concentrations of older adults often lose their younger adult residents to metropolitan areas with greater economic opportunities. Because aged communities often differ greatly by population density and by economic prospects, it is crucial to account for these community characteristics in the analyses given their potential for housing different aged populations with various political dispositions.⁶ Although the aged locations vary in size and economic means, they all have one major characteristic in common: they are home to a lot of older residents, relative to other age groups, creating a unique social context with consequences for political knowledge.

After analyzing the relationships between the older neighborhood context and political knowledge, I will also briefly explore the theoretical foundations for these relationships.

In the above section, I note that it is very difficult to separate the mechanisms at work within the aged communities. Can we really separate the unique opportunities for social interaction from the density of resources available in older locales? In the last part of the data analysis section, I attempt to gain some leverage on these theories by addressing each proposed mechanism, albeit with some limitations. I use items from the National Annenberg Election Survey that measure political discussion during the week to better understand the role of discussion networks in aged communities. I look at measures of local political engagement to touch on the density of political resources available in these older communities. The political news available in the aged communities may be especially relevant for seniors, encouraging enhanced local engagement.

Hierarchical general linear modeling is also used to assess these relationships. Overdispersed Poisson models are employed for the day-count variables (number of days discussing politics and number of days reading the newspaper), and a binary model with a logit link function is used for the measures of attention to the local news on television and in the newspaper.

Results

Before discussing the results for the multilevel models, Figure 1 provides descriptive information from the NAES data on the knowledge scores for a few age categories. The thick black line represents respondents of all ages as a comparison. The first panel in the figure shows the scores for political system knowledge. As expected, a higher percentage of people in the older age groups responded correctly to 3 or 4 answers than younger age groups. Approximately 35% of people age 75 and older scored 3 out of 4 points, but only 23% scored the maximum of 4 points—only slightly higher than the percentage for the youngest age group.

Figure 1.

Knowledge of political system, candidate issue positions, and candidate senior issue positions by age category.

The second panel presents the knowledge scores for candidates’ positions on a variety of issues. The age groups do not differ very much on these knowledge scores with the exception of the oldest group. Many more of the 75 and older respondents answer fewer questions correctly than the other age groups. Finally, the third panel provides the scores for the candidates’ positions on senior-relevant issues in particular. The emergent older adults, those ages 65 to 74, scored the highest on these older-age-related items. Few respondents in the youngest age group and in the oldest age group achieved 3 or 4 on the senior issue knowledge index.

The descriptive information supports past research and provides additional evidence that older generations are generally more politically knowledgeable, but that this knowledge may taper into the latest decades of life. In addition, the percentages of correct answers varied by the types of knowledge. It may be especially notable that the 65- to 74-year-olds were the highest scorers on the age-specific knowledge items. The hierarchical multivariate models provide a test for whether these relationships (between age and political knowledge) vary with the context of the local age distribution.

Table 1 presents the results for the hierarchical models for the three dependent variable knowledge scores with zip code–level context measures. Table 2 contains the same model specifications but with county population data at Level 2. I discuss results for each type of knowledge index at a time, comparing the results for the different context measures in Tables 1 and 2.

Table 1.

Zip Code HGLM Results for Knowledge Measures.

	Political system knowledge 2004 NAES	Candidate issue position knowledge 2000 NAES	Candidate senior issue position knowledge 2000 NAES
Level 2 variables
Intercept
Intercept	1.013** (0.007)	1.677** (0.012)	0.714** (0.014)
Old age (65 to74)
Population density (1000 per sq mile)	0.001 (0.001)	0.003 (0.002)	0.001 (0.001)
Median household income (in US$1,000s)	0.002** (0.000)	0.001 (0.001)	0.003** (0.001)
Aged community proportion	0.002* (0.001)	0.001 (0.002)	0.005* (0.002)
Old age (75 and above)
Population density (1000 per sq mile)	0.001 (0.002)	0.001 (0.001)	0.002 (0.002)
Median household income (in US$1,000s)	0.001* (0.001)	−0.002 (0.002)	−0.000 (0.001)
Aged community proportion	0.001 (0.001)	−0.005 (0.003)	−0.001 (0.003)
Level 1 variables
Young age	−0.145** (0.007)	0.009 (0.010)	−0.057** (0.013)
Old age (65 to 74)	−0.079* (0.028)	−0.093 (0.065)	−0.125* (0.054)
Old age (75 and above)	−0.103* (0.035)	−0.024 (0.088)	−0.001 (0.087)
Four-year college degree	0.154** (0.006)	0.120** (0.011)	0.139** (0.013)
Female	−0.167** (0.005)	−0.125** (0.010)	−0.150** (0.011)
Partisan strength	0.085** (0.006)	0.116** (0.010)	0.138** (0.012)
Less than 10K	−0.309** (0.022)	−0.141** (0.025)	−0.187** (0.028)
Income 35 to 50K	−0.002 (0.008)	−0.000 (0.013)	−0.006 (0.015)
Income 50 to 75K	0.076** (0.024)	0.058** (0.012)	0.059** (0.014)
Income 150K or more	0.157** (0.008)	0.139** (0.021)	0.173** (0.022)
Non-White	−0.187** (0.100)	−0.027* (0.013)	−0.088** (0.016)
Level 1 N =	25,308	7,755	12,522
Level 2 N =	9,022	4,920	6,618
Reduction in error variance =	0.75	0.71	0.85

Note. Source: NAES 2000, 2004 and U.S. Census 2000, 2004 estimates.

p < .05. **p < .001.

Table 2.

County HGLM Results for Knowledge Measures.

	Political system knowledge 2004 NAES	Candidate issue position knowledge 2000 NAES	Candidate senior issue position knowledge 2000 NAES
Level 2 variables
Intercept
Intercept	1.000** (0.007)	1.676** (0.013)	0.711** (0.015)
Old age (65 to74)
Population density (1,000 per sq mile)	0.001 (0.001)	0.004** (0.001)	0.003 (0.001)
Median household income (in US$1,000s)	0.002** (0.001)	0.001 (0.002)	0.005** (0.001)
Aged community proportion	0.004* (0.002)	−0.006 (0.004)	0.011** (0.003)
Old age (75 and above)
Population density (1000 per sq mile)	0.001 (0.001)	0.002 (0.001)	−0.000 (0.002)
Median household income (in US$1000s)	0.001 (0.001)	0.001 (0.003)	0.002 (0.003)
Aged community proportion	0.001 (0.002)	−0.001 (0.006)	0.006 (0.006)
Level 1 variables
Young age	−0.145** (0.007)	0.009 (0.010)	−0.057** (0.012)
Old age (65 to 74)	−0.130* (0.047)	−0.016 (0.107)	−0.257* (0.088)
Old age (75 and above)	−0.042 (0.055)	−0.210 (0.194)	−0.170 (0.162)
Four-year college degree	0.154** (0.005)	0.121** (0.011)	0.138** (0.013)
Female	−0.168** (0.005)	−0.126** (0.010)	−0.150** (0.012)
Partisan strength	0.085** (0.006)	0.117** (0.010)	0.139** (0.012)
Less than 10K	−0.316** (0.022)	−1.142** (0.026)	−0.184** (0.027)
Income 35 to 50K	0.001 (0.008)	−0.001 (0.013)	−0.006 (0.016)
Income 50 to 75K	0.072** (0.007)	0.056** (0.013)	0.057** (0.014)
Income 150K or more	0.153** (0.009)	0.134** (0.021)	0.175** (0.022)
Non-White	−0.187** (0.011)	−0.030* (0.012)	−0.089** (0.015)
Level 1 N =	23,285	7,729	12,464
Level 2 N =	2,514	1,762	2,084
Reduction in error variance =	0.57	0.68	0.52

Note. Source: NAES 2000, 2004 and U.S. Census 2000, 2004 estimates.

p < .05. **p < .001.

First, individual income, party attachment, gender, education, and race strongly predict respondents’ political knowledge for all three measures. Consistent with past research, the more affluent and higher educated respondents tend to be more politically engaged and knowledgeable. Having a high income, being strongly identified with a political party, being of male gender, having graduated from a 4-year college and being White are all associated with higher scores on the three dependent variables measuring political knowledge. These relationships are consistently evident for both the zip code–level and county-level results.

A person’s age is also a good indicator of his or her political system knowledge and his or her knowledge of candidate positions for senior issues. These effects are similar for both the models with zip code contextual data and those with county contextual data. Tables 1 and 2 show that younger and older people have less knowledge of the political system than respondents of middle age, controlling for all of the other factors in the models. However, older adults living among their peers may differ in their political knowledge stores compared with older adults living in places without the same aged context.

To understand the effect of the aged community, I am most interested in the relationship between the local age structure and older residents’ political knowledge scores. To explore this effect, I model the age categories in the HGLM analyses with the aged community measures. This modeling choice allows for the comparison of the effect for older adults living in places ranging from an insignificant aged context (0%, 3%; zip code, county) to the political knowledge of their peers residing in locations with the highest proportions of older adults (83%, 35%; zip code, county).

The relationships between the aged community measures and the political knowledge scores are very similar for the zip code and county results, with only slightly larger coefficients for the county results.⁷ Older adults living in the aged community zip codes and counties have higher levels of political system knowledge than their peers living in places with fewer older adults. However, this same effect is not evident for the oldest respondents. Knowledge of the political system for people age 75 and older does not vary with the age structure of their local community.

As for substantive significance, the 65- to 74-year-olds⁸ living in an aged community on average answer 0.36 more questions correctly than their peers living without the same older community. While the difference of 0.36 in the scores may seem small, it represents 9% of the total possible score. If the respondents at both ends of the aged community spectrum received quiz grades for their responses, the effect would be similar to receiving a score of a 62 (least aged communities) compared with a score of a 71 (most aged communities).

The results for the models in Tables 1 and 2 predicting knowledge of candidates’ issue positions are also comparable. Overall, age and the age context are not good indicators for knowledge of the candidates’ positions, controlling for all of the other factors. Scores drop slightly for respondents in the most aged communities. There is a 0.89 (9%) decline in the score for candidate issue position knowledge among 65- to74-year-olds living in the most aged communities, compared to their peers in less aged locales. The decline is steady but not statistically significant.

Results may be more easily understood when measuring respondents’ knowledge of candidates’ positions for senior-related issues alone. Once again, results are consistent across the two context measures in the two tables. Younger and older adults have less knowledge of the aging-related issues, when controlling for all of the other factors. However, the aged community measure coefficient is significantly positive when modeled on the slope for the emergent older adults, age 65 to 74.

Figure 2 highlights the substantive effect.⁹ People age 65 to 74 who live among large numbers of their peers, relative to people of other ages, answer on average 0.83 additional questions (out of 4) correctly than their peers in locations without the aged community. This would be like going from a 48 on an exam to a grade of 69. Both grades do not seem very good, but one is significantly more informed on candidate positions for senior-related issues—a nice effect for the aged community context. However, as with the measure of political system knowledge, the same effect is not observed for the oldest adults, aged 75 and older.

Figure 2.

Candidate senior issue position knowledge scores for emergent older adults.

It should also be noted that the aged community effects hold even when considering aged communities as different as Lake County, Florida and Pawnee County, Nebraska.¹⁰ Sometimes the places with higher local population densities and greater levels of household incomes are associated with higher levels of political knowledge, so it is important to account for these differences between locations in the models—ensuring even more confidence in aged context findings.

Finally, I examine the mechanisms behind the relationships described above. Why does the age context of a community relate to the knowledge, especially for senior-related issues, of the older residents? Past work suggests that two possible factors making the aged locales unique. The models in Table 3 examine the idea that the aged communities may encourage increased opportunities for social interaction with their older adult peers. Although the measures of discussion (number of days discussing politics with family and friends during the week), do not necessarily measure opportunities for discussion, they do measure the levels of political discussion among senior residents. Looking at the results in Table 3, there does not appear to be a strong relationship between the local age context and the levels of political discussion among the older residents. For the emergent older adults, the relationship was negative in 2004 but positive in 2000, but neither relationship comes close to statistical significance.

Table 3.

County HGLM Results for Discussion Measures.

	Discussed politics with friends or family (week days) 2004 (NAES)	Discussed politics with family or friends (week days) 2000 (NAES)
Level 2 variables
Intercept
Intercept	0.920** (0.009)	0.608** (0.015)
Old age (65 to74)
Population density (1,000 per sq mile)	−0.003 (0.002)	0.001 (0.002)
Median household income (in US$1,000s)	0.002* (0.001)	0.005* (0.002)
Aged community proportion	−0.004 (0.003)	0.001 (0.004)
Old age (75 and above)
Population density (1,000 per sq mile)	0.001 (0.001)	0.003 (0.003)
Median household income (in US$1,000s)	0.002 (0.001)	0.005 (0.002)
Aged community proportion	0.003 (0.003)	0.006 (0.006)
Level 1 variables
Young age	−0.126** (0.007)	−0.114** (0.011)
Old age (65 to 74)	0.018 (0.064)	−0.126 (0.098)
Old age (75 and above)	−0.253** (0.083)	−0.345* (0.155)
Four-year college degree	0.098** (0.007)	0.158** (0.013)
Female	−0.030** (0.006)	−0.118** (0.011)
Partisan strength	0.314** (0.007)	0.410** (0.011)
Less than 10K	−0.283** (0.021)	−0.321** (0.030)
Income 35 to 50K	−0.006 (0.009)	0.026 (0.014)
Income 50 to 75K	0.064** (0.008)	0.112** (0.015)
Income 150K or more	0.255** (0.010)	0.332** (0.022)
Non-White	−0.117** (0.010)	−0.154** (0.015)
Level 1 N =	68,561	37,035
Level 2 N =	2,908	2,717
Reduction in error variance =	0.86	0.44

Note. Source: NAES 2000, 2004 and U.S. Census 2000, 2004 estimates.

p < .05. **p < .001.

Table 4 presents the models exploring the other proposed theory: aged communities provide a greater density of older-age-related resources than other places. The communities are political hotbeds. I consider these factors by using measures of local engagement (great deal of attention to political news on local television, political news in the newspaper, and number of days reading the newspaper).¹¹ A glance at Table 4 shows positive relationships between the aged community context and local engagement, however, the relationships do not reach the level of statistical significance where p < .05. However, emergent older adults in the most aged communities were 8 percentage points more likely to watch political news on local television than these seniors in the locations at the other end of the aged spectrum. The p value for this coefficient is .068.

Table 4.

County HGLM Results for Local Engagement Measures.

	Great deal of attention political news on local TV 2004 (NAES)	Great deal of attention political news in newspaper 2004 (NAES)	Read newspaper this week (week days) 2000 (NAES)
Level 2 variables
Intercept
Intercept	−2.169** (0.035)	−1.722** (0.031)	−0.491** (0.029)
Old age (65 to74)
Population density (1,000 per sq mile)	−0.011* (0.005)	−0.003 (0.003)	−0.004 (0.003)
Median household income (in US$1,000s)	0.001 (0.003)	0.008* (0.002)	0.015** (0.004)
Aged community proportion	0.015 (0.008)	0.011 (0.007)	0.010 (0.008)
Old age (75 and above)
Population density (1,000 per sq mile)	0.007 (0.005)	0.008 (0.005)	0.005 (0.006)
Median household income (in US$1,000s)	0.003 (0.003)	0.006* (0.003)	0.011* (0.005)
Aged community proportion	0.004 (0.011)	0.018* (0.008)	0.003 (0.012)
Level 1 variables
Young age	−0.348** (0.030)	−0.313** (0.027)	−0.911** (0.024)
Old age (65 to 74)	0.160 (0.207)	−0.055 (0.183)	0.233 (0.227)
Old age (75 and above)	0.207 (0.258)	−0.038 (0.204)	0.647* (0.314)
Four year college degree	−0.010 (0.031)	0.081** (0.025)	0.174** (0.028)
Female	−0.007 (0.024)	−0.267** (0.022)	−0.230** (0.023)
Partisan strength	0.603** (0.028)	0.625** (0.025)	0.295** (0.020)
Less than 10K	0.158* (0.062)	−0.187* (0.065)	−0.707** (0.056)
Income 35 to 50K	0.002 (0.031)	−0.071* (0.030)	−0.010 (0.029)
Income 50 to 75K	−0.017 (0.033)	−0.022 (0.029)	0.217** (0.029)
Income 150K or more	−0.055 (0.053)	0.385** (0.041)	0.692** (0.049)
Non-White	0.341** (0.036)	0.032 (0.037)	−0.234** (0.034)
Level 1 N =	55,114	53,783	43,094
Level 2 N =	2,827	2,820	2,774
Reduction in error variance =	0.15	0.24	−0.04

Source: NAES 2000, 2004 and U.S. Census 2000, 2004 estimates.

p < .05. **p < .001.

The older adults (age 65 to 74) in the most aged locales were 7 percentage points more likely to pay attention to political news in the newspaper than their peers in the least aged places. The p value for this coefficient is .117. This relationship is not as strong as with the measure of local attention to political news on television, but it is still close enough to traditional levels of statistical significance to take note. Attention to political news among the oldest residents (age 75 and above) in the aged communities is statistically significant with a p value of .023. These oldest residents are 13 percentage points more likely to pay a great deal of attention to political news in the newspaper than their peers in the least aged communities.

The final model presents the results for the relationship between the aged community context and levels of newspaper reading. The emergent older adults in the most aged communities were 9 percentage points more likely to read the newspaper every day of the week than their peers living in the least aged locales. However, the coefficient for this relationship is a less impressive 0.189.

Discussion

The findings show that living in an older adult community is associated with greater levels of political knowledge for older residents—with some caveats. First, the oldest residents, those age 75 and over, did not receive the same boost to their political cognition as the emergent older adults, those between the ages of 65 and 74. Second, the aged context effect was not evident for every type of political knowledge examined. Knowledge of the political system and knowledge of candidates’ positions on senior citizen-related issues both seem to be influenced by the aged community environment, but this same effect was not found for knowledge of candidates’ positions for a wider variety of issues.

The first qualification is fully in line with the hypotheses proposed earlier in the article. The decline of political cognition, using a number of political knowledge measures, among those of the most advanced ages has been well documented in recent years (Lau & Redlawsk, 2006, 2008; Riggle & Johnson, 1996). Although aging neighborhoods have been shown to delay or prevent the onset of dementia in older adults (Clarke et al., 2012), the proposed mechanisms presuppose integration with a social network of peers and engagement with and participation in activities provided by the widely available senior-focused resources (Clarke et al., 2012; Fratiglioni et al., 2004; Fratiglioni, Wang, Ericsson, Maytan, & Winblad, 2000). Even with cognition intact later in older adulthood, the oldest of the old are more likely to be frail due to a combination of limitations and ailments associated with advanced aging (Hamerman, 1999). These impairments compromise the autonomy of the oldest members of society and leave fewer opportunities for social interaction that would encourage the maintenance of or prevent decline in political cognition (Hebert, 1997).

As for the second addendum, the findings fully support two out of the three hypotheses. Lau and Redlawsk (2008) showed different-sized effects for a few types of political memories. They find that age has less of an effect for shorter term candidate memories, compared with memories rooted in a lifetime of political experience. The present study made the same prediction for the aged context effect for political knowledge of candidates’ issue positions but found no relationship.

This article makes an important contribution to the study of aging and politics, by examining political knowledge especially relevant for older adults. The findings indicate that knowledge of candidates’ positions on issues specific to the concerns of senior citizens is greater among emergent older adults (age 65 to 74) residing in aged communities. These results make intuitive sense and provide additional insight for the proposed underlying mechanisms of the aged context.

Older adults living in places with very large older populations may be the target of interest groups focused on aging concerns and/or campaigns wanting to mobilize seniors based on these issues. Older residents in these contexts likely have greater opportunities for acquiring information on these issues and perhaps have the life experiences and skills necessary to take advantage of the unique political spheres in aged communities. Even less politically interested older adults may not be able to ignore the older-favored information supply (Huckfeldt, 1986). They may learn about these age-specific candidate positions simply by living in a place with an older cognitive content (Burbank, 1995).

This study also contributes to the growing body of work on the importance of the aged neighborhood context by providing a data-based examination of the underlying mechanisms. While the surveys used for this study did not include questions specifically asking about obtaining news on senior-related issues, measures of political discussion and local political news engagement were considered. Overwhelming support was lacking for both the theories of social interaction and density of resources, but the relationships between the aged context and the local engagement measures are worth further consideration.

The older adults living in the most aged communities did appear to be more engaged in local political news than their peers living elsewhere. The measures are certainly not ideal and do not necessarily provide substantial evidence that political resources and information are more abundant in the most aged locations, but they do point to a higher sense of political engagement in the communities among the older adults. This analysis is helpful in pointing future research in the direction of concentrations of political resources. Are the aged communities really hotbeds of political discourse, information, and/or resources?

It is telling that The Villages in Central Florida, one of the premier retirement locations in the nation, has become a popular campaign stop for candidates (especially Republicans) running for office in the last few election cycles (Erisman, 2011). Vice presidential candidate Paul Ryan visited The Villages with his mother in August of 2012 (Real Clear Politics, 2012).

A brief Internet search reveals that the presidential and vice presidential candidates have visited places like Wytheville (Virginia), Henderson County (North Carolina), Carbon County (Pennsylvania), Allegheny County (Pennsylvania), and Pinellas County (Florida) during the 2012 presidential election campaign. All of these places are located in swing/leaning states, but they are communities with significant aged populations. Not every aged community will receive a visit from a presidential candidate, but future research should include a systematic analysis of campaign visits (from candidates at all levels) and local media coverage to provide additional evidence for the political activities and resources available in the aged communities, compared with other locations.

In sum, the current work contributes significantly to the study of (a) the maintenance of political knowledge into older adulthood and (b) the impact of an aging community for cognition—by bringing them together. In sum, an aged context is associated with the political cognition of individual residents. These findings are encouraging as the country’s population grays and as communities are increasingly home to overwhelming numbers of senior citizens. The aged communities may act as political knowledge preservers, encouraging political engagement into retirement age. These communities may be particularly important for socializing their younger citizens into politically active dispositions, supporting democratic government well into the future.

This research supports the ongoing work showing that older communities may influence a host of political experiences, attitudes, and behaviors of the locals. The observed relationships are statistically significant and substantively important and quite robust, with very similar effects across two different measures of context. Even with this confidence in the findings, some will argue that the author has not ruled out the explanation of self-selection.

Perhaps older adults who settle into aged communities are already more knowledgeable, and the community context does not have an impact on knowledge levels. I think this is a valid concern for any study of context, especially when people choose where to live. However, I am not as concerned about selection effects for the current study because many older adults do not have the choice to move to the aged communities. As noted earlier in the article, many older people find themselves nested in an aged community by retiring in place. They do not necessarily select their communities, and may have little choice in the matter.

In order to fully address the question of self-selection, I reiterate that additional community level data should be collected. We must become as confident as we can that the aged communities are providing a unique political context. This study has certainly moved the body of research in that direction. While the models indicate that individual characteristics still play a very large explanatory role when predicting levels of political knowledge, the aged context also counts.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Notes

Author Biography

Brittany H. Bramlett is an Assistant Professor of political science at Albright College. Her research interests include religion and politics, political geography, civic education, political socialization and academic journal transparency.

References

Alford

R. R.

Scoble

H. M.

(1968). Sources of local political involvement. American Political Science Review, 62, 1192-1206.

Blum

T. C.

Kingston

P. W.

(1984). Homeownership and social attachment. Sociological Perspectives, 27, 159-180.

Burbank

M. J.

(1995). The psychological basis of contextual effects. Political Geography, 14, 621-635.

Cagney

(2006). Neighborhood age structure and its implications for health. Journal of Urban Health, 83, 827-834.

Clarke

P. J.

Ailshire

J. A.

House

J. S.

Morenoff

J. D.

King

Melendez

Langa

K. M.

(2012). Cognitive function in the community setting: The neighbourhood as a source of “cognitive reserve”? Journal of Epidemiology and Community Health, 66, 730-736.

Delli Carpini

M. X.

Keeter

(1993). Measuring political knowledge: Putting first things first. American Journal of Political Science, 37, 1179-1206.

Delli Carpini

M. X.

Keeter

(1997). What Americans know about politics and why it matters. New Haven, CT: Yale University Press.

Erisman

(2011, August 20). Politics in The Villages. The Villages Florida Book. Retrieved from http://www.thevillagesfloridabook.com/politics/

Fratiglioni

Paillard-Borg

Winblad

(2004). An active and socially integrated lifestyle in late life might protect against dementia. Lancet Neurology, 3, 343-353.

10.

Fratiglioni

Wang

H.-X.

Ericsson

Maytan

Winblad

(2000). Influence of social network on occurrence of dementia: A community-based longitudinal study. Lancet, 355, 1315-1319.

11.

Frey

W. H.

(2011). The uneven aging and “Younging” of America: State and metropolitan trends in the 2010 census. Washington, DC: Brookings. Retrieved from http://www.brookings.edu/research/papers/2011/06/28-census-age-frey

12.

Galston

W. A.

(1991). Liberal purposes: Goods, virtues, and diversity in the liberal state. Cambridge, UK: Cambridge University Press.

13.

Galston

W. A.

(2001). Political knowledge, political engagement, and civic education. Annual Review of Political Science, 4, 217-234.

14.

Hamerman

(1999). Toward an understanding of frailty. Annals of internal medicine, 130, 945-950.

15.

Hart

Atkins

Markey

Youniss

(2004). Youth bulges in communities the effects of age structure on adolescent civic knowledge and civic participation. Psychological Science, 15(9), 591-597.

16.

Hartley

(2006). Changing role of the speed of processing construct in the cognitive psychology of human aging. In Birren

J. E.

Schaire

K. W.

(Eds.), Handbook of the psychology of aging (6th ed., pp. 183-207). Amsterdam, Netherlands: Elsevier Academic Press.

17.

Hasher

Zacks

R. T.

(1988). Working memory, comprehension, and aging: A review and a new view. In Bower

G. H.

(Ed.), The psychology of learning and motivation: Advances in research and theory (Vol. 22, pp. 193-225). London, UK: Elsevier Academic Press.

18.

Hebert

(1997). Functional decline in old age. Canadian Medical Association Journal, 157, 1037-1045.

19.

Hess

T. M.

(2005). Memory and aging in context. Psychological Bulletin, 131, 383-406.

20.

Hoyer

W. J.

Verhaeghen

(2006). Memory aging. In Birren

James E.

Schaie

K. W.

(Eds.), Handbook of the psychology of aging (6th ed.). San Diego, CA: Elsevier Academic Press.

21.

Huckfeldt

R. R.

(1986). Politics in context: Assimilation and conflict in urban neighborhoods. New York, NY: Algora.

22.

Lang

I. A.

Llewellyn

D. J.

Langa

K. M.

Wallace

R. B.

Huppert

F. A.

Melzer

(2008). Neighborhood deprivation, individual socioeconomic status, and cognitive function in older people: Analyses from the English longitudinal study of ageing. Journal of the American Geriatrics Society, 56, 191-198.

23.

Lau

R. R.

Redlawsk

D. P.

(2006). How voters decide: Information processing in election campaigns. New York, NY: Cambridge University Press.

24.

Lau

R. R.

Redlawsk

D. P.

(2008). Older but wiser? Effects of age on political cognition. Journal of Politics, 70, 168-185.

25.

Longino

C. F.

McClelland

K. A.

Peterson

W. A.

(1980). The aged subculture hypothesis: Social integration, gerontophilia and self-conception. Journal of Gerontology, 35, 758-767.

26.

Mata

Schooler

L. J.

Rieskamp

(2007). The aging decision maker: Cognitive aging and the adaptive selection of decision strategies. Psychology and Aging, 22, 796-810.

27.

Nie

N. H.

Verba

Kim

(1974). Political participation and the life cycle. Comparative Politics, 6, 319-340.

28.

Plutzer

(2002). Becoming a habitual voter: Inertia, resources, and growth in young adulthood. American Political Science Review, 96(1), 41-56.

29.

Quillian

M. R.

(1968). Semantic memory. In Minsky

, (Ed.), Semantic information processing (pp. 216-270). Cambridge, MA: MIT Press.

30.

Real Clear Politics. (2012, August). Paul Ryan addresses The Villages with his mother. Real Clear Politics. Retrieved from http://www.realclearpolitics.com/video/2012/08/18/paul_ryan_addresses_the_villages_with_his_mother.html

31.

Riggle

E. D. B.

Johnson

M. M. S.

(1996). Age difference in political decision making: Strategies for evaluating political candidates. Political Behavior, 18(1), 99-118.

32.

Schaie

K. W.

(2005). Developmental influences on adult intelligence: The Seattle longitudinal study. New York, NY: Oxford University Press.

33.

Seeman

T. E.

Lusignolo

T. M.

Albert

Berkman

(2001). Social relationships, social support, and patterns of cognitive aging in healthy, high-functioning older adults: MacArthur studies of successful aging. Health Psychology, 20, 243-255.

34.

Smith

E. E.

Shoben

E. J.

Rips

L. J.

(1974). Structure and process in semantic memory: A featural model for semantic decisions. Psychological Review, 81, 214-241.

35.

Stern

(2006). Cognitive reserve and Alzheimer Disease. Alzheimer Disease & Associated Disorders, 20(2), 112-117.

36.

Strate

J. M.

Parrish

C. J.

Elder

C. D.

Ford

(1989). Life span civic development and voting participation. American Political Science Review, 83, 443-464.

37.

Sykes

G. M.

(1950). The differential distribution of community knowledge. Social Forces, 29, 376-382.

38.

Taylor

Gorard

Fitz

(2003). The modifiable areal unit problem: Segregation between schools and levels of analysis. International Journal of Social Research Methodology, 6(1), 41-60.

39.

Wang

H.-X.

Karp

Winblad

Fratiglioni

(2002). Late-life engagement in social and leisure activities is associated with a decreased risk of dementia: A longitudinal study from the Kungsholmen project. American Journal of Epidemiology, 155, 1081-1087.

40.

Ward

R. A.

LaGory

Sherman

S. R.

(1985). Neighborhood and network age concentration: Does age homogeneity matter for older people? Social Psychology Quarterly, 48(2), 138-149.

41.

Wight

R. G.

Aneshensel

C. S.

Miller-Martinez

Botticello

A. L.

Cummings

J. R.

Karlamangla

A. S.

Seeman

T. E.

(2006). Urban neighborhood context, educational attainment, and cognitive function among older adults. American Journal of Epidemiology, 163, 1071-1078.

42.

Williams

R. W.

(1999). The contested terrain of environmental justice research: Community as unit of analysis. Social Science Journal, 36, 313-328.

43.

Wolf

(2001). Hot towns: The future of the fastest growing communities in America. Piscataway, NJ: Rutgers University Press.

44.

Wolfinger

R. E.

Rosenstone

S. J.

(1980). Who votes? New Haven, CT: Yale University Press.

45.

Zarit

S. H.

Zarit

J. M.

(2007). Mental disorders in older adults: Fundamentals of assessment and treatment (2nd ed.). New York, NY: Guilford Press.