Gerontographics and consumer behavior in later life: Insights from the life course paradigm

The present article applies the life course paradigm to the study of older consumers. It is based on the general life course premise that events experienced in later life in the form of biophysical, social, and psychological changes create demand for readjustment and adaptation to new life conditions that define the multidimensional processes of aging and influence consumer behavior. These aging processes are collectively integrated into a model to develop stages of aging, known as “gerontographics,” that people go through in later life. To test the model’s efficacy in predicting consumptionrelated activities, a convenience sample (N = 383) of adults aged 45 and older is used to develop the gerontographics model and compared it to commonly used measures of aging (chronological and cognitive age) in predicting select consumer behaviors relevant to people in later life. The results reveal the value of the gerontographics model in understanding and explaining the consumer behavior of older adults. Implications of these findings and directions for further research are also discussed.

Gerontographics and consumer behavior in later life:

Insights from the life course paradigm

George P Moschis a, Anil Mathur band Thuckavadee Sthienrapapayut c

Marketing and International Business, Frank G Zarb School of Business, Hofstra University, Hempstead, NY,

ABSTRACT

The present article applies the life course paradigm to the study of

older consumers It is based on the general life course premise that

events experienced in later life in the form of biophysical, social, and

psychological changes create demand for readjustment and

collectively integrated into a model to develop stages of aging, known

conve-nience sample (N = 383) of adults aged 45 and older is used to develop

the gerontographics model and compared it to commonly used

mea-sures of aging (chronological and cognitive age) in predicting select

consumer behaviors relevant to people in later life The results reveal

the value of the gerontographics model in understanding and

find-ings and directions for further research are also discussed.

老年图表学与晚年消费行为: 以生命历程范式为视角

而言, 晚年经历的一些事情会让人的生理、社会和心理产生变化,

这些变化要求人们重新调整, 适应新的生活条件, 这些条件明确了

老过程, 整合为一个模型, 将人们的老年生活划分为各个衰老阶段,

果, 研究采用45岁及以上成人的便利样本（数量: 383）来开发“老

年图表学”模型, 将其与衰老的常用度量（生理年龄和认知年龄）

显示出了老年图表学模型在理解和解释老年人消费行为方面的价

ARTICLE HISTORY Received 31 January 2019 Revised 15 February 2019 Accepted 28 February 2019 KEYWORDS

Gerontographics; aging; life course; older consumers; market segmentation 关键词

老年图表学；衰老；生命 历程；老年消费者；市场 细分

1 Introduction

The study of older consumers’ behavior and well-being is becoming of increasing interest to several groups of people, including social workers, marketers, educators, and policy makers (e.g., Drolet, Schwarz, & Yoon, 2010) Efforts to understand

CORRESPONDENCE TO Anil Mathur Anil.Mathur@hofstra.edu

https://doi.org/10.1080/21639159.2019.1613908

consumer behavior in later life fall into two broad categories: descriptive studies that report observed differences among older consumers possessing various attributes such

as demographics (e.g., Moschis, 1992) and psychological characteristics (e.g., age iden-tity, values) (e.g., Sudbury & Simcock, 2009a), and studies that examine age-related differences in behaviors using aging theories and models (e.g., Mathur, Ong, Fatt, Rakrachakarn, & Moschis, 2017; Moschis, 2000; Sudbury & Simcock, 2009b) Descriptive studies, albeit useful in profiling consumers and measuring market beha-vior, provide little explanation for the observed differences For example, it is not clear that the observed age-related differences in consumer behavior are the result of aging, period, or cohort (Moschis, 2012)

In contrast, the vast majority of studies of older consumers attempt to explain differences in their consumption habits by using age as a proxy measure of the aging processes, which are believed to explain the observed differences across consumers in different age groups They assume that those theories that were developed to explain how and why the body, mind, and roles people occupy change with age can also adequately explain differences or changes in behaviors They infer that the observed age differences in consumer behaviors are due to differences in the aging processes that people experience (Moschis, 2012) The notion that aging, whether biological, psycho-logical, or social changes, also affects consumer behavior has not been rigorously tested (Moschis, 2000, 2012) In addition, people develop strategies to compensate for many of these changes (Moschis, Mosteller, & Kwaifatt, 2011; Salthouse, 2010)

Efforts to use theory-based measures that tap aging to explain consumer behavior in later life tend to be limited to a single aging process, such as psychological (self-perceived age in particular), when it is widely acknowledged that aging is multi-dimensional––i.e., psychological, biophysical, social, and even spiritual (e.g., Mathur

et al., 2017; Moody, 1988; Moschis, 2012; Sudbury & Simcock, 2009b) And when measures of aging dimensions are used in developing segmentation models and explaining consumer behavior (e.g., Sudbury & Simcock, 2009b), their explanatory power is not compared to other theory-based measures or models reported in the literature (for some exceptions, see Mathur et al., 2017; Sthienrapapayut, Moschis, & Mathur, 2018)

In sum, studies have been limited in that they either tend to infer changes in behavior as a result of chronological age, which is used as surrogate measure of aging processes, or because measures of aging processes that are linked to consumption rely heavily on psychological dimensions of aging, largely ignoring the social, biophysical, and spiritual dimensions of aging processes Age per se has long been viewed as

“empty” or “worthless” variable has little explanatory power (e.g., Settersten & Godlewski, 2016; Settersten & Mayer, 1997) because it merely measures the passage

of time As aging is not uniform among people, the assumption that all people of a certain age behave the same because of their uniform experiences of aging is not valid Within-group differences are often greater than between group differences in aging (Salthouse, 2010) Most previous studies do not account for individual differences in aging processes, nor do they directly link measures of such processes to consumer behaviors, nor do they compare the efficacy of measures of such processes and models derived from them to other measures and models

In view of these limitations, researchers have developed measures of the main dimensions of aging––i.e., biophysical, social, and psychological––that have been used

to explain consumer behavior These measures have been named “gerontographics” (Moschis, 1992, 1996); they draw heavily from a relatively recent approach that is widely used across disciplines known as “life course paradigm” (Moschis, 1996, 2019a, 2019b), which is considered as the most important development in social and behavioral sciences (Colby, 1998)

However, the gerontographic measures used differ across studies, raising validity issues Because previous studies do not show how each the measures of the three main types of aging affects behavior, they have yet to empirically demonstrate that the observed differences in consumer behavior of older adults are due to differences in the three main aging processes they experience The present study attempts tofill these gaps in the literature It tests relationships between select consumer behaviors most relevant to people at later stages in life and the three measures of aging that form the bases of the gerontographics model Second, it assesses the efficacy of the three measures in predicting consumption-related activities relevant primarily to older adults

by comparing them to commonly used measures of age and cognitive age

2 Background

“Gerontographics” is a concept that taps individual differences in aging processes (Moschis, 1996) It was coined by researchers at the Center for Mature Consumer Studies to explain consumption-related activities of people in later life as manifestations

of individual differences in the aging processes and life circumstances they experience The increasing heterogeneity in these aging processes is assumed to also affect con-sumer needs, motives, attitudes and behaviors in general, as people age differently biologically, psychologically, and socially, and at different rates The theories of aging that explain variability in the aging processes and the presumed resultant differences in consumer behaviors are incorporated within the broader life course paradigm (e.g., Elder & Johnson, 2002; Moschis, 2019a, 2019b) According to this multi-theoretical paradigm, life events or changes which people experience at different stages in life serve

as markers of biological, psychological, and social aging The timing and circumstances

of these experiences account for individual differences in the aging processes that explain the variability in consumer behaviors (Moschis, 2019a)

Thus, the term “gerontographics” refers to a person’s experience of the aging processes as evidenced by the extent to which he or she has acquired the characteristics associated with old age, such as chronic conditions (e.g., Kaufman & Elder, 2002) or displays patterns of thought and actions that characterize old age and old age-related roles (Herzog & Markus, 1999) And the gerontographics approach to the study of consumer behavior is based on the assumption that differences in consumer behavior are the result of differences in the aging processes in the form of biological, social, and psychological changes that are not uniformly experienced at a specified chronological age (Moody, 1988) For example, research has linked consumer behaviors to interna-lization and acceptance of old age status as a result of biological and life-stage changes (e.g., Mathur & Moschis, 2005; Moschis, 2012)

The gerontographic measures that are derived from the life course paradigm have consistently produced four life-course stages, which are defined in terms of the aging processes people experience in later life (Figure 1) These stages describe segments of consumers with different needs for products and services (for description of these segments, see e.g., Moschis, 1996; Moschis, Lee, & Mathur, 1997) Since its introduction

to thefield of marketing 30 years ago (e.g., Marketing Communications, 1988), this life-stage model has been applied across industries and countries And the gerontographic-based segments have been found better predictors than age-gerontographic-based segments with regard

to older consumer responses to marketing stimuli in general, as well as preferences for specific products and services such as financial services, travel and leisure services, homefire safety services, long-term care insurance, apparel and footwear, brands and department stores, food and grocery stores, restaurants, retirement communities, hous-ing, healthcare products and services, and use of the Internet (e.g., Moschis, 1996; Moschis et al., 1997; Correia & Elliot, 2006; Sthienrapapayut et al., 2018)

However, previously-used measures of psychological and social aging tend to mea-sure psychological aging (e.g., Moschis, 1996; Sthienrapapayut et al., 2018), raising issues of convergent and discriminant validity because psychological aging may or may not be the result of experienced and anticipated life transitions (Sthienrapapayut

et al., 2018) In a similar vein, subjective measures of biological aging (e.g., perceived health) are not likely to capture this dimension of aging, as people adapt to their

Ailing outgoers

Healthy

s

Healthy hermits

Time

Figure 1 Life-course stages of aging consumers

Speci fic arrows indicate that people may move to the next stage in life due to physiological, psychological, and social aging Arrows pointing downwards denote psychosocial aging, and arrows pointing upwards denote biophysical aging Source: Moschis (2019a).

frailties and continue to report little change in their health well into late life (Moschis, 1992) To overcome such threats to validity, we operationalize the three aging process differently We use objective measures of biological aging to assess deterioration of the person’s bodily systems, of social aging that measure the person’s actual and anticipated experience of life transitions into old-age roles (e.g., retiree, empty nester, grandparent), and measures of psychological aging as evidenced by the aging person’s awareness of his or her increased proximity to the end of life The latter conception, which we label

“old-age identity,” differs from those of subjective or cognitive age, which are subject to the individual’s interpretation of his or her status and to change with contexts (Guido, Amatulli, & Peluso, 2014; Moschis, 2012; Settersten & Godlewski, 2016)

3 Hypotheses

The life course paradigm assumes that behaviors at any stage in life can be viewed either

as responses in the form of choices and adjustments to present and anticipated life conditions, or as products of one’s adaptation to various events and circumstances previously experienced (see, e.g., Moschis, 2019b) Adaptation and adjustments to events that people experience and define aging result in changes in behaviors These events take various forms Some are in the form of biological changes (e.g., onset of chronic conditions) while others mark transitions into social roles (e.g., retiree, grand-parent) Regardless of their nature, life events create emotional demands (e.g., stress) to which people must adapt (Moschis, 2019a, 2019b) by changing their mindsets, personal goals, and behaviors These changes may reflect the person’s enactment of newly acquired social roles, efforts to cope with stress, or the development of new skills or knowledge, with three theoretical perspectives––normative, stress, and human capital–– offered as explanations for changes, respectively (e.g., Mathur, Moschis, & Lee, 2008; Moschis, 2019a, 2019b; Moschis & Ong, 2012)

According to the life course paradigm, expected or unexpected changes people experience or anticipate and choices they make, including environmental, biological, psychological, and social demands during an individual’s life, define typical life events and social roles that serve as turning points and affect behavior (e.g., Elder, 1998; Elder

& Johnson, 2002) These events also defining the three widely-accepted types of aging processes: biological, social, and psychological aging (e.g., Moody, 1988; Moschis, 1992)

3.1 Biophysical aging

Biological aging in the form of changes in the bodily systems (e.g., onset of chronic conditions) are experienced with the passage of time in later life (Moschis, 1992) Such changes signify transition into old age and affect the older person’s ability to perform consumption-related tasks For example, the onset of a chronic condition, such as diagnosis of hearing deficit or loss, is associated with old age and a sense of frailty; it signifies transitions into a physically-impaired role (i.e., “hearing-impaired”) (Moschis, Lee, Mathur, & Rigdon, 2015) Such biological changes have been linked to the inter-nalization and acceptance of the “old age” status that subsequently affect consumer behaviors (e.g., Mathur & Moschis, 2005; Moschis, 2012) Similarly, deterioration of bodily systems (e.g., perceptual, visual) is a sign of biophysical aging that affects the

older person’s ability to function effectively in the marketplace (e.g., read information

on packages) (Moschis, 1992) These changes suggests the following hypothesis: H1: The deterioration of the bodily systems is positively associated with the person’s engagement in old-age consumer behaviors

3.2 Social aging

Life events which signify transitions into social roles that are ascribed to old age––i.e., old-age-related roles, such as retiree and grandparent––affect the old person’s identity (Elder, 1998) For example, events such as retirement and becoming a grandparent that signify transition into the old-age roles of retiree and grandparenthood affect a person’s age identity, producing an older cognitive age (Mathur & Moschis, 2005) and accel-erating “the pace of aging, including a sense of being old” (Elder & Johnson, 2002, p 67) Such transitions may begin not only upon experiencing the transitional event, but also in anticipation of the event (Mathur et al., 2008; Moschis, 2019a, 2019b) Thus, adaptation to experienced and anticipated life events entails changes in one’s mindset as

to how the person feels about his or her age and about behaviors appropriate for old-age-related roles Therefore, it is hypothesized that:

H2: The older person’s propensity to engage in old-age consumer behaviors is positively related to the number of events that signify transition into old-age roles the person has experienced

H3: The older person’s propensity to engage in old-age consumer behaviors is positively related to the number of events that signify transition into old-age roles the person anticipates to experience in the near future

3.3 Psychological aging

People at later stages in life tend to differ with respect to perceptions of their own aging and old-age identity, as they do not experience events that signify the arrival of old age at the same chronological age Within the life course paradigm, the contextual view of human capital that characterizes developmental stages points to the emotional and mental states brought about by psychosocial crises (events) associated with aging, both experienced (e.g., physical and social losses) and anticipated (e.g., end of life) that require adaptation (Moschis, 2019a; Thomae & Lehr, 1986) The perceived loss of control over the aging person’s environment makes the employment of problem-focused strategies a riskier proposition (Heckhausen & Schulz, 1995) As a result, the effectiveness of strategies that help the elderly cope with age-related losses and the approaching end of their lives (Belk, 1988) tend to be replaced with emotion-focused coping strategies, such as religiosity and spirituality (Lazarus & Folkman, 1984) as well as preservation of“self” beyond death (Belk, 1988) Such strategies reflect psychological aging and define one’s old-age identity, as they

underscore the aging person’s awareness of his or her increased proximity to the end of life (Moschis, 2019a) Therefore, it is hypothesized that:

H4: The stronger the person’s old-age identity the greater his or her likelihood of engagement in old-age-related behaviors

4 The study

4.1 Sample

Workers enrolled in Amazon’s Mechanical Turk marketplace panel were asked to participate

in the study They were offered $1.00 as a compensation for completing the survey, which took approximately 10 min to complete Only those panel members who were residents of the United States were invited to participate The survey was made available to specific age groups

in stages to get respondents from all age groups in the sample Completed responses were obtained from 404 participants However, since we were interested in studying behaviors usually associated with later life, only those participants who were 45 years of age or older and had complete responses to study questions were included in the study for test of hypotheses A total of 383 respondents who satisfied the age requirement and had complete responses and were included in the analysis

Although some researchers have questioned the representativeness of samples drawn from

a MTurk panel of workers and the quality of data obtained from them because of concerns related to character misrepresentation (cf., Goodman & Paolacci, 2017), in this study we took steps to minimize negative effects and improve data quality Goodman and Paolacci (2017) suggest several practical steps for users of MTurk panel so that potential negative effects could

be minimized Many of those suggestions were incorporated in this study Study participants were assured of anonymity and that there were no right or wrong answers to any of the statements used in the questionnaire Moreover, study participants were not aware of any possible relationships among survey questions Finally, the survey included several items that asked participants to input data from the keyboard (rather than just clicking on the box) (e.g., what is the year of your birth?) to minimize monotony and maintain participants’ interest in the survey Although sample obtained from the panel of Mechanical Turk workers cannot be considered a representative of the total U.S population, it is generally acceptable because it produces adequate variance for hypothesis testing (Buhrmester, Kwang, & Gosling, 2011) And we had no reason to expect the consumption activities and aging processes of individuals

in our sample to be different from those of the general population

There was diversity in the sample in terms of demographic characteristics The sample was somewhat balanced in terms of gender (females = 53.8%, males = 46.2%) Almost one third of the respondents were between 45 and 54 years of age In addition, 19.3% of respondents were between 55 and 64 years of age, 43.8% were between 65 and

74 years of age and 3.7% were 75 years of age and older All income and education groups were adequately represented in the sample Finally, in terms of employment status, the vast majority of respondents (61.0%) reported that they were employed full-time or part-full-time (including those who were employed after retirement) and 39% of the respondents indicated that they were retired or not employed

4.2 Operational definitions and measures

Old-age consumer behaviors refer to cognitive and overt consumption-related activities relevant to people in later life We relied on studies that report on activities that older people engage in, compared to their younger counterparts (e.g., Moschis, 1992; Moschis

& Mathur, 2007) Our review of the literature revealed 12 behaviors (see the Appendix) For each, respondents were asked to indicate whether they had done it, thought about

it, or had not done it or thought about it, assigning the numbers of 2, 1, and 0 to measure their responses, respectively Scores were summed across all 12 behaviors to form a 0-to-24-point index

Deterioration of the bodily systems refers to aging of the bodily systems As the various bodily system begin to age, they become less efficient and require medication or assistive devices (e.g., hearing aid) We used the number of drugs presently used for chronic conditions as a proxy variable to measure the presence of deterioration or a lack of optimal functioning of the various bodily systems

Events that signify transition into old-age roles that people experience reflect the extent to which a person has experienced social aging Five events that signify transition into various roles that define roles common to people at later stages in life were used: retirement, birth of first grandchild, last child leaving home, first chronic condition diagnosed, and death of the second parent to measure transitions into the roles of retiree, grandparent, empty nester, physically-impaired, and parentless, respectively Affirmative responses to the five events were used to form a 0-to-5-point index of old-age roles experienced A similar approach was used to measure transition into anticipated old-age roles, using three life events: empty nest, grandparenthood, retirement Affirmative responses to these events that respondents expected to experience in the following 12 months were used to form a 0-to-3-point index

of anticipated old-age roles

Old-age identity was our operational measure of psychological aging We draw from research in consumer behavior, gerontology and social sciences that reveal cognitive, emotional, and behavioral changes that tend to occur with advancing age in later life, especially with anticipation of increasing proximity to the end of life (event) (e.g., fear

of deteriorating health and dependency on others, stronger spirituality, desire for extending one’s self beyond death) (e.g., Belk, 1988; Charles & Carstensen, 2010; Coleman, Schröder-Butterfi, & Spreadbury, 2016; Moody, 1988; Moschis, 2012) We developed items to tap these changes in emotions, cognitions, and behaviors For example, the spiritual dimension of aging was reflected in the statement, “Nowadays,

I find that I rely more on my spirituality and faith” (Coleman et al., 2016); and the meaningful emotional interactions in the statement “Nowadays, I value relationships with my loved ones more than I did 10 years ago” (Charles & Carstensen, 2010) Exploratory factor analysis revealed four items that appear to relate to the psychological dimension of aging (see the Appendix) The alpha reliability coefficient was 60, which

is lower than the minimum accepted level of 70 for established measures (Nunnally, 1978) Although this latter value is not a particularly high level of internal consistency, scales with alpha values in the 60 to 70 range are still deemed to be acceptable for research purposes (Lance, Butts, & Michels, 2006), especially for development of measures at the exploratory stage of their development (Nunnally, 1967)

Chronological age was measured by asking respondent to indicate their birth date The cognitive age’s measure was identical to that used in previous studies (e.g., Mathur & Moschis, 2005) (see the Appendix) Finally, due to the sensitivity of age and aging-related questions, we used a short version of social desirability scale (Strahan & Gerbasi, 1972)

5 Analysis and results

5.1 Hypothesis testing

Partial correlations we computed for all study variables by controlling for social desirability The results of the resultant correlations are shown in Table 1 These correlations were also used in testing the hypothesized relationships

Hypothesis 1 posits a positive relationship between deteriorations of the person’s bodily systems and his or her propensity engage in old-age related consumer behaviors The correlation coefficient between these two variables was positive but not statistically significant (r = 058, n.s.), providing no support for Hypothesis 1

Hypothesis 2 posits a positive relationship between the number of transitions into old age roles, as measured by the number of events experienced that signify such transitions, and the person’s propensity to engage in old-age consumer behaviors The relationship between these two variables was significant (r = 0.286, p < 001), providing strong support for Hypothesis 2

Hypothesis 3 posits a positive relationship between the older person’s propensity to engage in old-age consumer behaviors and the number of events that signify transition into old-age roles the person anticipates to experience in the near future The partial correlation between these two variables was not significant (r = 0.003, n.s.), providing no support for this hypothesis Finally, Hypothesis 4 predicts a positive relationship between the person’s old-age identity and his or her likelihood of engagement in old-age-related behaviors The resultant correlation between these two variables was positive and significant, as expected (r = 0.312, p < 001), providing strong support for Hypothesis 4

These results suggest that biophysical aging, as measured by the number of prescrip-tion drugs used, does not predict the person’s old-age-related consumer behaviors used in this study In contrast, psychological aging (as measured by old-age identity) and social aging (as measured by the events signifying transitions into old age roles) are strong predictors of old-age consumer behaviors Anticipation of transitions into old age roles apparently does not lead to engagement in old-age consumer behaviors, suggesting that the aging person may defy the assumption of an old-age identity in view of forthcoming events that signify transition into old age status

5.2 Comparison of alternative models

It has been argued that cognitive age, which is a domain of psychological aging––i.e., self-concept in terms of age is another surrogate measure of passage of time (Moschis, 2012) If these views are valid, our gerontographics model that is defined by the three aging processes should be a stronger predictor of consumer behaviors in later life, compared to chronological and cognitive age Therefore, a major objective of the

Table