Subjective Control and Health Among Mexican-Origin Elders in Mexico and the United States: Structural Considerations in Comparative Research

Abstract

Objectives

This study examines the joint impact of psychological and structural factors on Mexican and Mexican American elders' sense of personal control over important aspects of their lives and health in Mexico and the United States.

Methods

We employ the Mexican Health and Aging Study (MHAS) and the Hispanic Established Populations for Epidemiologic Studies of the Elderly (H-EPESE) to explore patterns of association among structural factors, personal characteristics, indicators of material and physical vulnerability, and expressed locus of control.

Results

The results suggest that an older individual's sense of personal control over important aspects of his or her life, including health, reflects real material and social resources in addition to individual predispositions. In Mexico, only the most privileged segment of the population has health insurance, and coverage increases one's sense of personal control. In the United States, on the other hand, Medicare guarantees basic coverage to the vast majority of Mexican Americans over 65, reducing its impact on one's sense of control.

Discussion

Psychological characteristics affect older individuals' sense of personal control over aspects of their health, but the effects are mediated by the economic and health services context in which they are expressed.

DURING the middle part of the previous century, the humanistic psychologist Abraham Maslow developed his theory of a hierarchy of human needs that posits that each of us is confronted with a series of developmental tasks that involve satisfying ever more complex social and personality needs beginning with the most basic of physiological needs and progressing through higher-level needs, such as the need for safety and security, love and belonging, the esteem and respect of others, and self-esteem. If, and only if, these more basic needs are satisfied does one arrive at the stage of self-actualization, which consists of a complex mix of traits, attitudes, and behaviors that express one's complete and engaged humanity (Maslow, 1962). As an ideal goal, or even as a metaphor for the good life, Maslow's stage of self-actualization reflects the greatest possible generativity, social engagement, and emotional maturity that the individual can achieve, and by extension, a stage at which one has subjective control over important aspects of one's life.

Maslow's theory remains appealing not because of its ability to make specific empirical predictions but because of its humanistic focus that draws attention to the major obstacles that marginalized and powerless individuals face in their progress toward full material and psychological security. We employ Maslow's theory rhetorically in order to frame the objective of this study, which is to make the point that successful aging and the potential for optimal development depends not only on individual characteristics but also on the particular economic, political, and social contexts in which an individual spends his or her life.

Our basic theoretical perspective views successful aging as dependent both on agency and structure. Individual characteristics and behaviors are clearly important in determining one's health, but health care system characteristics and the ability to access high-quality care are essential if one is to achieve optimal health in later life. Optimal physical and mental health depends on adequate material resources, both to prevent illness and to regain one's health when one falls ill. As part of our empirical analysis, we examine the extent to which various health outcomes, as well as one's sense of control over important aspects of one's life and health, reflect individual-level factors and structural factors related to economics and the health care system in two very different national contexts, Mexico and the United States. Rather than focusing on specific scores or attempting to make specific cardinal comparisons, which we find debatable in comparative research in general, our objective is to determine the extent of structural similarity among predictors of health and one's sense of control among older Mexicans and Mexican-origin U.S. residents. Ultimately, the purpose of the study is to theorize about and examine patterns of self-reported subjective health and individuals' sense of personal control over health in two very different socioeconomic contexts while holding aspects of ethnicity and culture constant.

THE MEANING AND SOCIAL CONTEXT OF SUCCESSFUL AGING

The construct of “successful aging” was developed primarily by psychologists and is largely defined in psychological terms. In general, researchers identify several functional domains as important components of successful aging (Havighurst, 1961; House, 2002; Neugarten, Havighurst, & Tobin, 1961). Among the most important of these are (a) a high level of cognitive functioning, (b) resilience and the ability to adapt to the losses and physical decline associated with aging, and (c) social integration and engagement. The general concept often includes a more general domain that taps (d) life satisfaction and self-actualization. A major theoretical focus in this tradition defines successful aging in terms of successful development, adaptation, and coping, as well as the achievement of one's peak level of physical, social, and psychological well-being (Berkman & Mullen, 1997; Fisher, 1992; Gibson, 1995; Palmore, 1995; Ryff, 1989; Seeman et al., 1994). For purposes of this discussion, we define successful aging broadly as optimal physical functioning, emotional well-being, and social engagement, but we also add a material dimension, as we explain below.

Most conceptual models of successful aging do not explicitly include, or they assume as implicit, an adequate level of material resources (Rowe & Kahn, 1987). Yet, it is critical to explicitly deal with structural factors as determinants of successful aging, especially when dealing with marginalized and poor groups or in comparative research that includes populations at different levels of development. As Farmer (1999) and other social scientific investigators observe, although marginalized groups may not share a common culture or a similar health care system, what they do have in common is poverty. The importance of material and structural factors in optimal aging is clearly revealed by the literature on racial and ethnic differences in the well-being of the elderly. Minority Americans have lower lifetime incomes and arrive at mature adulthood with far less wealth than majority Americans. As a consequence, they are at high risk of dependency on others in old age (R. J. Angel & J. L. Angel, 2006). A large literature shows that for structural reasons, these individuals often lack the ability to satisfy the more basic deficiency needs in the Maslowian hierarchy.

As is the case with the other dimensions of the concept of successful aging, the role of material well-being is complex and includes both absolute and relative components. An income of less than $2 a day represents the worst of absolute poverty and clearly robs one of the ability to age successfully or even to survive (Farmer, 2001). Incomes that are higher, but that place one in the lower tiers of the income distribution, can leave one with few choices in living arrangements and few possibilities for expressing one's generativity materially. The lack of material resources precludes the possibility of the sort of generativity that those who can help children buy houses or pay for a grandchild's college education enjoy. In both cases, the individual lacks the material basis of a satisfying and productive old age.

LOCUS OF CONTROL

Our focus on one's subjective sense of control emerges from the theoretical and empirical literature on locus of control. In recent years, interest in understanding the relationship between the concept of locus of control and a variety of different health attitudes, behaviors, and outcomes has grown (Antonucci, 2001; Black, Markides, & Miller, 1998; Hunt, Valenzuela, & Pugh, 1998; Lachman & Weaver, 1998a, 1998b; Norman, Bennett, Smith, & Murphy, 1998; Wickrama, Surjadi, Lorenz, & Elder, 2008). Also referred to as “personal control beliefs,” this theoretical construct refers to an individual's beliefs regarding the extent to which he or she is able to control or influence important life events, including health outcomes. As defined by Rotter in the mid-1960s, the construct is based on social learning theory and posits two major orientations that might be seen as two ends of a continuum that characterize individuals' beliefs concerning the causes of various outcomes. One end of the continuum is characterized by individuals with an internal locus of control profile. These individuals attribute events or outcomes in their lives to their own efforts. At the other extreme, those with an external locus of control profile attribute these same events and outcomes to external forces including chance and supernatural forces. Most individuals fall somewhere between these two extremes.

The more general concept of locus of control was applied specifically to health by Wallston, Wallston, Kaplan, and Maides (1976), who developed the concept of Health Locus of Control (HLC) in the 1970s. This concept focuses on the degree to which individuals believe that their health is largely under their own control rather than depending on external forces or chance. Two of the most well known scales used to measure this construct are the Health Locus of Control Scale and the Multidimensional Health Locus of Control Scale (Lachman & Weaver, 1998b; Wallston, 2005). The scale includes items such as “if I get sick, it is my own behavior which determines how soon I get well again,” “no matter what I do, if I am going to get sick, I will get sick,” “health professionals control my health,” and “if I take the right actions, I can stay healthy.” Responses range from “strongly agree” to “strongly disagree” on a Likert scale. In the last three decades, a large body of research has linked an internal locus of control profile to positive health behaviors and greater knowledge about the causes of age-related disease (Krause, 1986, 1987, 2007; Lefcourt & Davidson-Katz, 1991; Norman et al., 1998; Shaw & Krause, 2001).

The concept and its measures have proved to be powerful predictors of positive health outcomes. Older adults who feel that they have greater subjective control over salient roles, such as parenting, grandparenting, and being a friend, homemaker, caregiver, volunteer, and church or club member, live longer than those who lack this sense of control (Krause, 2007). These favorable mortality experiences may reflect the positive health behaviors that are associated with greater internal locus of control. Those individuals with a lower sense of control are more likely to engage in behaviors such as drinking alcohol to excess and smoking, and they are more likely to be obese than those with a greater sense of control (Lachman & Weaver, 1998a). In contrast to individuals with an internal locus of control, those with an external locus report lower levels of physical and mental well-being and cope poorly with stress and financial strain (Krause, 1987). Certain institutional factors appear to reduce a sense of control over aspects of one's life, or to at least have this association (Shaw & Krause, 2001).

GENOTYPE OR PHENOTYPE

The construct of locus of control implies the existence of a personality trait, either inborn or learned through experience or, perhaps, both. One is born with certain genotypic predispositions, and those are expressed as the result of the organism's interaction with the environment. Characterized as a trait, the implication is that such a behavioral predisposition is fairly structured and stable and that it can be used to predict particular outcomes or to explain differences among people and even groups. David McClelland's (1961) needs theory is an example. According to McClelland, individuals can be characterized as manifesting different levels of three need traits: the need for achievement, the need for affiliation, and the need for power. One's particular mix may be based on fundamental characteristics, but it is also shaped over time in response to one's life experiences. Those experiences include the values one learns from one's culture, and McClelland's theory is very similar to Max Weber's attribution of capitalist acquisitiveness to Calvinist religious values (Weber, 1930). McClelland employed the theory to explain differences in levels of development among nations (McClelland). The use of such supposed psychological traits to characterize groups or to explain highly complex social, economic, or political differences, though, represents a potentially serious attributional error. More specifically, it runs the risk of attributing the effects of structural factors to culture and group beliefs. We suspect the same risk applies to all supposed psychological traits that are used to explain group differences in behaviors.

Other personality theorists question the existence or the stability of such personality traits, especially as they relate to emotional characteristics. Over a quarter of a century ago, Mischel (1969), for example, pointed out that supposed emotional personality characteristics are highly contextual and that one responds very differently to emotional stimuli in different social contexts. To the extent that locus of control includes emotional components or is shaped by emotional experience, it too is likely to be contextually influenced.

SUBJECTIVE CONTROL AMONG OLDER MEXICAN-ORIGIN INDIVIDUALS

Although the physical and psychological benefits of an internal locus of control are well documented in the general U.S. population, little is known about how locus of control is affected by culture, economics, and social circumstances (Menec & Chipperfield, 1997). Few international comparative studies exist. We propose a conceptualization of successful aging that places material well-being at the center of the association among resources, locus of control, and health outcomes. This conceptualization is based on the general proposition that among groups with few material resources, feeling that one has little control over one's life, including little control over one's health, reflects the reality of powerlessness and limited access to health care. As part of a learned helplessness process, such a situation may well result in a fairly stable orientation similar to a personality trait. Theoretically and practically, self-efficacy or subjective control over various aspects of one's life can be seen as a learned general or global self-conception that can be relatively stable over time and across situations but that can change when the situation changes markedly.

Our testable research proposition holds that one's expressed locus of control, in addition to symptoms of distress, self-reported poor health, and self-reports of functional impairment, will be greatly affected by material well-being reflected both in national context and in individual characteristics related to the ability to control one's life such as level of education, income, financial strain, and health insurance. We expect that older Mexican Americans will be more likely to express greater sense of control over important aspects of their health than Mexicans because of better access to primary and specialty health care that results from the nearly universal access to health care through Medicare.

DATA AND METHODS

In order to identify major predictors of locus of control and other health outcomes among elderly individuals in two different national contexts, we employ the Hispanic Established Populations for Epidemiologic Study of the Elderly (H-EPESE) in the United States and the Mexican Health and Aging Study (MHAS) with similar, but not identical, data in Mexico. Both surveys include detailed information on demographics, socioeconomic characteristics, migration history where applicable, and physical and mental health and functioning. Although they provide useful comparative information, the two data sets differ in important respects, and we focus more on similarities or differences in general patterns rather than on specific coefficients or prevalence rates. The H-EPESE is a longitudinal study based on an original probability sample of 3,050 Mexican-origin individuals aged 65 years and over in the five southwestern states of Arizona, California, Colorado, New Mexico, and Texas who were first interviewed in 1993/1994 (Markides, 1999). The baseline H-EPESE response rate was 86%. Although the MHAS consists of a national probability sample of 15,186 Mexicans aged 50 years and over interviewed in Mexico in 2001, we have selected respondents aged only 65 years and over. The baseline MHAS response rate was 90% (Wong & Espinoza, 2004). In both surveys, information on respondents who could not answer for themselves was provided by proxy respondents. These cases have been deleted from the analyses because of missing or invalid responses on selected outcome variables (MHAS, n = 449; H-EPESE, n = 316). H-EPESE respondents were given the option of taking the survey in Spanish, and more than three quarters (77.8%) did so.

From these two data sources, we created five groups: (a) Mexican residents with no history of residence in the United States (MHAS, n = 3,875), (b) U.S. residents who were born in Mexico and migrated to the United States between the ages of 1 and 19 years (H-EPESE, n = 383), (c) U.S. residents who were born in Mexico and migrated to the United States between the ages of 20 and 49 years (H-EPESE, n = 578), (d) U.S. residents who were born in Mexico and migrated to the United States between the ages of 50 and 90 years (H-EPESE, n = 232), and (e) U.S. residents who were born in the United States (H-EPESE, n = 1,541). We excluded 578 individuals in the MHAS who reported that they had lived or worked in the United States in order to isolate the MHAS reference group from significant exposure to U.S. systems and culture. Spanish and English versions of the health questions are presented in Appendix A.

MEASURES

Locus of Control: MHAS

The comparisons we are able to make are not exact because we do not have identical measures of locus of control in the two surveys. This shortcoming should be kept in mind in interpreting the results of the two surveys. In the MHAS, the focus is on the more general construct of locus of control as measured by eight items adapted from Rotter (1966). MHAS respondents were asked to rate the degree to which they agreed or disagreed with the following statements (see the Appendix for the Spanish version): (a) “There is no sense in planning a lot for the future”; (b) “The really good things that happen to me are mostly luck”; (c) “Most of one's problems are due to bad luck”; (d) “One has little control [over] the bad things that happen to him/her”; (e) “One is responsible for his/her own successes”; (f) “One can do just about anything he/she puts his/her mind to”; (g) “One's misfortunes are the result of his/her own mistakes”; and (h) “One is responsible for his/her own failures.” Response categories for these items were coded (1) disagree, (2) somewhat disagree, (3) somewhat agree, and (4) agree. We sum the eight locus of control items after reverse coding the external locus of control questions (e–h) and divide by 8 to create an index that ranges from 1 to 4. A higher score reflects greater internal locus of control and less of a sense that one is in control of his or her life.

Health Improvement Beliefs: MHAS

MHAS respondents were asked the following question to measure health improvement beliefs: “Do you think that a person your age can improve his/her health through regular exercise, balanced diet, or by stopping smoking?” Response categories for this item were coded (1) for yes and (0) for no.

Self-Care Concern: H-EPESE

H-EPESE respondents were asked the following question to assess self-care concern: “How concerned are you about being unable to be independent and take care of yourself and your affairs in the future?” Response categories for this item were coded (1) not concerned at all, (2) somewhat concerned, and (3) very concerned.

Health Sense of Control: H-EPESE

H-EPESE respondents were asked the following question to measure health sense of control: “To what extent do you feel you can control the general state of your health through your own actions?” Response categories for this item were coded (1) not at all, (2) somewhat, and (3) a great deal. In general, the H-EPESE questions are more general and nonspecific than the MHAS questions that ask about more specific behaviors such as control of health through regular exercise, balanced diet, or by stopping smoking.

Psychological Distress

We employ seven items from the Center for Epidemiologic Studies–Depression (CES-D) scale to measure psychological distress. In both surveys, respondents were asked whether during the past week they had felt (a) depressed, (b) that everything they did was an effort, (c) their sleep was restless, (d) unhappy, (e) lonely, (f) they did not enjoy life, or (g) sad. In order to increase the comparability of the CES-D between the two samples, the H-EPESE items were recoded (1) most or all the time and (0) otherwise. The final psychological distress measure represents a summed index of the seven items.

Health Risk Factors

Our measurement of health risk factors includes body mass, smoking, and drinking behavior. Using the standard formula and documented thresholds provided by the Centers for Disease Control, we coded body mass as (1) for obese, a body mass index of 30 or over, and (0) otherwise. In the MHAS, height and weight were reported by the respondent, whereas in the H-EPESE, respondents were measured and weighted by the interviewer. Smoking behavior is operationalized as pack years. We measure pack years by multiplying the number of years the respondent smoked by the average number of packs (20 cigarettes per pack) during that time.

To measure heavy drinking and drinking problems, we use items adapted from the CAGE instrument (the first letter of a key word in each question spells CAGE, e.g., cut down, annoyed, guilty, eye opener) (Ewing, 1984). In both surveys with slight variations in wording, respondents were asked: (a) “Have you ever felt you should cut down on your drinking?” (b) “Have people annoyed you by criticizing your drinking?” (c) “Have you ever felt bad or guilty about your drinking?” (d) “Have you ever had a drink first thing in the morning to steady your nerves or to get rid of a hangover (eye opener)?” Following the work of Saitz and colleagues (2007), in both surveys respondents who answered “yes” to any of the four questions were coded (1) for problem drinker and (0) otherwise, including individuals without any drinking problem and those who never drink.

Chronic Conditions

Our assessment of chronic conditions is a summative measure of six items worded slightly differently. Both sets of respondents were asked by the doctor whether they had ever had arthritis, diabetes, high blood pressure, a heart attack, a stroke, or cancer.

Self-Rated Health

Self-rated health was measured with single items in both surveys, with categories ranging from excellent to poor. In order to compare self-rated health across surveys, we recoded these items (1) fair or poor and (0) otherwise.

Disability

Our measure of disability assesses limitations in performing routine activities of daily living (ADLs). Respondents were asked to indicate whether they had any difficulties (because of a health problem) with the following activities: (a) bathing or showering, (b) eating, (c) getting in or out of bed, (d) using the toilet, (e) walking across a room, and (f) dressing. Original response categories were coded (1) for yes and (0) for no. All the items were summed to form an activities of daily living (ADL) index, with scores ranging from 0 to 7.

Financial Strain

Exposure to life stressors such as chronic economic strain can adversely affect psychological well-being (Krause, 1987). H-EPESE respondents were asked two questions to assess financial strain: (a) “How much difficulty do you have in meeting monthly payments on your bills?” Response categories for this item were coded (1) none, (2) a little, (3) some, and (4) a great deal. (b) “At the end of the month, do you usually end up with some money left over, just enough to make ends meet, or not enough to make ends meet?” Response categories for this item were coded (1) some money left over, (2) just enough to make ends meet, and (3) not enough money to make ends meet. We measure financial strain as a mean index of these two items. Note that these items have been standardized to account for metric differences. We assessed financial strain in MHAS respondents with a single item. Respondents were asked, “Would you say your financial situation is excellent, very good, good, fair, or poor? Because less than 2% of respondents could be categorized as either “excellent” or “very good,” these categories were combined with “good.” Thus, response categories for this item were coded (1) good, (2) fair, and (3) poor.

Insurance Status

Health insurance is an important factor in determining health care use in both Mexico and the United States (Wong, Diaz, & Espinoza, 2006). The most recent data available indicate that slightly over half of those over age 60 years have health care coverage of some sort (Wong et al.). Access as well as the quality of care received by Mexican people, including the elderly, depend on their employment status, the sector in which they work (e.g., informal economy, formal sector), and income. Those with the most complete access are individuals who were in salaried jobs prior to retirement. They are covered by various health funds that make up the social security system, which also includes smaller funds or other plans such as those sponsored by the military or Petróleos Mexicanos (PEMEX), the Mexican petroleum monopoly (Organisation for Economic Co-operation & Development, 2005). The public system for the elderly poor receives less funding and is clearly inferior to the system for salaried employees. Because of extensive poverty and inadequate coverage, many Mexicans, including the elderly, face serious barriers to health care. Formal long-term care is unavailable, and the infirm elderly must rely on their families when they can no longer care for themselves. Many of the most seriously underserved Mexicans live in rural areas or in the poorer states of the South. In order to improve access for the poor in 2004, Mexico introduced a new program entitled Seguro Popular (Public Insurance), which is intended to extend health insurance to the nation's 50 million uninsured by 2010. This program may well improve health care, which today is often unavailable and of inferior quality, especially for rural residents and residents of the southern states.

H-EPESE respondents were classified as having health insurance if they reported having Medicare, Medicaid, or private insurance. H-EPESE respondents were also classified as having health insurance if they reported receiving Social Security or Supplemental Security Income, which qualifies respondents for Medicare. MHAS respondents were classified as having health insurance if they reported receiving coverage from any of the following sources: Mexican Social Insurance Institute, Social Services and Security Institute for State Employees, PEMEX, private insurance, or some other form of insurance.

Sociodemographic Characteristics

We control for several known sociodemographic correlates of health status, including age, sex, marital status, educational attainment, and personal income in the Mexican-origin population (Angel, Angel, & Hill, 2008). Like non-Hispanic whites, as older Mexican-origin people age, they may feel less in control of their own lives as the result of infirmity. In addition, education may boost the sense of one's control (Mirowsky, 1995). We categorized education to reflect meaningful differences in the samples. In the MHAS, education is coded into three categories: (a) no formal years of education, (b) 1–5 years of formal education, and (c) 6 or more years of formal schooling. In the H-EPESE, we create a dummy variable for education (less than high school as reference category) to reflect the limited educational attainment of older Mexican Americans in the United States. We include sex, coded (1) for females and (0) for males, because women may have a lower sense of control than men do (Ross & Mirowsky, 2002). Age is coded into three categories: (a) 65–69, (b) 70–79, and (c) 80 and over. Marital status is coded (1) for currently married and (0) otherwise. Because the income distributions in the United States and Mexico are so different, personal income is divided into three categories in each country in order to capture relative income levels in order to differentiate very poor households from those with higher incomes.

RESULTS

To restate the logic of the analysis, although one's beliefs concerning his or her locus of control no doubt reflect innate or socially learned traits, empirical measures of control also reflect an individual's accurate assessment concerning how much control an individual has over aspects of his or her life in specific contexts (Krause, 1986). In this section, we predict the locus of control indicators in Mexico and the United States in stratified analyses. The predictor equations include the demographic, migration, and health measures as independent variables to determine how each affects the locus of control variables. Table 1 presents descriptive statistics for the Mexican and U.S. samples. Both samples are roughly similar in their age distributions, proportion female, and percent married. Although neither sample is highly educated, educational levels are much lower in Mexico, and a far smaller fraction of the Mexican sample has health insurance. Over one third (36%) of the MHAS sample has no formal education compared with 16% of the H-EPESE; 10% of H-EPESE respondents report at least some college. None of the MHAS respondents have any college education. The H-EPESE sample has very low income by U.S. standards. Finally, slightly over half of H-EPESE respondents were born in the United States.

Table 1.

Descriptive Statistics by Country of Residence (range in parentheses)

Mexican Elderly Aged 65 Years and Older	Mexican Resident (MHAS 2001), n = 3,875		U.S. Resident (H-EPESE 1993), n = 2,734
Age (years)
65–69	—	38.0	—	41.0
70–79	—	46.0	—	44.0
80 and older	—	16.0	—	15.0
Female	—	57.0	—	58.0
Married	—	51.0	—	56.0
Education
No formal education	—	36.0	—	16.0
1–5 years	—	37.0	—	44.0
6 or more years	—	27.0	—	40.0
≥12 years of education	—	—	—	10.0
Personal income
Low	—	35.0	—	35.0
Middle	—	35.0	—	51.0
High	—	30.0	—	14.0
Financial strain index	2.0 (1–3)	—	0.0 (−2 to 2)	—
Currently insured	—	62.0	—	97.0
Nativity/migration age
United States/—	—	—	—	56.0
Mexico/1–19	—	—	—	14.0
Mexico/20–49	—	—	—	21.0
Mexico/50–90	—	—	—	9.0
Psychological distress	2.7 (0–7)		0.6 (0–7)
Control beliefs
Locus of control	3.0 (1–4)	—	—	—
Health improvement	—	87.0	—	—
Self-care concern	—	—	1.9 (1–3)	—
Health sense of control	—	—	2.4 (1–3)	—
Psychological distress	2.7 (0–7)		0.6 (0–7)
Fair/poor self-rated health	—	70.0	—	60.0
Health risk factors
Obese	—	18.0	—	30.0
Pack years	1.2 (0–15)	—	2.2 (0–15)	—
Problem drinker	—	15.4	—	13.0
Chronic conditions index	1.0 (0–5)	—	1.3 (0–6)	—
ADL disability index	0.3 (0–2)	—	0.2 (0–2)	—

Note: MHAS = Mexican Health and Aging Study; H-EPESE = Hispanic Established Populations for Epidemiologic Study of the Elderly.

Table 1 also presents information on control beliefs and the prevalence of health problems in each group. In the MHAS, locus of control ranges from 1 to 4, with a higher score reflecting a greater sense of control. The mean of 3 reflects a relatively high sense of control. On the second measure, which asks, “do you think a person your age can improve his/her health through regular exercise, balanced diet, or by stopping smoking?”, 87% of Mexicans responded positively. In the H-EPESE, the measures self-care concern ranges from 1 to 3, and a higher score reflects a greater sense that one will not be able to be independent and take care of oneself in the future. The second measure, health sense of control also ranges from 1 to 3, and a higher score reflects the sense that one will be able to control one's health in the future. Of those who responded, most expressed a moderate sense of concern over their capacity to maintain independence (1.9) and relatively high sense of control over health (2.4). MHAS respondents report almost three times more distress (2.7 vs. 0.6) than H-EPESE respondents, and they were more likely to report fair or poor general health (70% vs. 60%). Older Mexican Americans, on the other hand, were nearly twice as likely as Mexican residents to report obesity and reported more pack years of smoking. A somewhat smaller percentage of the Mexican sample reported problem drinking. Both samples were relatively similar in chronic conditions and ADL disability.

We next turn to multivariate analyses of the four locus of control measures on the demographic and health indicators for the two samples. Two models are presented for each locus of control outcome, and each column is numbered for reference. The first model for each subjective control outcome controls for demographic and economic factors (Model 1) and the second introduces the psychological, health, and health behavior variables (Model 2). Table 2 presents the results for the MHAS and Table 3 for the H-EPESE.

Table 2.

OLS and Logistic Regressions of Control Beliefs: MHAS

	Mexico Resident (MHAS 2001), n = 3,875
	Greater Internal Locus of Control		Ability to Improve Health
	Model 1	Model 2	Model 1	Model 2
Age (years)
65–69	—	—	—	—
70–79	−0.04	−0.03	−0.34	−0.36
80 and older	−0.06	−0.05	−0.43	−0.45
Female	−0.02	−0.02	−0.03	−0.06
Married	−0.01	−0.01	−0.13	−0.10
Education
No formal education	—	—	—	—
1–5 years	0.07*	0.07*	0.19	0.20
6 or more years	0.23***	0.22***	1.04***	1.06***
Personal income
Low	—	—	—	—
Middle	−0.03	−0.04	0.03	0.02
High	0.05	0.04	0.02	0.05
Financial strain index	−0.02	−0.01	0.22	0.23
Currently insured	0.06*	0.06*	0.27	0.36
Psychological distress	—	−0.00	—	0.19**
Fair/poor self-rated health	—	−0.08**	—	−0.29
Health risk factors
Obese	—	0.06*	—	−0.22
Pack years	—	0.00	—	−0.05
Problem drinker	—	−0.00	—	0.13
Chronic conditions index	—	0.01	—	−0.23*
ADL disability index	—	0.02	—	−0.00
Intercept	2.98	3.02	—	—
R2 (Nagelkerke)	.08	.09	(.04)	(.07)

Note: OLS = ordinary least squares; MHAS = Mexican Health and Aging Study; ADL = activity of daily living.

Table 3.

Ordered Logistic Regressions of Control Beliefs: H-EPESE

	U.S. Resident (H-EPESE 1993–1994), n = 2,734
	Greater Self-Care Concern		Greater Sense of Health Control
	Model 1	Model 2	Model 1	Model 2
Age (years)
65–69	—	—	—	—
70–79	−0.06	−0.09	−0.31***	−0.24*
80 and older	0.22	0.08	−0.59***	−0.42**
Female	0.38***	0.07	−0.04	−0.08
Married	0.21*	0.26**	−0.05	−0.02
Education
≥12 years of education	−0.13	−0.13	0.16	0.04
Personal income
Low	—	—	—	—
Middle	0.09	0.07	0.19*	0.20*
High	−0.02	−0.08	0.16	0.24
Financial strain index	0.47***	0.39***	−0.24***	−0.17**
Currently insured	0.02	−0.06	−0.22	−0.30
Nativity/migration age
United States/—	—	—	—	—
Mexico/1–19	0.03	−0.03	0.19	0.22
Mexico/20–49	−0.42***	−0.47***	−0.03	−0.07
Mexico/50–90	−0.65***	−0.72***	−0.00	−0.10
Psychological distress	—	0.37***	—	0.02
Fair/poor self-rated health	—	0.14	—	−0.48***
Health risk factors
Obese	—	−0.09	—	−0.01
Pack years	—	−0.03*	—	−0.03*
Problem drinker	—	−0.53***	—	−0.15
Chronic conditions index	—	0.13***	—	−0.12**
ADL disability index	—	0.38***	—	−0.21*
R2 (Nagelkerke)	(.08)	(.15)	(.03)	(.07)

Note: H-EPESE = Hispanic Established Populations for Epidemiologic Study of the Elderly; ADL = activity of daily living.

In Table 2, the first variable we examine is Internal Locus of Control, and the results are presented using unstandardized ordinary least squares regression coefficients in the first two columns. In Model 1, education and having health insurance increase one's internal locus of control. In Model 2, in which we add the health-related predictors, fair or poor self-reported health reduces one's internal locus of control, but obesity increases it. Columns 3 and 4 present results for the logistic regressions (binary logit) related to the sense that one can improve one's health. In Model 1, as was the case for internal locus of control, education is a strong predictor of a greater sense of control. None of the other demographic predictors are significant. In Model 2, education remains highly significant, and psychological distress emerges as a positive predictor of one's sense of one's ability to improve health. On the face of it, one might expect distress to undermine one's sense of control, but in this case, it may make the issue of control more salient. In Model 2, chronic conditions decrease one's ability to improve health for older Mexicans.

Table 3 displays the results for the H-EPESE sample employing ordered logistic regression procedures. Columns 1 and 2 present logit coefficients for self-care concern, and columns 3 and 4 one's sense of control over health. In Model 1, for self-care concern, females and married people report greater concern over their capacity to care for themselves in the future. In addition, financial strain is strongly correlated with concerns over one's ability to remain independent. Because nativity and age at migration are such important factors in the Mexican American experience, we control for these variables for the U.S. sample. Native born is the reference category, and the results indicate that older age at migration decreases self-care concerns. Immigrants who arrive later in life are more likely to live with and be dependent on family, and this fact may make the issue of control irrelevant. Model 2 introduces the psychological, health, and health behavior variables. The introduction of these variables has little effect on the demographic and financial strain variables other than to drop female below the level of significance. Of the psychological variables, higher distress is associated with greater self-care concern. In addition to psychological distress, physical health problems and ADL disability increase self-care concerns. In this model, smoking and problem drinking decrease concerns about one's ability to care for oneself. Such behaviors may reflect a general lack of concern with health. In the HLC model, age is statistically significant.

Columns 3 and 4 of Table 3 present results for one's sense of control over one's health. In Model 1, older age decreases one's sense of control. Personal income in the middle range increases one's sense of control relative to those in the lowest category. Although the patterns for the highest income category are similar to those for middle income, they are not significant. Financial strain significantly increases self-care concern. In the U.S. sample, health insurance is not significant in any model. In this sample, Medicare coverage is basically universal.

Model 2 introduces the psychological, health, and health behavior variables for HLC. In this model, fair or poor self-rated health is associated with a lower sense of control over one's health. Of the health behaviors, only smoking has a significant effect in reducing concern over the ability to control one's health. The presence of chronic health conditions reduces one's sense of control, as does the presence of functional problems (ADL index).

DISCUSSION

Our findings build upon and extend a growing body of research that investigates issues of aging, locus of control, and health (e.g., Krause, 2007; Lachman and Weaver, 1998a, 1998b; Norman et al., 1998; Ross & Mirowsky, 2002; Shaw & Krause, 2001; Wickrama et al., 2008) by examining issues of a sense of control in a comparative context. Our analyses of two culturally similar samples of older individuals in two very different national contexts reveal that social characteristics associated with poverty and social marginalization result in a lower sense of control over important aspects of one's life, including one's health. The findings also indicate that access to health insurance, which is nearly universal for elderly Americans in the form of Medicare, is an important factor in removing one's sense of insecurity and a lack of control over one's health. In Mexico, where health care coverage among the elderly is not universal, access to health care coverage directly affects one's ability to obtain care (Wong et al., 2006) and is therefore associated with one's subjective sense of control.

Our findings suggest that universal health care coverage would have clear psychological benefits in addition to the practical benefits associated with access to care. From a public policy perspective, the contribution of health coverage to psychological as well as physical health has important implications for the well-being of older individuals given the increasing need for medical care that accompanies aging. If health problems are not treated or not treated adequately, the quality of individual lives can be undermined and the eventual long-term care burden on families and society increased. The adequate control of chronic illnesses and the minimization of their negative effects represent core objectives of primary care for the elderly. In Mexico, the inability of many older individuals to avoid or reduce the consequences of chronic illnesses objectively reduces their control over their health (Angel et al., 2008). This fact has important implications for comparative research in which subjective health-related, or really any other subjective, information is collected.

Although human beings share a common evolutionary history and pass through the life course on the basis of a basic set of similar developmental possibilities, the actual experience of aging is strongly conditioned by the material, political, and social contexts in which one lives one's life. We began this study by calling upon Abraham Maslow's humanistic theory of human development to emphasize the fact that the possibilities for optimal development depend on an adequate material foundation. We end by calling upon another classic author, T. H. Marshall, who over half a century ago characterized the modern welfare state as the culmination of three centuries of the consolidation of basic human rights that define full citizenship (Marshall, 1950). During the 18th century, citizens, or at least white European males, gained basic legal-civil rights; during the 19th century, they gained greater political rights; and during the 20th century, citizens of the modern welfare state won social rights, which include not only the right not to be detained without cause and the right to vote but also the right to full participation in the economy and to partake of the material necessities of a civilized and productive life.

Marshall's characterization has been criticized for referring only to Europe or, perhaps, even just Great Britain. Yet it expresses an ideal of equity in which all citizens have the right to the necessities of a dignified life. Clearly, many groups in the developing world, and even many in the developed world, including racial and ethnic minorities and colonized indigenous groups, have not achieved full political, let alone social, rights. For them, the route to self-actualization and the generativity that comes with economic and social security remains elusive. In this paper, we have raised issues related to the role of national context in determining access to adequate material supports and the impact of those differences on one's sense of control over one's life for older people in the United States and Mexico. The construct of locus of control has been applied usefully to health. Such complex constructs are clearly important in helping us understand health and illness behavior, and they can even be useful in designing public health interventions. The contribution of this discussion to the literature is to emphasize the central importance of structural factors related to economics and the health care system on what we might see as purely psychological phenomena. The issue is particularly salient as the amount of comparative research expands and as countries, states, and regions with very different socioeconomic and political, in addition to cultural, profiles are compared.

Today, the best practices in comparative research focus on careful questionnaire construction and translation. We have clearly learned much about the best ways to ask complex questions concerning issues that are affected by different cultural contexts. In addition to measurement, though, conceptual modeling and the treatment of the context itself remain problematic. Although most comparative researchers are well aware of the complexities that different cultural, political, and economic contexts introduce into the response task, few analyses, especially those based on secondary data, specifically discuss the context, including the nature of the health care financing and delivery systems, in which the study takes place in detail.

Given the differences in survey design and variable definition in our Mexican and Mexican American samples, the degree of similarity in patterns between the two surveys seems remarkable. The differences that emerge largely reflect differences in economic and social contexts and the reality of one's ability to exercise control over one's life. There can be little doubt that in addition to structural factors, psychological factors affect one's sense of control. Yet it seems imperative to emphasize that psychological traits or states, especially as they are measured using survey techniques, may themselves reflect structural and economic factors. Individuals who do not have health insurance and have not been to the doctor most likely do not know that they have diabetes and they may be misinformed about what they need to do to deal with symptoms (Angel et al., 2008). Contact with health professionals educates one in terms of real capacities for self-care and can affect one's subjective sense of control.

For the most part, our results reflect what one would expect in the association among health and one's sense of control over one's health. Poor health reduces one's sense of control in both countries but not for all outcomes. Chronic conditions increase concern over health and decrease one's sense that he or she has control over health in the U.S. sample. Health-related behaviors are not just the result of personal choices but are affected by social inequalities and differences in access to education, housing, jobs, and human services. As Jervis, Beals, Fickenscher, and Arciniegas (2007) argue, these are important considerations to take into account in establishing optimal methods for assessing the cultural relevance of measures of cognitive impairment in ethnically diverse older populations. The patterns that emerge from this analysis suggest that the measurement of psychological characteristics in countries that differ greatly in level of economic development, health care system characteristics, and social class structure must deal with more than psychometrics and translation. In order to understand health in the broadest sense, the context in which health is constructed cognitively and emotionally must also be understood. Healthful aging and one's sense of control are clearly complex and multifaceted constructs (J. L. Angel & R. J. Angel, 2006). As we discussed in the introduction, successful aging has been characterized in terms of psychological processes and outcomes, and indeed, it would be hard to deny that a deep sense of satisfaction and generativity, perhaps summarized by the term self-actualization, represents the ultimate end point of the successful life course. It is also quite clear that individuals arrive in life with unique personalities, capacities, talents, and different levels of physical vitality and health. These represent each person's unique human capital that he or she employs as a rational actor in the game of life to achieve personal and collective goals. Yet each of us also arrives in a different social location which determines the possibilities we inherit in terms of material capital and opportunities, as well as the social stigma often associated with specific categories of race, ethnicity, gender, religion, tribe, or caste. In terms of the material resources that one has at one's disposal as the result of the luck of one's birth, life may simply not be equitable. The interaction of individual personalities and capacities and the social structures within which individuals express their potential represent the traditional structure versus agency problem in western thought. Psychologists focus most heavily on the personality traits and states, and organizational and political theorists focus most heavily on the structural. Gecas (1986) believes self-efficacy consists of three dimensions that represent self-concept, combined with self-esteem and self-authenticity. From this perspective, self-efficacy or subjective control is viewed as part of a general or global self-evaluation that tends to be stable across situations but that can change over time depending upon one's specific circumstances. Unfortunately, our data do not allow us to determine how an individual's sense of subjective control varies as he or she ages.

Both psychologists and sociologists, of course, recognize the legitimacy and importance of one another's level of analysis, but the focus on one or the other tends to dominate individual theories and research. Psychologists clearly dominate the study of successful aging. The objective of this research has been to emphasize the operational difficulties inherent in attempting to isolate psychological traits from situational factors. Personality characteristics and traits clearly exist and lead people to respond differently to different situations. Yet the situations in which such traits are expressed have independent or more likely interactive effects. The importance of such an observation, which may even seem obvious, is to warn against the temptation to attribute excessive causal force to psychological factors in situations in which poverty, violence, and general social disorganization reign. The widespread suspicion of culture of poverty explanations for the multigenerational persistence of the social disorganization characteristic of some poor communities and neighborhoods arises from the fear that the effects of structural factors are being misattributed to culture. Such misattributions are certainly epistemologically incorrect, but they are also potentially quite damaging in attributing to individual and group pathologies the negative effects of oppressive structural disadvantage. This discussion and analysis has focused on the material aspects of aging and on social structure in the determination of the physical and emotional well-being of older individuals in two very different economic and political contexts. Given the differences in those contexts and the fact that the two surveys on which the analyses are based do not have identical measures, we have avoided the direct comparison of specific effect sizes. Given the need for translation and the fact that comparative studies cannot eliminate the effects of structural and cultural differences, our position is that specific cardinal comparisons may be unjustifiable whatever one does (Angel, 2006). Despite these limitations, the strength of this study lies in the development of a heuristic model and the call for a closer examination of structure in comparative research dealing with subjective states.

CONCLUSION

For the topic of successful aging and the comparative study of health, generally, our message is clear and simple. Although it is clear that the ultimate potential for human development over the life course can be defined in terms of personal satisfaction, social engagement, effectiveness, generativity, and more, we cannot discount or forget the central role of material well-being. Effective social security and health care delivery systems for the old, as well as full employment and educational systems for the young and middle aged, are the bedrock of successful aging. These are also the bedrock of the modern welfare state, which is in a period of neoliberal and now post-neoliberal reform. For the developing nations of the world, as well as for the developed nations, providing the material basis for successful aging in a period of fiscal austerity and rapidly aging populations presents perhaps insurmountable challenges. With the dignity and peace of mind that material security brings, a person is far more likely to develop to the highest psychological level possible. In the study of successful aging, then, and especially in comparative studies that include different cultures, social classes, races, and ethnicities, the material basis of successful aging, as well as the economic, political, and organizational factors that structure the opportunities that determine one's place in the social hierarchy, must be directly addressed.

FUNDING

This research was funded, in part, from National Institutes of Health/National Institute on Aging (NIH/NIA) RO1 AG10939.

Appendix

Coding

Question	Item	Control/Concern Index
MHAS
(1) External locus of control—Adapted from Rotter (1966)
I'm going to read what people say sometimes about their lives. For each statement, please tell me if you: Agree, Somewhat Agree, Somewhat Disagree, or Disagree	19a There's no sense in planning a lot for the future	Lower external control = 19a + 19b + 19e + 19f
	19b The really good things that happen to you are mostly luck
	19e Most of one's problems are due to bad luck
	19f One has little control on the bad things that happen to him/herself
Enseguida le voy a leer cosas que las gente dice a veces. Para cada una de las expresiones, por favor digame si Ud.: Está de acuerdo, algo de acuerdo, algo en desacuerdo, o en desacuerdo	19a No tiene case planear mucho para el futuro
	19b Las cosas muy buenas que le suceden a uno son por buena suerte
	19e La Mayoria de los problemas se deben a la mala suerte
	Uno tiene poco control sobre las cosas malas que le suceden
(2) Internal locus of control—Adapted from Rotter (1966)
I'm going to read what people say sometimes about their lives. For each statement, please tell me if you: Agree, Somewhat Agree, Somewhat Disagree, or Disagree	19c One is responsible for his/her own successes	Higher internal control = 19c + 19d + 19g + 19h
	19d One can do just about anything he/she put his/her mind to happen to him/her
	19g One's misfortunes are the result of his/her own mistakes
	19h One is responsible for his/her own failures
Enseguida le voy a leer cosas que las gente dice a veces. Para cada una de las expresiones, por favor digame si Ud.: Está de acuerdo, algo de acuerdo, algo en desacuerdo, o en desacuerdo	19c Uno es responsible de sus propios exitos
	19d Uno puede hacer casi cualquier cosa que se proponga
	19g Las desgracias que le ocurren a uno son resultado de sus errores
	19h Uno es responsable de sus propias fallas
(3) Health locus of control (Wallston et al., 1978)
Do you think that a person your age can improve his/her health through regular exercise, balanced diet, or by stopping smoking?	Yes ..............1	Dichotomy (1,0)
	No ................2	Higher health control
¿Usted free que una persona de sue dad pueda mejorar su salud con ejercicio regular, alimentacion adecuada, o dejar de fumar?	Si ....................1
	No ...................2
H-EPESE
(4) Locus/coherence (general; Antonovsky, 1979) of control
How concerned are you about being unable to be independent and take care of yourself and your affairs in the future?	(1) Not concerned at all	Greater self-care concern
	(2) Somewhat concerned
	(3) Very concerned
¿Qué tan preocupado(a) está usted de no ser independiente ó poder atender sus asuntos personales en el futuro? Diria usted gue esta.	(1) Muy preocupado
	(2) Un poco preocupado
	(3) No se preocupa
(5) Health locus of control (Wallston et al., 1978)
To what extent do you feel you can control the general state of your health through your own actions?	(1) Not at all	Higher health control
	(2) Somewhat
	(3) A great deal
¿A cuál grado siente usted que puede controlar el estado general de su salud a través de sus propias acciones	(1) Mucho
	(2) Un poco
	(3) De ninguna manera